Descargar el fichero

Departamento de Física de la Tierra, Astronomía
y Astrofísica I (Geofísica y Meteorología).
Facultad de Ciencias Físicas
Universidad Complutense de Madrid
Desarrollo de modelos numéricos para investigar la isla de
calor en ciudades y estudio de la sensibilidad de distintos
parámetros urbanos
Development of numerical models to investigate the urban
heat island in cities and sensitivity study of different urban
parameters
Francisco Salamanca Palou
Memoria de Tesis presentada para optar al grado de Doctor
Madrid 2010
Departamento de Física de la Tierra, Astronomía
y Astrofísica I (Geofísica y Meteorología).
Facultad de Ciencias Físicas
Universidad Complutense de Madrid
Desarrollo de modelos numéricos para investigar la isla de
calor en ciudades y estudio de la sensibilidad de distintos
parámetros urbanos
Development of numerical models to investigate the urban
heat island in cities and sensitivity study of different urban
parameters
Francisco Salamanca Palou
Memoria de Tesis presentada para optar al grado de Doctor
Dirigida por
Dr. Alberto Martilli
Dr. Carlos Yagüe Anguís
Madrid 2010
A mis padres Antonio y Juana
A mis hermanos Bernardino y Candy
A Blanca
AGRADECIMIENTOS
Una tesis doctoral es difícil de llevar a cabo sin el apoyo de un buen equipo humano.
Por este motivo quiero aprovechar estas pocas líneas para dar las gracias a todas aquellas
personas que me han ayudado durante estos cuatro años en la realización de este trabajo.
Quisiera agradecer:
•
Al Dr. Fernando Martín del Centro de Investigaciones Energéticas,
Medioambientales y Tecnológicas (CIEMAT) por haberme acogido en el
grupo de modelización atmosférica que él dirige.
•
Al Dr. Alain Clappier de la École Polytechnique Fédérale de Lausanne
(EPFL), por hacer que mis dos visitas a Suiza fueran más agradables y
fructíferas.
•
Al Dr. Andrea Krpo por ofrecerme su amistad y compartir parte del trabajo
realizado en esta tesis.
•
Al Dr. Fei Chen del National Center for Atmospheric Research (NCAR) por
hacer que mis dos estancias en los EEUU hayan sido inolvidables y muy
provechosas.
•
Al Dr. Mukul Tewari del NCAR por el interés mostrado en mi trabajo y por
poner a mi disposición su gran experiencia y conocimiento.
•
Al Dr. José Luís Santiago por brindarme su amistad desinteresada. Los viajes,
congresos, y momentos compartidos en el CIEMAT han sido inolvidables.
•
A los nuevos compañeros del grupo de modelización del CIEMAT, Manuel
Santiago, Juan Luís Garrido y María Ángeles González por compartir muchos
momentos agradables en esta mi última etapa de la tesis.
•
A mi querido amigo Héctor Hernández con el que inicié esta aventura hace ya
casi cuatro años. Aunque trabajamos en temas totalmente distintos, mucho es
lo que nos une.
•
A la Dra. Begoña Aceña del CIEMAT por su sentido del humor y ofrecerme su
amistad.
Quisiera profundamente agradecer al Dr. Carlos Yagüe de la Universidad
Complutense de Madrid (UCM) sus valiosos consejos y experiencia, que siempre han
contribuido a mejorar el trabajo realizado.
Finalmente una mención especial merece el Dr. Alberto Martilli del CIEMAT. Alberto
me dio la oportunidad de hacer esta tesis hace ya cuatro años. Confió en mí desde el primer
momento y es la persona con la que he ido creciendo en el mundo de la modelización
atmosférica desde entonces. La comunicación con los ordenadores no siempre me resultó
sencilla y Alberto me enseñó a hacerlo. Este trabajo sin él no hubiera sido posible. A él le
debo muchísimo de lo aprendido y vivido durante esta etapa. Gracias por tu amistad y por
compartir conmigo estos maravillosos años.
Por último, gracias a mis padres y hermanos, que son mi apoyo, y en especial a Blanca
por compartir estos inolvidables años conmigo y mostrar siempre un gran respeto hacia mi
trabajo.
Este trabajo ha sido financiado a través de una beca para formación de personal investigador
(BOE nº 191 de 11 de Agosto de 2005) y a través del proyecto “Simulación a mesoescala del
clima urbano y desarrollo de una técnica de evaluación de estrategias de reducción de la isla
de
calor
urbana”
del
Ministerio
de
Medio
Ambiente
de
España
(expediente
200800050084408).
This work has been funded through a research fellowship (BOE nº 191 published 11th August
2005) and thanks to the project “Mesoscale simulations of urban climate and development of
an evaluation technique of urban heat island mitigation strategies” funded by the Ministry of
Environment of Spain (file 200800050084408).
ÍNDICE
RESUMEN……………………………………………………………………………………1
ORGANIZACIÓN DE LA MEMORÍA…………………………………………………….7
SUMMARY………………………………………………………………………………...…9
ORGANIZATION OF THE THESIS……………………………………………………...15
1. ESTADO DEL ARTE Y DESCRIPCIÓN DEL TRABAJO...………………...……….17
1.1 Introducción………………………………………………………………………………18
1.2 La capa límite atmosférica………………………………………………………………..19
1.2.1 La capa límite urbana……………………………………………………………….23
1.3 Flujos turbulentos y balance energético superficial………………………………………24
1.4 Modelización a mesoescala. Ecuaciones que gobiernan los flujos turbulentos……..........27
1.4.1 Modelización de la capa límite atmosférica sobre zonas urbanas. Parametrizaciones
urbanas……………………………………………………………………………………32
1.4.2
El
modelo
energético
(BEM)
para
simulaciones
del
clima
urbano……………………………………………………………………………………..36
1.4.3
Problemas
abiertos
en
las
simulaciones
del
clima
urbano
a
mesoescala………………………………………………………………………………..39
1.5
El
modelo
atmosférico
WRF.
Parametrizaciones
urbanas
en
WRF…………………………………………………………………………...………..…….40
1.5.1
La
base
de
datos
NUDAPT.
La
versión
de
WRF
adaptada
a
NUDAPT………..………………………………………………………………………..43
APÉNDICE…………………………………………………………………………………...47
2. OBJETIVOS………………………………………………………………………………50
2.1 Introducción………………………………………………………………………………51
i
2. OBJECTIVES……..……………………………………………………………………...57
2.1 Introduction……………………………………………………………………………….58
3. RESULTADOS……………………………………………………………………………63
(Publications by the author)
3.1 Un nuevo modelo energético de edificios acoplado a una parametrización urbana para
simulaciones del clima urbano-Parte-I. Formulación, verificación y análisis sensitivo del
modelo…………………………………………………………………………………...……64
3.2 Un nuevo modelo energético de edificios acoplado a una parametrización urbana para
simulaciones del clima urbano-Parte-II. Validación con simulaciones (off-line) en una
dimensión vertical…………………………………………………………………………….80
3.3 Derivación de las propiedades térmicas de un material representativo de un área
heterogénea de la ciudad……………………………………………………………………...94
3.4 Un estudio de la capa límite urbana usando diferentes parametrizaciones urbanas y
resoluciones de parámetros morfológicos que describen una ciudad con el modelo atmosférico
WRF…………………………………………………………………………………………104
3.5 Estudio numérico de la capa límite urbana sobre la ciudad de Madrid durante la campaña
DESIREX (2008) con WRF y evaluación de diferentes estrategias de mitigación de la isla de
calor urbana………………………………………………………………………………….132
4. DISCUSIÓN INTEGRADORA………………………………………………………...154
4. DISCUSSION…………………………..………………………………………………..167
5. CONCLUSIONES Y FUTURAS LÍNEAS DE INVESTIGACIÓN……………….....179
5.1 Conclusiones…………………………………………………………………….180
5.2 Futuras líneas de investigación……………………………………………….…183
5. CONCLUSIONS AND FUTURE RESEARCH LINES………………………………186
ii
5.1 Conclusions……………………………………………………………………...187
5.2 Future research lines……………………………………………………………..189
REFERENCIAS……………………………………………………………………………192
APPENDIX A……………………………………………………………………………....198
Abstracts of publications co-authored by the author
APPENDIX B………………………………………………………………………………203
Numerical treatment to solve the heat diffusion equation using an energetic balance as
boundary conditions.
CV………………………………………………………………………………….………..207
iii
iv
RESUMEN
Siempre motivado por lograr una mejor calidad de vida, el ser humano a lo largo de la
historia se ha caracterizado por sus continuos desplazamientos migratorios. Gracias al
desarrollo económico, social e industrial ligado a la vida moderna, asistimos a un imparable
crecimiento de la población en las zonas urbanas. Mientras las zonas rurales van perdiendo
habitantes, las ciudades van incrementando su número de forma progresiva. Actualmente la
mitad de la población mundial vive en ciudades y se estima que en las próximas décadas este
porcentaje aumente hasta las tres cuartas partes. Este fenómeno social ha derivado en la
aparición de grandes núcleos urbanos con poblaciones que exceden de los diez millones de
habitantes (megacities). El número de megacities en el mundo crece de forma rápida y
consecuentemente se acentúan los problemas asociados a la vida en las grandes ciudades.
La sustitución del suelo rural por los materiales de construcción, la emisión de
contaminantes debida al tráfico y a la actividad industrial, el calor antropogénico generado
por las diferentes actividades humanas, el gran consumo energético asociado a las necesidades
de regulación térmica interna en los edificios, etc., son factores que modifican el clima, no
solo a escala local sino que también a escala global (Oleson et al., 2010). No es de extrañar
que el interés por los problemas derivados de la vida moderna en las ciudades vaya creciendo
y despierte cada vez más la curiosidad
en la comunidad científica. Dos problemas
fundamentales que en los últimos años han centrado de forma notable el interés de los
estudiosos de la atmósfera son la conocida isla de calor urbana y la polución atmosférica.
El sistema de ecuaciones diferenciales que describen los flujos y movimientos
atmosféricos son altamente no lineales y requieren de técnicas numéricas para su estudio.
Gracias al aumento creciente de la capacidad de cálculo de los ordenadores y al desarrollo de
los modelos numéricos, hoy podemos estudiar y comprender mejor el impacto de las ciudades
1
en la atmósfera y en el sistema climático en general.
Esta memoria está enfocada fundamentalmente al estudio de la isla de calor que tiene
lugar en las ciudades con el fin de lograr una mayor comprensión de este fenómeno, así como
la evaluación de diferentes estrategias para su mitigación. No dejaremos de lado el estudio del
consumo energético asociado al uso de los aires acondicionados, así como su impacto en la
temperatura del aire. La evaluación de estrategias de mitigación tanto de la isla de calor
urbana como del consumo energético serán los ejes principales de este trabajo.
La idea es contribuir al desarrollo de una herramienta numérica capaz de calcular de
forma cuantitativa los efectos de las ciudades en la atmósfera, para poder así diseñar y evaluar
estrategias de mitigación de la isla de calor y del consumo energético asociados al crecimiento
y desarrollo de las ciudades.
En primer lugar se ha procedido al desarrollo de un modelo energético de edificios
(Building Energy Model, BEM) para poder estudiar los flujos de calor intercambiados
(sensible/latente) entre los edificios y la atmósfera. Estudios recientes indican (Kondo &
Kikegawa, 2003; Ohashi et al., 2007) que los flujos de calor procedentes de los edificios
(concretamente los originados por los sistemas de aire acondicionado) podrían tener un efecto
importante en la temperatura del aire y deberían ser considerados en los estudios del clima
urbano. En este nuevo modelo energético un edificio es considerado como una “caja”
compuesta por un apilamiento de plantas (pisos) y para cada planta se resuelven de forma
explícita:
•
la difusión del calor a través de las paredes, suelos y tejados,
•
la ventilación natural, así como la reflexión y emisión radiativa que tiene lugar
entre las superficies interiores del edificio,
•
el calor generado por los equipos domésticos y las personas,
2
•
el flujo de calor intercambiado por los sistemas de aire acondicionado y el
exterior.
Edificios con distinto número de niveles (plantas) pueden considerarse en el modelo y la
evolución temporal de la temperatura y humedad interiores son calculadas separadamente
para cada nivel.
Después de la validación del BEM donde se ha demostrado la capacidad del modelo
para simular los fenómenos esenciales de transferencia de calor, éste se ha acoplado a una
parametrización urbana multicapa llamada BEP (Building Effect Parameterization, Martilli et
al., 2002). Este esquema es uno de los más completos y detallados para simulaciones a
mesoescala del clima urbano y permite una interacción directa con la capa límite atmosférica.
En BEP se reconoce la naturaleza tridimensional de las superficies de los edificios y que éstas
actúan como fuentes/sumideros de calor y momento. En BEP se tiene en cuenta el impacto de
las superficies horizontales y verticales sobre la velocidad del viento, la temperatura y la
energía cinética turbulenta. Además, para el cálculo de la radiación que alcanza las calles y
muros de los edificios se consideran múltiples reflexiones y los efectos de las sombras.
En segundo lugar, simulaciones en una dimensión vertical (1D off-line) han sido
llevadas a cabo calculando la radiación neta y diferentes flujos turbulentos con el nuevo
esquema BEP+BEM. Estos flujos han sido comparados con valores observados recogidos
durante la campaña meteorológica BUBBLE (Rotach et al., 2005) que tuvo lugar en la ciudad
suiza de Basel. Los resultados muestran que el nuevo esquema urbano BEP+BEM es capaz de
reproducir de forma satisfactoria los flujos medidos y que los resultados han mejorado cuando
han sido comparados con los obtenidos con el viejo esquema urbano BEP. Finalmente se ha
participado en un proyecto internacional donde los flujos derivados de diferentes
parametrizaciones urbanas han sido comparados con valores observados. Los resultados
3
obtenidos con los modelos urbanos BEP y BEP+BEM han sido totalmente satisfactorios.
En los modelos numéricos, las escalas espaciales y temporales usadas fijan el límite de
los posibles fenómenos físicos que pueden ser capturados por un determinado modelo. En el
marco teórico de las distintas parametrizaciones físicas de los modelos atmosféricos se hacen
simplificaciones que podrían perder su validez si el tamaño de las celdas o rejillas numéricas
usadas en la simulación no es el adecuado. Para el caso de la determinación de los
intercambios de calor que tienen lugar en terrenos heterogéneos como las ciudades, además
nos podemos encontrar con la presencia de diferentes materiales con diferentes propiedades
térmicas dentro de una misma celda numérica (~1 km 2 ). La cuestión que aparece ahora es,
cuáles son los valores de las propiedades térmicas que determinan los flujos reales a esta
escala de 1 km 2 si la celda numérica está compuesta por parches de diferentes materiales con
diferentes propiedades físicas. En esta memoria abordamos esta pregunta y proponemos una
respuesta que ha mejorado lo hecho hasta la fecha.
En tercer lugar, el nuevo esquema urbano (BEP+BEM) ha sido incorporado de forma
oficial en el modelo atmosférico WRF v3.2 (Weather Research and Forecasting) desarrollado
en el NCAR (National Center for Atmospheric Research) en Boulder, CO, EEUU. Con el
nuevo esquema urbano se han simulado dos ciudades: Houston en el estado norteamericano
de TEXAS, y la ciudad de Madrid en España.
Cuando se simula una ciudad es importante disponer de datos morfológicos de la
misma con una resolución similar a la resolución usada en la rejilla numérica más fina. Los
esquemas urbanos necesitan de diferentes parámetros para la estimación de los flujos de calor
intercambiados con la atmósfera. Algunos ejemplos de parámetros urbanos son: la fracción
urbana, la altura media de los edificios, la anchura media de las calles, el porcentaje de área
cubierta por los edificios, etc. Esta información no está disponible para la mayoría de las
4
ciudades y una primera aproximación cuando se pretende simular una ciudad es la de definir
varias clases urbanas (a cada punto de la rejilla numérica donde existe suelo urbano se le
asocia una clase) y asignar a cada clase los valores más realistas de los correspondientes
parámetros urbanos. Así es como trabaja la versión estándar del modelo atmosférico WRF
cuando se usa una parametrización urbana y así se hizo para el caso de Madrid. A partir de la
base de datos de usos de suelo CORINE (http://www.eea.europa.eu) se definieron tres clases
urbanas. Los datos meteorológicos necesarios para la validación fueron recogidos durante la
campaña DESIREX que tuvo lugar en el verano del 2008 (Sobrino et al., 2009). Buenas
estimaciones de la temperatura del aire así como de la isla de calor urbana han sido obtenidas
en esta tesis con las simulaciones del nuevo esquema urbano BEP+BEM. La isla de calor
urbana sobre Madrid llega a alcanzar una intensidad entre 5 y 6 ºC durante la noche y el calor
antropogénico fue el responsable de un aumento de la temperatura de hasta 1.5-2 ºC en
algunos lugares de la ciudad. Diferentes estrategias para mitigar la isla de calor y reducir el
consumo energético debido al uso de los aires acondicionados fueron también evaluadas.
Reducciones de hasta el 10% en el consumo energético se obtuvieron modificando el albedo,
algunas propiedades de los materiales y eliminando el calor antropogénico proveniente de los
sistemas de aire acondicionado.
Para el caso de Houston existía información morfológica detallada de la ciudad con
una resolución de 1 km 2 . De hecho existe una base de datos NUDAPT (Ching et al., 2009)
con información morfológica detallada de unas 50 ciudades de los EEUU. Para poder utilizar
esta información punto a punto en el dominio numérico se tuvo que preparar una versión
diferente a la estándar de WRF (WRF_Nudapt) la cual utilizaba la información de la
morfología urbana como nuevas variables en los ficheros de entrada del modelo. Los
esquemas urbanos BEP y BEP+BEM inevitablemente tuvieron que ser modificados ya que en
5
un principio estaban diseñados para trabajar con clases urbanas. Finalmente se simuló la
ciudad usando las dos versiones, la versión estándar de WRF definiendo tres clases urbanas
derivadas de la base de datos NLCD (National Land Cover Data para los EEUU) y la versión
modificada WRF_Nudapt. Esta última versión está siendo utilizada por personal del NCAR
(Dr. Mukul Tewari) y de la EPA (Environmental Protection Agency, Dr. Jason Ching). El
consumo energético sobre la ciudad de Houston fue calculado con el nuevo esquema urbano y
comparado con el consumo energético obtenido en otros estudios con diferentes metodologías
(bottom-up y top-down). Cuando se utilizó información detallada de la morfología de la
ciudad (WRF_Nudapt) se obtuvieron buenas estimaciones.
El trabajo realizado en esta memoria contribuye a la mejora en la capacidad de estudio
del impacto de las ciudades en la atmósfera. Gracias a la mayor complejidad de la nueva
parametrización (BEP+BEM) el impacto de las zonas urbanas está mejor representado y
estrategias de mitigación tanto de la isla de calor como del consumo energético pueden
evaluarse. Los modelos atmosféricos junto con parametrizaciones urbanas detalladas son
importantes herramientas numéricas que nos ayudan a planificar el desarrollo y a evaluar el
impacto de futuros escenarios de nuestras ciudades.
6
ORGANIZACIÓN DE LA MEMORIA
En el primer capítulo se presenta un estado del arte de las características principales de
la capa límite atmosférica sobre zonas rurales y urbanas. Las ecuaciones utilizadas en la
modelización atmosférica a mesosescala junto con algunas propiedades generales de las
parametrizaciones urbanas estarán incluidas también en este capítulo, además de una detallada
descripción del trabajo llevado a cabo en esta memoria. En el segundo capítulo se presentan
los objetivos específicos de este trabajo. El capítulo tercero contiene los resultados en forma
de artículos científicos ordenados en función de los objetivos marcados en el capítulo anterior.
En el capítulo cuarto de presenta una discusión integradora de los resultados obtenidos y
finalmente en el capítulo quinto se presentan las conclusiones y futuras líneas de
investigación. El apéndice A contiene varias publicaciones en las cuales el autor de esta
memoria ha colaborado a lo largo de la elaboración de la tesis doctoral.
7
8
SUMMARY
Always motivated for achieving a better quality of life, the human being along the
history has been characterized by his continuous migratory displacements. Thanks to the
economic, social and industrial development tied to the modern life, we attend at an
unstoppable growth of the population in the urban zones. While the inhabitants in the rural
zones are decreasing, the cities are increasing their number of inhabitants in progressive form.
Nowadays half of the world population lives in cities and it is estimated that in the next
decades this percentage will increase up to three quarters. This social phenomenon has
derived in the appearance of big urban cores with populations who exceed ten million
inhabitants (megacities). The number of megacities in the world grows rapidly and
consequently the problems associated with the life in the big cities become more pronounced.
The substitution of the rural soil for building materials, the emission of pollutants
owed to the traffic and to the industrial activities, the anthropogenic heat generated by the
different human activities, the great energetic consumption associated with the needs of
indoor thermal control in the buildings, etc., are factors that modify the climate, not only at
local scale but also at global scale (Oleson et al., 2010). It is not surprising that the interest for
the problems derived from the modern life in the cities is increasing and increasingly awakens
the curiosity in the scientific community. Two fundamental problems that in the last years
have attracted notably the interest of the experts of the atmosphere are the urban heat island
and air pollution.
The system of differential equations that describes the atmospheric flows are highly
not linear and need to be studied with numerical techniques. Thanks to the increasing capacity
of calculation of the computers and the development of the numerical models, today we can
study and understand better the impact of the cities on the atmosphere.
9
This thesis is focused fundamentally on the study of the heat island that takes place in
the cities in order to achieve a major comprehension of this phenomenon, as well as the
evaluation of different strategies for its mitigation. We will not forget the study of energetic
consumption associated with the use of the air conditioning systems as well as their impact on
the air temperature. The evaluation of strategies of mitigation both of the urban heat island
and of the energetic consumption will be the principal axes of this work.
The idea is to contribute to the development of a numerical tool capable to compute
the effects of the cities on the atmosphere, with the idea to design and to evaluate strategies of
mitigation of the heat island and of the energetic consumption associated with the growth and
development of the cities.
In the first part of the work a building energy model (BEM) has been developed to
study the heat fluxes (sensible/latent) exchanged between the buildings and the atmosphere.
Recent studies indicated (Kondo & Kikegawa, 2003; Ohashi et al., 2007) that the heat fluxes
coming from the buildings (concretely those originated by the air conditioning systems) might
have an important effect on the air temperature and should be considered in the studies of the
urban climate. In this new building energy model a building is treated as a pile of boxes each
box representing a particular floor, and for every floor the following is solved:
•
the diffusion of heat through the walls, floors and roofs,
•
the natural ventilation, as well as the radiative reflection and emission that take
place between the indoor surfaces of the building,
•
the heat generated by the domestic equipments and persons,
•
the heat flux exchanged by the air conditioning systems and the exterior.
Buildings with different number of levels (floors) can be considered in the model and the
temporal evolution of the indoor air temperature and humidity are calculated separately for
10
every floor.
After the validation of BEM where there has been demonstrated the capacity of the
model to simulate the essential phenomena of heat transfer, BEM has been coupled to a
multilayer urban canopy parameterization called BEP (Building Effect Parameterization,
Martilli et al., 2002). This scheme is one of the most complete and detailed for mesoscale
simulations of the urban climate and allows a direct interaction with the planetary boundary
layer. In BEP are recognized the three-dimensional nature of the surfaces of the buildings and
that these act as source/sink of heat and momentum. BEP accounts for the impact of the
horizontal and vertical surfaces on the wind speed, air temperature and turbulent kinetic
energy. In addition, for the calculation of the radiation that reaches the streets and walls of the
buildings shadowing effects and multiples reflections are considered.
Secondly, simulations in a vertical column (1D off-line) have been carried out
calculating the net radiation and different turbulent flows with the new urban scheme
BEP+BEM. These fluxes have been compared with observed values gathered during the
meteorological campaign BUBBLE (Rotach et al., 2005) that took place in the Switzerland
city of Basel. The results show that the new urban scheme BEP+BEM is able to reproduce
satisfactorily the measured fluxes and that the results have improved when they are compared
with the ones obtained with the old urban scheme BEP. Finally one has taken part in an
international project where the flows derived from different urban parameterizations have
been compared with observed values. The results obtained with the urban schemes BEP and
BEP+BEM have been totally satisfactory.
In the numerical models, the spatial and temporal scales used fix the limit of the
possible physical phenomena that can be captured by a certain model. In the theoretical
framework of the different physical parameterizations used in the atmospheric models,
11
simplifications are supposed that might lose their validity if the size of the numerical cells
used in the simulations is not suitable. For the case of the determination of the heat exchanges
that take place in heterogeneous areas as cities, in addition we can encounter the presence of
different materials with different thermal properties inside the same numerical cell (~1 km 2 ).
The question that appears now is which are the values of the thermal properties that determine
the real fluxes at this scale of 1 km 2 if the numerical cell is composed by patches of different
materials with different physical properties. In this memory we approach this question and
propose a response that improves what is done up to date.
Thirdly, the new urban scheme (BEP+BEM) has been incorporated officially in the
atmospheric model WRFv3.2 (Weather Research and Forecasting) developed at NCAR
(National Centre for Atmospheric Research) in Boulder, CO, EEUU. With the new urban
scheme two cities have been simulated: Houston in Texas (EEUU), and the city of Madrid in
Spain.
When a city is simulated it is important to have morphological information with a
resolution similar to the resolution used in the finest numerical domain. The urban schemes
need different parameters for the estimation of the heat fluxes exchanged with the
atmosphere. Some examples of urban required parameters are: the urban fraction, the average
height of the buildings, the average width of the roads, the percentage of area covered by the
buildings, etc. This information is not available for most of the cities and the first
approximation done is to define several urban classes (at every point of the numerical domain
where exists urban soil an urban class is associated) and to assign to every class the most
realistic values of the corresponding urban parameters. In this way works the standard version
of the atmospheric WRF model when an urban parameterization is used and it was done for
the case of Madrid. From the land use/cover database CORINE (http://www.eea.europa.eu)
12
three urban classes were defined. The meteorological information necessary for the validation
was gathered during the DESIREX campaign that took place in the summer of 2008 (Sobrino
et al., 2009). Good estimations of the air temperature as well as of the urban heat island were
obtained in the simulations with the new urban scheme BEP+BEM. Different strategies to
mitigate the heat island and to reduce the energetic consumption due to the use of the air
conditioning systems were also evaluated. Reductions of up to 10 % in the energetic
consumption were obtained modifying the albedo, some properties of the materials and
eliminating the anthropogenic heat coming from the air conditioning systems. The urban heat
island over Madrid reached between 5 and 6 ºC during the night and the anthropogenic heat
was responsible of an increase in the air temperature up to 1.5-2 ºC in some places of the city.
For the case of Houston morphological detailed information existed with a resolution
of 1 km 2 . In fact there exists a database called (NUDAPT, Ching et al., 2009) with
morphological detailed information of approximately 50 cities of the USA. To be able to use
this information point to point in the inner numerical domain, it is necessary to modify the
WRF´s standard version to use the information of the urban morphology as new variables in
the input files of the model. The urban schemes BEP and BEP+BEM inevitably had to be
modified as at the beginning they were designed to work with urban classes. Finally the city
was simulated using both versions, WRF´s standard version defining three urban classes
derived from the database NLCD (National Land Cover Data) for the US and the modified
version WRF_Nudapt. The latter version is being used by personal of the NCAR (Dr. Mukul
Tewari) and of the EPA (Environmental Protection Agency, Dr. Jason Ching). The energetic
consumption over the city of Houston was calculated with the new urban scheme and
compared with the energetic consumption obtained in other studies by different
methodologies (bottom-up and top-down). When there the detailed information of the
13
morphology was used (WRF_Nudapt) good estimations were obtained.
The work done in this memory contributes to the improvement in the capacity of
studying the impact of the cities in the atmosphere. Thanks to the major complexity of the
new urban scheme (BEP+BEM) the impact of the urban zones is better represented and
strategies of mitigation both of the urban heat island and of the energetic consumption can be
evaluated. The atmospheric models together with urban detailed parameterizations are
important numerical tools that help us to plan the development and to evaluate the impact of
future scenarios of our cities.
14
ORGANIZATION OF THE THESIS
In the first chapter a state of the art of the principal characteristics of the planetary
boundary layer over rural and urban zones is presented. The equations used in the mesoscale
atmospheric modelling together with some general properties of the urban parameterizations
are included also in this chapter, together with a detailed description of the work carried out in
this thesis. In the second chapter the specific aims of this work are presented. The third
chapter contains the results in form of scientific papers arranged depending on the aims
marked in the previous chapter. In the fourth chapter a discussion of the results obtained is
presented and finally in the fifth chapter the conclusions and future lines of investigation are
presented. The appendix A contains several publications co-authored by the author of this
thesis.
15
16
Capítulo 1
CAPÍTULO 1
ESTADO DEL ARTE Y DESCRIPCIÓN DEL TRABAJO
17
Capítulo 1
1.1 Introducción
Hoy en día la mitad de la población mundial vive en las ciudades y se espera que en
las próximas décadas esta proporción aumente hasta las tres cuartas partes. Consecuentemente
el bienestar de la mayoría de la población mundial está ligado al entorno urbano. El calor
antropogénico debido al tráfico y a las actividades industriales, las propiedades físicas y
geométricas particulares de las ciudades que acentúan una mayor absorción de la radiación de
onda corta y una menor emisión de la radiación de onda larga por encima de los edificios, el
mayor almacenamiento de la energía solar debido a las propiedades térmicas de los materiales
utilizados en la construcción junto con la polución del aire hacen que la temperatura de la
atmósfera sobre la ciudad pueda diferir sustancialmente cuando se la compara con la
temperatura de su entorno rural más próximo (esta diferencia es conocida como isla de calor
urbana). Debido a la alta rugosidad de las superficies urbanas los efectos mecánicos también
juegan un papel importante afectando notablemente a la velocidad del viento. Las
dimensiones de los edificios y su irregular distribución espacial hacen que el viento en las
capas inferiores de la atmósfera pueda diferir notablemente del viento vecino próximo a las
ciudades. Además, los gradientes térmicos existentes entre la ciudad y sus alrededores pueden
originar flujos medios desde las zonas rurales hacia la ciudad desviando la dirección de los
vientos regionales. Estos últimos flujos tienen su origen en la diferente velocidad de
enfriamiento del suelo urbano y del suelo rural que ocurre después de la puesta del sol
(Haeger-Eugensson et al., 1999).
Claramente la estructura de la ciudad modifica el balance energético de la superficie y
la composición de la atmósfera (Oke, 1988; Landsberg, 1981). Las interacciones físicoquímicas que tienen lugar entre la ciudad y la atmósfera modifican el clima urbano y pueden
acentuar los efectos negativos que en determinados episodios (olas de calor, episodios con
18
Capítulo 1
alta concentración de contaminantes, etc.) pudieran hacer que la vida en la ciudad fuera más
desagradable y peligrosa para la salud de sus ciudadanos.
Las interacciones entre la ciudad y la atmósfera involucran fenómenos de diferente
naturaleza y diferentes escalas espacio temporales (locales y mesoescalares). Todo ello junto
con la no linealidad de las ecuaciones atmosféricas hace que los modelos numéricos sean la
mejor herramienta que permita su estudio de una forma más detallada. Afortunadamente, a
pesar de todas estas complejidades el interés por el estudio del clima urbano se ha visto
incrementado notablemente en la última década.
1.2 La capa límite atmosférica
La atmósfera es una fina capa (comparándola con el radio de la Tierra) que cubre la
totalidad de ésta y gracias a su composición la vida es posible en ella. Cerca del 100 % de su
masa está compuesta de cuatro especies químicas: nitrógeno (78 %), oxígeno (21 %), argón
(0.93 %) y dióxido de carbono (0.03 %). En la parte baja de la atmósfera (primeros centenares
de metros) encontramos una gran variabilidad en la concentración de agua en forma de gas o
bien condensada o sublimada en forma de nubes.
Considerando la dirección vertical, la atmósfera puede dividirse en cuatro capas
principales generalmente asociadas con su distribución vertical de la temperatura. La capa
más externa es la termosfera. Ésta se extiende desde unos 80 km desde la superficie terrestre
hasta unos 500-600 km. La temperatura del aire en esta capa sufre un fuerte gradiente que va
desde los 1000-2000 K en su borde más externo hasta unos 190 K en su límite inferior. Aquí
es donde se registra la menor temperatura de toda la capa atmosférica. Posteriormente
tendríamos la mesosfera. La mesosfera es la capa comprendida entre los 50 y los 80 km sobre
la superficie terrestre. En esta capa la temperatura experimenta un gradiente más o menos
constante que va de los 190 K en su parte más externa hasta los 273 K en su límite inferior.
19
Capítulo 1
Aquí la alta temperatura alcanzada es debida a la absorción de los rayos UV y a la formación
y descomposición del ozono estratosférico. La tercera capa sería la estratosfera. Esta capa
tendría un espesor de unos 30 km. En el límite inferior la temperatura es de unos 213-223 K.
Al límite inferior de la estratosfera se la conoce con el nombre de tropopausa, debajo de la
cual aparece una zona caracterizada por fuertes vientos (Jet Stream) bien conocidos por los
pilotos de aviación. La última capa es la troposfera. Su grosor varía desde los 8 km en las
latitudes más altas hasta los cerca de 20 km en las zonas ecuatoriales. Está caracterizada por
fuertes gradientes positivos de temperatura (~ 6 K por km) hacia la superficie terrestre y por
una alta variabilidad en la concentración de vapor de agua y agua condensada y cristales de
hielo en forma de nubes.
La capa límite atmosférica es aquella zona (perteneciente a la troposfera) influenciada
directamente por la rugosidad y el balance energético que tiene lugar en la superficie. La capa
límite atmosférica puede extenderse desde unos pocos metros (~ 100 m o menos bajo
condiciones estables) hasta un par de kilómetros bajo condiciones convectivas. En esta capa la
velocidad del viento, la temperatura y la humedad del aire presentan grandes fluctuaciones y
existe una importante mezcla vertical (Stull, 1988).
Debido al calentamiento de la superficie terrestre originado por el sol, el aire en
contacto con ésta experimenta movimientos verticales turbulentos produciéndose una rápida
mezcla desde las primeras horas de la mañana. Esto hace que la capa límite atmosférica vaya
creciendo y alcance su máxima altura unas horas después del medio día. La temperatura
potencial y humedad del aire son prácticamente constantes cuando la capa límite está
totalmente desarrollada (en las primeras horas de la tarde) y es una capa bien mezclada debido
a los movimientos turbulentos verticales que tienen lugar en ella y que pueden alcanzar
incluso dimensiones cercanas al tamaño de la propia capa límite. La capa límite está acotada
20
Capítulo 1
por una capa con fuerte inversión térmica donde la temperatura crece con la altura y por
encima de la cual se encuentra la atmósfera libre. Cerca del suelo coexiste una fina capa
superficial donde la temperatura y humedad del aire decrecen y la velocidad del viento crece
rápidamente con la altura. Desde la puesta de sol hasta el amanecer la capa límite atmosférica
evoluciona dando lugar a una capa estable sobre la superficie de unos pocos metros de altura
(originada por el enfriamiento radiativo del suelo) y una capa residual por encima de ésta
donde la temperatura potencial y humedad del aire son prácticamente constantes. Esta capa
residual tiene su origen en la capa mezclada originada durante las horas diurnas del día
previo. Debido al menor enfriamiento radiativo que tiene lugar en las ciudades (esto se
explicará con más detalle más adelante en la sección 1.4.1 de esta memoria) y a la liberación
del calor almacenado durante el día en los edificios y superficies pavimentadas, la capa
estable y la capa residual pueden llegar a fusionarse sobre la ciudad, dando lugar a una capa
neutra mezclada con un espesor que puede alcanzar varios centenares de metros. Por encima
de la capa límite atmosférica el viento dominante es el viento casi geostrófico y la turbulencia
es escasa.
Esta evolución idealizada (ver la Fig. 1.1) de la capa límite atmosférica donde hemos
supuesto que los efectos térmicos dominan sobre los mecánicos puede sufrir notables
modificaciones debido a las diferentes condiciones sinópticas y a la topografía del suelo
(montañas, valles, costa marina, etc.).
21
Capítulo 1
Figura 1.1. Sección vertical mostrando la evolución idealizada de la capa límite atmosférica.
Los fenómenos atmosféricos dentro de la capa límite atmosférica están gobernados por
diferentes escalas espacio-temporales. Por ejemplo, los fenómenos microescalares (escalas
espaciales de unos pocos de metros y temporales de unos segundos) determinan como son
transmitidos los contaminantes desde sus fuentes (los vehículos por ejemplo) hacia la capa
límite atmosférica. El transporte de estos contaminantes por la capa límite atmosférica vendría
gobernado por los fenómenos mesoescalares (escalas espaciales de varias decenas de
kilómetros y temporales de varias horas) como pudieran ser una brisa marina en una zona
costera o los vientos catabáticos producidos en un valle. Cuando se quieren determinar las
diferentes variables meteorológicas en la capa límite atmosférica, ambas escalas deben ser
consideradas y el uso de los modelos atmosféricos mesoescalares junto con diferentes
parametrizaciones
son
las
herramientas
más
adecuadas.
Por
ejemplo,
con
las
parametrizaciones urbanas se intentan representar los fenómenos microescalares que tienen
lugar en las ciudades y acoplarlos a los fenómenos mesosescalares que gobiernan las
condiciones meteorológicas del clima regional.
22
Capítulo 1
1.2.1 La capa límite urbana
La capa límite urbana es la capa límite atmosférica que existe por encima de los
edificios de nuestras ciudades. La altura de la capa límite urbana puede diferir de la altura de
la capa límite atmosférica en zonas rurales debido al mayor calentamiento que se produce en
las superficies pavimentadas y a la mayor rugosidad propia de los edificios. Una de las
características principales que diferencia a la capa límite urbana de la capa límite atmosférica
sobre zonas naturales es el exceso de temperatura observada en un área metropolitana cuando
se la compara con sus áreas rurales vecinas (isla de calor urbana).
Comentamos (en la sección anterior) que durante la noche una capa estable se forma
junto al suelo rural debido al enfriamiento radiativo de éste y que puede alcanzar varios
centenares de metros. El aire en contacto con el suelo se enfría mas rápidamente que el que
está por encima originando un gradiente térmico inverso que inhibe los movimientos
verticales. Sin embargo sobre la ciudad ocurre algo bastante diferente. El enfriamiento
radiativo es mucho menor debido al atrapamiento de la radiación que tiene lugar entre las
calles (canyon urbano) y los edificios (Oke, 1981). Además el flujo de calor que durante el
día se ha ido almacenando en las superficies pavimentadas ahora es liberado en forma de calor
sensible y el aire en contacto con las superficies urbanas puede seguir calentándose. A pesar
de que lógicamente la temperatura del aire sigue bajando en la capa límite urbana durante la
noche, se puede llegar a formar una capa neutra e incluso ligeramente inestable de unos
cientos de metros por encima de la ciudad (Bornstein, 1968; Godowitch et al., 1985; Uno et
al., 1988; Oke, 1995). La intensidad de la isla de calor urbana y la estabilidad de la capa límite
sobre la misma dependerá de las características morfológicas de la ciudad y de las
condiciones meteorológicas que han estado presentes a lo largo del día.
Durante las horas diurnas debido al calentamiento del suelo rural, la capa estable va
23
Capítulo 1
desapareciendo dando lugar a una capa inestable bien mezclada donde los flujos turbulentos
predominan en toda la capa límite atmosférica. Las diferencias entre la capa límite sobre la
ciudad y las zonas rurales son pequeñas durante las horas diurnas y dependen mucho de las
condiciones meteorológicas presentes en la región. Un componente importante que afecta a la
temperatura sobre la ciudad y que no existe en las zonas rurales es el calor antropogénico. Sin
embargo éste durante el día se distribuye a lo largo de toda la capa límite (que puede alcanzar
varios kilómetros) y aunque pueda tener valores significativamente importantes su efecto se
ve reducido. Por consiguiente, la diferencia de temperatura entre las dos capas límites (la
urbana y la rural) puede ser nula o pequeña durante el día, aunque esta diferencia depende
mucho del tipo de suelo y de su contenido en agua. En varias ciudades se han medido durante
el día temperaturas del aire menores sobre la ciudad cuando se han comparado con las
temperaturas del aire de las áreas rurales cercanas. Si la zona urbana está caracterizada por
edificios altos y calles estrechas la radiación solar tiene más dificultades para calentar
determinadas superficies verticales y las calles pavimentadas, permitiendo que el aire dentro
de la canopy urbana pueda estar más frío que el aire en las zonas rurales colindantes
(Georgakis & Santamouris, 2009).
Lógicamente, la isla de calor puede desaparecer e incluso no formarse si las
condiciones meteorológicas sobre la región son desfavorables. Fuertes vientos y cielos
cubiertos dificultan la formación de la isla de calor debido a la mezcla de los aires urbano y
rural y a la reducción de la radiación solar que alcanza finalmente el suelo.
1.3 Flujos turbulentos y balance energético superficial
El balance energético en la superficie terrestre es el que determinará la energía
disponible para la generación de los diferentes flujos turbulentos que existen en la parte baja
de la atmósfera. La cantidad de energía disponible está gobernada por el ciclo solar diario que
24
Capítulo 1
actúa como fuente externa y depende de la naturaleza propia del suelo (urbano o rural) así
como de las diferentes propiedades térmicas de éste. Asumiendo la no existencia de advección
horizontal, el balance energético en un volumen de aire próximo al suelo y que contiene a éste
puede ser descrito como (la explicación detallada de los símbolos pueden verse en el apéndice
de este capítulo):
Q * + Q AH = Q H + Q LE + QG ,
(1)
donde el término Q * es la radiación neta, Q AH es el flujo de calor antropogénico, QH es el
flujo de calor sensible, QLE es el flujo de calor latente y QG es el flujo de calor que fluye hacia
o desde el suelo (Oke, 1988). Este balance energético nos dice que las fuentes de energía
radiativa y el calor generado por las actividades humanas se transforman en calor sensible,
latente y en un flujo de calor que penetra en la superficie o fluye desde ella. Sobre una
superficie natural ( Q AH = 0 ) el balance energético puede escribirse como:
Q * = (1 − alb) S ↓ +ε ( L ↓ −σTs4 )
______
Q H = ρ C P w′θ ′
(2)
______
QLE = ρ LV w′q ′
donde el término QG = Q * − QH − QLE sería el término residual e igual al flujo que penetra en
o fluye desde la superficie (QG = − k
∂T
∂Z
).
Z =0
En las zonas urbanas además del calor antropogénico, el cómputo de la radiación neta
es bastante diferente cuando se la compara con las zonas rurales. Al no disponer de una única
superficie (superficies verticales y horizontales) el reparto energético es diferente y la
geometría de los edificios condiciona fuertemente la radiación que alcanza los muros y las
calles. Todo ello junto con las diferentes propiedades térmicas hacen que el reparto energético
difiera entre las zonas rurales y las zonas urbanas. Además en las zonas rurales el calor latente
25
Capítulo 1
es en general mayor que en las zonas urbanas disminuyendo así la cantidad de energía
disponible en forma de calor sensible. Por consiguiente, la atmósfera sobre las ciudades puede
presentar diferencias de temperatura y humedad cuando se la compara con la atmósfera sobre
las zonas rurales vecinas. Las características propias de las zonas urbanizadas favorecen la
formación de la isla de calor urbana.
Los flujos turbulentos suelen calcularse de forma estándar usando la conocida teoría K
( K − theory ). Esta teoría local también conocida como la teoría del gradiente-transporte
establece que los flujos turbulentos asociados a una variable ξ pueden escribirse como:
______
u i′ξ ′ = − K
∂ξ
,
∂xi
(3)
______
donde K es el coeficiente de intercambio turbulento, u i′ξ ′ es el flujo turbulento de la
variable ξ en la dirección i y ξ es el valor medio (ensemble average) de la variable en
cuestión (temperatura, humedad o una componente del viento). Para valores positivos de K
los flujos tienen lugar en sentido contrario a la dirección local del gradiente de los valores
medios de la variable; cuando K es negativo se habla entonces de flujo contra-gradiente, y la
ec. (3) se suele ver modificada para tener en cuenta este efecto.
Aplicando la ec. (3) a las anteriores ecs. (2), podemos escribir los flujos turbulentos de
calor sensible y latente como (para más detalles de los símbolos consultar el apéndice del
capítulo):
∂θ
∂z
∂q
= − ρ LV K q
∂z
QH = − ρ C P K θ
QLE
(4)
donde K θ y K q son los coeficientes verticales de intercambio turbulento para la temperatura
y humedad respectivamente. Estos coeficientes turbulentos distan mucho de ser constantes,
26
Capítulo 1
estando bastante influenciados por la estabilidad, y deben ser parametrizados a través de algún
esquema turbulento (presentaremos un esquema turbulento en la sección 1.4 de esta memoria)
para poder ser calculados.
1.4 Modelización a mesoescala. Ecuaciones que gobiernan los flujos
turbulentos.
En las últimas décadas, el rápido aumento en la capacidad de memoria y de cálculo de
los ordenadores ha contribuido notablemente en la mejora y desarrollo de diferentes tipos de
modelos atmosféricos mesoescalares que simulan las variables atmosféricas en prácticamente
la totalidad de la troposfera. Para los estudios de calidad del aire, los resultados obtenidos con
los modelos atmosféricos son introducidos en los modelos fotoquímicos que resuelven las
ecuaciones del transporte y las diferentes transformaciones químicas (Jacobson, 1999). En
algunos casos ambos modelos corren separadamente pero hoy en día es fácil encontrarlos
acoplados (el modelo WRF-Chem es un ejemplo de ello, Grell et al., 2005) con la ventaja de
que los campos meteorológicos pueden ser utilizados instantáneamente por el modelo
fotoquímico en cada paso de tiempo.
Aunque dependiendo del modelo utilizado las ecuaciones a resolver puedan diferir
ligeramente debido a algunas simplificaciones, expondremos en esta sección las ecuaciones
fundamentales básicas de los flujos turbulentos. Separando las diferentes variables
meteorológicas en su valor medio (ensemble average) y en su parte turbulenta, podemos
obtener las ecuaciones generales que rigen los movimientos atmosféricos turbulentos (para
una derivación completa de las ecuaciones puede consultarse por ejemplo el libro de Stull,
1988). La ecuación que describe la conservación de la masa en un volumen dado de aire es
∂U i
= 0,
∂xi
(5)
27
Capítulo 1
donde hemos supuesto la condición de incompresibilidad
∂U i
d ln ρ
<<
, válida en los
dt
∂xi
movimientos turbulentos propios de estas escalas (todos los símbolos están explicados en el
apéndice del capítulo).
La ecuación de conservación del momento para la componente i de la velocidad del
viento en un sistema de coordenadas situado sobre la superficie terrestre puede escribirse
como:
_______
∂U i
∂U i
∂ 2U i ∂ u i′u ′j
1 ∂P
+U j
= −δ i 3 g + fε ij 3U j −
+ν
−
+ Dui ,
∂t
∂x j
ρ ∂xi
∂x j
∂x 2j
(I)
(II)
(III) (IV)
(V)
(VI)
(VII)
(6)
(VIII)
donde el primer término (I) de la izquierda representa la variación local de la velocidad media
y el segundo término (II) la advección debida al viento medio. El primer término del lado
derecho (III) es no nulo sólo en la dirección vertical y representa la aceleración debida a la
gravedad. El siguiente término (IV) es el término de Coriolis y describe la influencia de la
rotación terrestre. El siguiente término (V) es la fuerza debida al gradiente de presión presente
en el fluido y el término (VI) representaría la influencia de su viscosidad. El término (VII) es
la divergencia del transporte de flujo turbulento de momento (Reynolds stress), que aparece
como consecuencia de la descomposición de las variables instantáneas en media y turbulenta,
y finalmente el último término (VIII) representa las fuerzas inducidas por la interacción entre
las distintas superficies (suelo rural o edificios) y el flujo atmosférico.
Aplicando la primera ley de la termodinámica a un volumen de aire dado, la
conservación del calor puede escribirse matemáticamente como:
28
Capítulo 1
_______
*
1 ∂Q j ∂ u ′jϑ ′
∂ϑ
∂ϑ
∂ϑ
+U j
=νϑ
−
−
+ Dϑ ,
∂t
∂x j
ρ C p ∂x j
∂x j
∂x 2j
2
(I)
(II)
(III)
(IV)
(V)
(7)
(VI)
donde no se ha especificado el término que representa el calor proveniente de los posibles
cambios de fase gas-líquido, líquido-sólido o gas-sólido que pudieran tener lugar en la
atmósfera. El primer término (I) del lado izquierdo es la variación local de la temperatura
potencial media y el segundo término (II) describe la advección de temperatura debida al
viento medio. El primer término (III) del lado derecho corresponde a la difusión térmica
molecular y el segundo término (IV) representa las pérdidas de calor asociadas a la
emisividad radiativa de la atmósfera. El término (V) es la divergencia del transporte de flujo
turbulento de calor y finalmente el término (VI) incluiría todas las fuentes de calor sensible
provenientes de la interacción de la atmósfera con las distintas superficies (rurales o edificios
en zonas urbanas).
La conservación de la cantidad de agua presente en la atmósfera, nos permite escribir
para la humedad específica la siguiente ecuación:
_______
∂q
∂q
∂ 2 q ∂ u ′j q ′
+U j
=ν q 2 −
+ Dq
∂t
∂x j
∂x j
∂x j
(I)
(II)
(III)
(IV)
(8)
(V)
donde el primer término (I) representa la variación local de la humedad específica media del
aire y el segundo término (II) de la izquierda la advección de la humedad debida al viento
medio. El primer término del lado derecho (III) representa la difusión molecular de la
humedad específica media del vapor de agua y el término (IV) es la divergencia del transporte
del flujo turbulento de humedad específica. Finalmente, el término (V) incluiría todas las
fuentes de humedad específica del aire. Lógicamente todos los términos que involucran
29
Capítulo 1
variables turbulentas en las anteriores ecs. (6-VIII), (7-V) y (8-IV) deben ser parametrizados
por algún esquema turbulento para poder ser calculados.
Finalmente para que el número de incógnitas ( ρ , P , ϑ , U i , q ) sea igual al número de
ecuaciones se hace uso de la conocida ecuación de estado P = ρ RT de los gases ideales.
Muchos son los esquemas turbulentos propuestos y utilizados en los diferentes
modelos atmosféricos. Debido a que los flujos turbulentos no pueden ser calculados
directamente éstos deben ser parametrizados por medio de algún esquema que describa de
forma aproximada el comportamiento turbulento de la atmósfera. El esquema turbulento más
utilizado en esta memoria ha sido un esquema turbulento k − l 1 desarrollado por Bougeault &
Lacarrère, (1989). En este esquema el flujo de transporte turbulento vertical se calcula de
acuerdo a la ec. (3) anterior y el coeficiente turbulento K z se obtiene como:
K z = C k l k TKE ,
(9)
donde C k es una constante de valor 0.4, TKE es la energía cinética turbulenta por unidad de
masa TKE =
__________ _________
1 __________
( (u ′) 2 + (v ′) 2 + ( w′) 2 ) y l k es una longitud de escala.
2
La longitud de escala l k se obtiene de acuerdo a las siguientes relaciones:
z + lup
β (ϑ ( z ) − θ ( z ′))dz ′ = TKE ( z )
z
z
β (ϑ ( z ′) − ϑ ( z ))dz ′ = TKE ( z ) ,
(10)
z −ldown
l k = mínimo(lup, ldown)
donde β = g / T es el coeficiente de flotabilidad térmico. Obsérvese que para el cómputo del
coeficiente turbulento K z se requiere de la resolución de una ecuación de pronóstico para el
cálculo de la energía cinética turbulenta TKE (más detalles de la ecuación a resolver pueden
1
Un modelo k-l significa que resuelve una ecuación para la energía cinética turbulenta (k) y que estima la
disipación por medio de una longitud de escala (l).
30
Capítulo 1
encontrarse en Bougeault & Lacarrère, 1989). En este esquema turbulento los coeficientes
verticales para el momento y el calor se consideran iguales (el número de Prandtl turbulento
es 1, en realidad en condiciones de estabilidad de moderada a fuerte esto no es del todo
realista). En la ec. (10), las longitudes de escala lup y ldown representan las distancias que
una parcela de aire originalmente situada en el nivel z y con energía cinética turbulenta
TKE ( z ) puede desplazarse hacia arriba o hacia abajo antes de detenerse debido a los efectos
de estabilidad térmica (flotabilidad). En la formulación de este esquema se asume la presencia
de una capa estable a una cierta altura (la altura de la capa límite atmosférica) sobre el suelo.
Es interesante apuntar que la longitud de escala ldown no puede ser mayor que la altura por
encima del suelo a la que se encuentra la parcela de aire ( ldown = z ).
Originalmente en los primeros modelos mesoescalares el transporte de flujo turbulento
horizontal no era considerado y sólo se tenía en cuenta el transporte de flujo turbulento en la
dirección vertical. Esto era debido principalmente a que la resolución numérica en la
dirección vertical en la capa límite era mucho mayor que la resolución en la dirección
horizontal. Sin embargo hoy en día gracias al aumento en la capacidad de cálculo de los
ordenadores la resolución horizontal se ha visto incrementada notablemente y simulaciones
con resolución horizontal por debajo de 1 km son bastante frecuentes. Por consiguiente, la
necesidad de uso de esquemas turbulentos tridimensionales es hoy una realidad. Sin embargo,
se debe prestar atención a la resolución horizontal utilizada con un modelo mesoescalar pues
estructuras celulares no reales podrían aparecer como parte de la solución numérica debido al
uso de una alta resolución en la dirección horizontal (Lemone et al., 2010). Este es un tema
abierto y probablemente la solución pueda estar condicionada al tipo de esquema turbulento
usado cuando simulamos con alta resolución ( ∆x, ∆y < 1 km).
31
Capítulo 1
1.4.1 Modelización de la capa límite atmosférica sobre zonas urbanas.
Parametrizaciones urbanas.
Las ciudades debido a la alta rugosidad de su superficie y a sus propiedades térmicas
pueden tener un impacto importante en la estructura de la capa límite atmosférica. El impacto
de la ciudad involucra tanto a los efectos mecánicos (turbulencia creada por la presencia de
los edificios) como a los efectos térmicos (atrapamiento radiativo y efectos de sombreado en
el canyon urbano). La teoría tradicional usada para representar a una superficie urbana en los
modelos de mesoescala era la Monin-Obukhov Similarity Theory (MOST). Esta teoría supone
que cerca de la superficie los flujos turbulentos son prácticamente constantes con la altura. La
diferencia en el tratamiento con respecto a las superficies rurales es el uso de un mayor valor
para la rugosidad del suelo y el uso de diferentes propiedades térmicas.
Sin embargo, diferentes medidas realizadas en diferentes ciudades (Rotach, 1993;
Roth & Oke, 1993) pusieron de manifiesto que los flujos turbulentos no son constantes con la
altura en la sub-capa de rugosidad urbana (urban roughness sub-layer), la parte de la capa
superficial entre el nivel de calle y 2-3 veces la altura media de los edificios. En la teoría
MOST no se tiene en cuenta el atrapamiento radiativo ni los efectos de sombreado que tienen
lugar en el canyon urbano, por consiguiente esta teoría no puede describir de forma
satisfactoria la isla de calor urbana ni el reparto energético. En la última década se han
desarrollado multitud de esquemas urbanos (Masson 2000; Kusaka et al., 2001; Martilli et al.,
2002) con la intención de obtener una mejor representación de los efectos de las ciudades en
la atmósfera. La parametrización urbana más utilizada en esta memoria ha sido la
parametrización BEP (Building Effect Parameterization) de Martilli et al., (2002). Esta
parametrización es una de las más detalladas y al ser multicapa (es decir, posee niveles
verticales propios definidos dentro del canyon urbano) permite una interacción directa con la
32
Capítulo 1
capa límite urbana. En esta parametrización una zona urbana está representada por una clase
urbana que a su vez está caracterizada por los siguientes parámetros urbanos.
•
La distancia media entre los edificios de una calle (W).
•
La anchura media de los edificios (B) y la altura media de los mismos (H).
•
La distribución vertical (en altura) de los edificios, expresada en términos de la
probabilidad γ ( z ) de que un edificio tenga una altura dada z . Además se utiliza una
densidad de probabilidad Γ( z ) que expresa la probabilidad de que un edificio tenga
una altura mayor o igual que z (ver la Fig. 1.2). Ambas funciones están relacionadas
por la expresión:
Γ( z iu ) =
nu
γ ( z ju ) ,
(11)
ju = iu
donde nu es el mayor nivel vertical en la rejilla urbana.
•
La orientación de las calles.
•
Las propiedades térmicas de los materiales (conductividad térmica, capacidad
calorífica, albedo y emisividad) de las distintas superficies urbanas (calles, muros y
tejados).
Figura 1.2. Ilustración de los parámetros geométricos urbanos: anchura media de las calles (W), anchura
media de los edificios (B) y distribución vertical de los edificios en términos de
γ
y
Γ.
33
Capítulo 1
Este esquema urbano calcula los flujos turbulentos y las diferentes variables meteorológicas
para una orientación particular de las calles. BEP utiliza una malla vertical propia (diferente
de la malla vertical del modelo atmosférico) con una resolución mayor definida por el
usuario. La extensión de esta fina malla alcanza el edificio más alto de cada clase urbana.
Después del cálculo de los flujos y las diferentes variables meteorológicas, éstas son
verticalmente interpoladas en la rejilla mesoescalar.
Los impactos de los edificios considerados en el esquema urbano BEP (Martilli et al.,
2002) son:
•
las fuerzas de arrastre debido a la presencia de las superficies verticales
(muros),
•
las fuerzas de fricción que generan las superficies horizontales (calles y
tejados),
•
la modificación de los flujos de calor debido al atrapamiento radiativo y a los
efectos de las sombras en el canyon urbano, y finalmente
•
la generación de energía cinética turbulenta a partir de la energía cinética
media del aire.
Los impactos de las diferentes superficies se calculan teniendo en cuenta la proporción
de cada una de ellas en los diferentes niveles verticales de la rejilla urbana. La pérdida de
momento debido a la fricción de las superficies horizontales utiliza la teoría estándar MOST
con la diferencia de que ahora la fuerza de fricción se distribuye verticalmente desde el suelo
(calles) hasta el edificio más alto de la clase urbana y es proporcional a la superficie
horizontal presente en cada nivel. Por otro lado, las superficies verticales (muros) inducen
fuerzas de arrastre, las cuales son nuevamente calculadas considerando la proporción de
superficie vertical presente en cada nivel vertical urbano. De este modo, la parametrización
34
Capítulo 1
urbana BEP tiene en cuenta los sumideros de momento debido a las distintas superficies de
los edificios en la totalidad de la canopy urbana.
Las superficies horizontales y verticales (muros, calles y tejados) son tratadas también
separadamente para el cálculo de los flujos de calor. Debido a que estos flujos dependen de la
diferencia de temperatura entre el aire y la capa exterior de la superficie en contacto con éste,
resolvemos la ecuación de difusión para cada una de las superficies urbanas y consideramos
como condición de contorno un balance energético en la capa externa. La otra condición de
contorno en la capa interna de una superficie se obtiene fijando la temperatura durante toda la
simulación. El valor fijado es libremente definido por el usuario y la solución final depende
de éste.
Para considerar los efectos de las sombras y el atrapamiento radiativo se calculan los
factores de vista (view factors) para cada nivel vertical dentro de la rejilla urbana. De esta
forma múltiples reflexiones de la radiación proveniente del sol y de la radiación de onda larga
pueden considerarse entre los muros y las calles. Conocida la radiación total que llega a una
superficie, se puede calcular el balance energético sobre la misma y los flujos de calor
intercambiados con el aire.
El impacto de las superficies urbanas en la producción de energía cinética turbulenta
utiliza dos aproximaciones diferentes para cada una de las superficies horizontales y verticales
presentes en la clase urbana. Aquí nuevamente los efectos son proporcionales a la distribución
vertical del área de las superficies presentes (ver más detalles en Martilli et al., 2002).
El término Dui de la ec. (6) representa la pérdida de momento debida a la presencia de
las superficies urbanas. De forma análoga, los términos Dϑ y Dq de las ecs. (7) y (8) serán
iguales a los flujos derivados del esquema urbano una vez que éstos estén interpolados en la
malla del modelo atmosférico. A la hora de calcular el volumen disponible para el aire dentro
35
Capítulo 1
de la celda numérica urbana se tiene en cuenta el volumen ocupado por los edificios dentro de
la misma. Además, para el transporte turbulento vertical se tiene en cuenta el área horizontal
ocupada por los edificios entre dos celdas numéricas adyacentes.
Por último es importante comentar que la vegetación en las zonas urbanas (por
ejemplo la vegetación presente en los parques y jardines) puede jugar un papel importante y
debe considerarse a la hora de calcular los diferentes flujos (Grimmond et al., 2010a). Debido
a que la interacción de los edificios con la atmósfera es muy diferente que la interacción de la
atmósfera con la vegetación, en el modelo atmosférico se define para cada clase urbana un
porcentaje de área urbana ( α ) donde se calculan los flujos de calor y momento con BEP, y el
resto ( 1 − α ) corresponde a un porcentaje de área rural, donde los flujos se calculan con un
modelo de suelo y vegetación. De esta forma, los flujos ( φ ) totales de calor y momento en el
modelo atmosférico se obtienen promediando ambas partes, es decir :
φ = αφ urbano + (1 − α )φ rural ,
(12)
donde φ urbano representa el flujo derivado del esquema urbano y φ rural el flujo derivado del
esquema rural.
1.4.2 El modelo energético de edificios (BEM) para simulaciones del clima
urbano.
A pesar de que con el desarrollo y el uso de las nuevas parametrizaciones urbanas se
ha comprendido mejor el fenómeno de la isla de calor, aún queda mucho trabajo por
desarrollar. En la mayoría de los esquemas urbanos no se calcula el calor antropogénico
derivado directamente de las actividades humanas (consumo energético) y por consiguiente
no se puede evaluar su efecto. Estudios recientes han demostrado que el calor antropogénico
liberado en las grandes ciudades puede contribuir al aumento de la isla de calor de forma
notable aumentando la temperatura. Por ejemplo, Kikegawa et al., (2003), demostró que los
36
Capítulo 1
efectos de los aires acondicionados sobre la ciudad de Tokio (Japón) aumentaban la
temperatura del aire en un valor promedio de 1 ºC. Si a esto le sumamos el calor producido
por el tráfico y las zonas industriales nos podemos hacer una idea de la importancia que puede
llegar a tener el calor derivado de las distintas actividades humanas. Para que a partir de un
esquema urbano acoplado a un modelo atmosférico se pueda calcular de una forma más o
menos realista el calor antropogénico generado en una ciudad, es preciso el uso de un modelo
energético de edificios. Al estar interesados en el calor antropogénico liberado en la atmósfera
y en la evaluación de estrategias de ahorro energético a escalas de varios kilómetros, no
podemos hacer uso de software especializado resolviendo individualmente cada uno de los
edificios que componen una ciudad. El volumen de datos necesarios para describir
detalladamente los edificios hace inviable tal cálculo. Sin embargo, desarrollando un modelo
energético lo suficientemente simple como para poder ser utilizado sobre todos los edificios
de una determinada clase urbana (donde suponemos que la diferencia entre ellos radique
únicamente en las dimensiones) y lo suficientemente detallado como para que los resultados
obtenidos sean razonables, podemos calcular cuantitativamente el calor antropogénico sobre
una ciudad una vez que este modelo energético esté acoplado a un modelo atmosférico. El
primer modelo energético diseñado con este propósito y acoplado a una parametrización
urbana fue el modelo energético desarrollado por Kikegawa et al., (2003). En este primer
modelo un edificio era tratado como una “caja” y la evolución de la temperatura y humedad
interiores eran calculadas separadamente. El calor liberado en la atmósfera era “proporcional”
al flujo responsable de la variación de la temperatura y humedad interiores.
En esta memoria se ha desarrollado un nuevo modelo energético BEM (Building
Energy Model) donde un edificio es considerado como un apilamiento de “cajas” y cada caja
representa una planta o un piso en particular. El modelo así definido presenta una ventaja
37
Capítulo 1
respecto al modelo anterior, ya que plantas situadas a diferentes alturas pueden recibir
diferente radiación exterior y por consiguiente liberar diferente cantidad de calor
antropogénico proveniente del uso de los aires acondicionados. Además se puede hacer
coincidir cada nivel de la rejilla urbana con cada una de las plantas propias de cada edificio
tipo y así los diferentes flujos pueden ser calculados en cada nivel vertical. En BEM la
evolución de la temperatura y humedad interiores se calculan separadamente para cada planta
y se resuelve:
•
La ecuación de difusión para las distintas superficies (muros, suelos y tejados)
imponiendo un balance energético tanto en la capa exterior como en la interior
de cada superficie.
•
La ventilación natural.
•
La radiación (de onda larga y corta) que alcanza las superficies interiores de las
paredes de cada planta considerando múltiples reflexiones en el interior de la
misma.
•
El calor generado por los ocupantes y equipos domésticos.
•
El calor sensible y latente intercambiado con el exterior gracias al desarrollo de
un modelo para los sistemas de aire acondicionado.
Con este modelo que describe el funcionamiento de los sistemas de aire
acondicionado, el usuario puede definir libremente un umbral de confort para la temperatura y
humedad interiores. De este modo la temperatura y humedad siempre estarán dentro de esa
banda de confort (cuando los sistemas de aire acondicionado estén funcionando lógicamente)
y tanto el consumo como el calor antropogénico intercambiados con la atmósfera (ver todos
los detalles del modelo en Salamanca et al., 2010a) podrán ser calculados. El modelo
energético BEM ha sido acoplado al esquema urbano BEP (Martilli et al., 2002) para poder
38
Capítulo 1
estudiar su impacto en las diferentes variables meteorológicas. Esta nueva parametrización
BEP+BEM acoplada a un modelo atmosférico nos permite evaluar el consumo sobre una
ciudad así como diferentes estrategias para la reducción tanto de la isla de calor como del
consumo energético.
Varias simulaciones (off-line) en una columna vertical han sido llevadas a cabo en esta
tesis con los esquemas urbanos BEP y BEP+BEM. Por encima del canyon urbano se forzaron
los distintos esquemas a los valores medidos del viento, la temperatura y la radiación dirigida
hacia el suelo (downward radiation). Los resultados obtenidos para la radiación neta y para
diferentes flujos turbulentos derivados con las parametrizaciones a diferentes alturas, fueron
comparados con valores medidos recogidos durante la campaña meteorológica BUBBLE
(Basel Urban Boundary Layer Experiment) llevada a cabo sobre la ciudad suiza de Basel
(Rotach et al., 2005). Los resultados mejoraron cuando el nuevo esquema urbano BEP+BEM
fue utilizado frente al viejo esquema BEP (Salamanca & Martilli, 2010).
1.4.3 Problemas abiertos en las simulaciones del clima urbano a mesoescala.
En una sección anterior (ver la sección 1.4) comentamos que una resolución horizontal
excesiva podría llevar a la formación de estructuras celulares artificiales que enmascaran la
verdadera solución de nuestro problema. Si bien estas soluciones no deseadas pudieran ser
dependientes del esquema turbulento utilizado, cuando modelizamos zonas urbanas se nos
puede presentar un nuevo problema añadido. En las distintas parametrizaciones urbanas, las
propiedades térmicas de los materiales son de vital importancia para la estimación de los
diferentes flujos intercambiados entre los edificios y la atmósfera. Esto es así ya que la
temperatura que alcanza una determinada superficie depende fuertemente de sus propiedades
térmicas. Si en las simulaciones utilizamos resoluciones horizontales del orden de uno o unos
pocos kilómetros y una determinada zona urbana presenta una gran heterogeneidad de
39
Capítulo 1
materiales, ¿cuáles deben ser los valores para la capacidad calorífica y conductividad térmica
que debemos utilizar para representar a esa zona en el modelo si ésta contiene diferentes
materiales con diferentes propiedades térmicas?
Si conocemos el porcentaje de los materiales presentes en la zona podemos derivar los
valores térmicos (capacidad calorífica y conductividad térmica) que representan a la clase
urbana de diferentes formas. En Salamanca et al., (2009), analizamos dos formas estándar de
calcular esos valores y proponemos una nueva que mejora los resultados obtenidos
notablemente. La idea consiste en suponer que con los valores térmicos representativos de la
clase urbana debemos obtener el mismo flujo de calor (sensible en este caso) que si
sumáramos el flujo de calor proveniente de los diferentes materiales teniendo en cuenta el
porcentaje presente de cada uno de ellos en la zona bajo estudio. Resolviendo la ecuación de
difusión del calor hemos podido analizar la ventaja de la nueva forma de “promediar” las
propiedades de los materiales frente a las estándar. Los métodos estándar consisten en
“promediar” los valores térmicos de las propiedades físicas de cada material y no los flujos
intercambiados con la atmósfera.
1.5 El modelo atmosférico WRF. Parametrizaciones urbanas en WRF.
El modelo WRF (Weather Research and Forecasting) es el modelo atmosférico
elegido para integrar la nueva parametrización (BEP+BEM) y con el que se han realizado
distintas simulaciones sobre varias ciudades. En esta sección no vamos a explicar detalles de
este modelo pues existe infinidad de documentación (http://www.mmm.ucar.edu/wrf/users/)
especializada dedicada íntegramente a ello. La ventaja del modelo WRF frente a otros
modelos es que es un modelo libre (totalmente gratuito), que está en continua evolución y
existe un apoyo para que los usuarios puedan consultar sus dudas o resolver sus problemas
gracias al NCAR que es la institución de los EEUU que lo mantiene y desarrolla.
40
Capítulo 1
Mucho es el trabajo que se ha hecho en el modelo WRF por parte de la comunidad de
modelizadores urbanos desde su aparición. Actualmente este modelo (v3.2) posee cuatro
parametrizaciones urbanas. La primera parametrización (bulk scheme) aparece en la versión
del año 2003. Este primer esquema representa los efectos de las superficies urbanas por medio
de una rugosidad de 0.8 m, un albedo de 0.15, una capacidad calorífica de 3.0 J/m 3 K y una
conductividad térmica de 3.24 W/mK . Debido a su sencillez el esquema es ideal para la
predicción meteorológica rutinaria y así lo prueban numerosos estudios (Liu et al., 2006). Sin
embargo, posee una desventaja frente a otras parametrizaciones más detalladas, ya que no
puede distinguir la heterogeneidad presente en una ciudad al no utilizar datos morfológicos de
ningún tipo. La segunda parametrización incluida en WRF fue la parametrización desarrollada
por Kusaka et al., (2001) y Kusaka & Kimura, (2004). Este esquema urbano está disponible
desde la versión oficial v2.2 del año 2006. Es un modelo con una sola capa (single-layer
urban canopy model) en el canyon urbano donde un perfil diario de calor antropogénico
puede ser añadido al flujo total de calor sensible. La geometría urbana está representada a
través de canyons urbanos infinitamente largos y tres diferentes superficies (muros, calles y
tejados) son reconocidas. En el canyon urbano múltiples reflexiones y los efectos de las
sombras son considerados. La tercera opción urbana es el esquema desarrollado por Martilli et
al., (2002). Esta parametrización es un esquema multicapa que permite una interacción directa
con la capa límite urbana. El esquema multicapa también distingue tres tipos de superficies
urbanas y tiene en cuenta su efecto en el momento, la energía cinética turbulenta y la
temperatura del aire. Múltiples reflexiones y efectos de sombreado son considerados también
en el canyon urbano. El modelo BEP de Martilli et al., (2002) existe de forma oficial en WRF
desde la versión v3.1 del año 2009. Por último, de forma oficial ha sido integrada en el
modelo WRFv3.2 la parametrización BEP+BEM desarrollada en esta memoria. Este nuevo
41
Capítulo 1
esquema BEP+BEM es el resultado del acoplamiento de la parametrización BEP (Martilli et
al., 2002) con un modelo energético de edificios BEM (Salamanca & Martilli, 2010). Esta
última incorporación permite calcular el consumo energético de los aires acondicionados así
como su efecto sobre la temperatura del aire en la ciudad. Con esta parametrización se calcula
el calor antropogénico, a diferencia de otras parametrizaciones donde se fija arbitrariamente
un perfil diario definido por el propio usuario. La necesidad de un mayor volumen de datos
para llevar a cabo las simulaciones se ha visto incrementada notablemente, a medida que la
complejidad de las parametrizaciones urbanas ha ido aumentando. Aunque en un principio
esto pudiera parecer una desventaja, el uso de las parametrizaciones más complejas permite
llevar a cabo estudios que de otra forma no serían posibles (comprensión del fenómeno de la
isla de calor, estrategias de mitigación, reducción del consumo energético, etc.).
Con el modelo WRFv3.2 se ha simulado la ciudad de Madrid. Usando los datos de la
campaña meteorológica DESIREX (Sobrino et al., 2009) que tuvo lugar en el verano del
2008, se ha validado la parametrización BEP+BEM sobre esta ciudad simulando dos días
consecutivos con buenas condiciones sinópticas. Se utilizó un fichero actualizado de usos de
suelo para la Comunidad de Madrid así como tres clases urbanas derivadas de la base de datos
de la Agencia Europea de Medio Ambiente CORINE (http://www.eea.europa.eu). El efecto de
los aires acondicionados, la magnitud de la isla de calor y la evaluación de diferentes
estrategias en la reducción del consumo energético arrojaron interesantes resultados (más
detalles en Salamanca et al., 2010c). La isla de calor urbana alcanzó entre 5-6 ºC durante la
noche y el calor antropogénico fue el responsable de un aumento de la temperatura de hasta
1.5-2 ºC en algunos puntos interiores de la ciudad. Las diferentes estrategias encaminadas al
ahorro del consumo energético (aumento del albedo de los tejados, el uso de materiales
aislantes y la eliminación del calor proveniente de los aires acondicionados) una vez que
42
Capítulo 1
fueron agrupadas disminuyeron notablemente la magnitud de la isla de calor y el ahorro
energético fue reducido en torno a un 10-10.5 %.
Aunque una cuantificación precisa del consumo energético sobre una ciudad requiere
del uso de una detallada base de datos morfológicos de la misma (ver la sección siguiente
1.5.1), estimaciones del ahorro energético y evaluación de diferentes estrategias de mitigación
tanto de la isla de calor como del consumo energético pueden llevarse a cabo sin el
conocimiento exacto de la geometría de la ciudad.
1.5.1 La base de datos NUDAPT. La versión de WRF adaptada a NUDAPT.
Cuando se simula una ciudad con una parametrización urbana, ésta es llamada por el
modelo atmosférico cuando el punto de la rejilla posee una fracción urbana no nula. En el
modelo WRF a cada punto de rejilla urbana se le asocia una sola clase urbana donde los
parámetros urbanos deben ser definidos por el propio usuario (con la excepción del esquema
bulk que no utiliza los valores definidos en la clase urbana sino que tiene asignados unos
valores por defecto). Por defecto, en el modelo WRF se pueden definir hasta tres tipos
diferentes de clases urbanas (comerciales o industriales, zonas residenciales con densidad alta
y zonas residenciales con densidad baja de edificios) siempre que se lo indiquemos en el
fichero de usos de suelo que alimentará la simulación. Por consiguiente, podemos distinguir
hasta tres tipos diferentes de zonas urbanas en nuestra malla numérica. Así es como se
procedió para el caso de las simulaciones que se llevaron a cabo sobre la ciudad de Madrid.
Probablemente para una gran cantidad de estudios sobre clima urbano sea suficiente
proceder de esta forma. Sin embargo, recientemente en Norteamérica se ha creado una base
de datos NUDAPT (National Urban Database and Access Portal Tool, Ching et al., 2009)
con información detallada de la morfología urbana de las principales ciudades de este país. La
fracción urbana ocupada por los edificios, la altura media de los mismos, la anchura media de
43
Capítulo 1
las calles, el número de edificios de una determinada altura, etc. son parámetros morfológicos
que se encuentran en esta base de datos con una resolución mínima de 1 km 2 . Un ejemplo
concreto de la información que existe en NUDAPT puede verse en la Fig. 1.3. En esta gráfica
se representa la altura media y la fracción de área cubierta por los edificios para una región
que cubre la totalidad de la ciudad de Houston y tiene una extensión aproximada de unos
5250 km 2 . La ventaja de utilizar esta información es que no necesitamos definir clases
urbanas cuando se quiere simular una ciudad. En cada punto de la rejilla urbana disponemos
de la información necesaria para que la parametrización urbana proporcione las condiciones
de borde al modelo atmosférico. De esta forma cada punto numérico urbano viene
representado por los parámetros reales que lo describen a una determinada resolución. Esta es
la manera más realista de proceder a la hora de simular una ciudad. En Salamanca et al.,
(2010b), se simuló la ciudad de Houston (TEXAS) haciendo uso de dos metodologías
totalmente diferentes. En la primera, se utilizaron tres clases urbanas derivadas de la base de
datos NLCD (National Land Cover Data) comparándose las cuatro parametrizaciones urbanas
existentes en WRF con medidas observadas. Los valores morfológicos definidos para cada
clase urbana fueron derivados de Burian & Han, (2003). En la segunda parte del trabajo
(segunda metodología) se utilizó la información que existe en NUDAPT para un área que
cubrió la ciudad de Houston.
44
Capítulo 1
2
Figura 1.3. a) Altura media y b) fracción del área ocupada por los edificios con una resolución de 1 km en la
región de Houston, TEXAS (EEUU).
En los ficheros de entrada del modelo atmosférico se integró la información de la base
de datos en forma de nuevas variables. De modo que ahora el modelo atmosférico WRF
(WRF_Nudapt) utilizaba para cada punto de la rejilla numérica urbana la información
45
Capítulo 1
existente en la base de datos NUDAPT y no la información definida para cada clase urbana.
Los resultados obtenidos muestran una mejoría en la predicción de la temperatura del aire
cuando se comparan con los obtenidos en la primera parte del trabajo. Sin embargo, el mayor
impacto se obtuvo en el cómputo del consumo energético total de la ciudad. Utilizando la
información de NUDAPT se obtuvieron interesantes estimaciones del consumo energético
cuando se compararon con los valores obtenidos (Heiple and Sailor, 2008) por otros métodos
totalmente diferentes (bottom-up y top-down). La versión del modelo atmosférico que es
capaz de utilizar la información de NUDAPT punto a punto en el dominio numérico
(WRF_Nudapt) está siendo utilizada por personal del NCAR (Dr. Mukul Tewari) y de la EPA
(Dr. Jason Ching) ya que esta versión representa el mayor desarrollo del modelo atmosférico
WRF para la modelización urbana a mesoescala.
46
Capítulo 1
APÉNDICE
Símbolos de las ecuaciones
alb
albedo
ε
emisividad
σ (W / K 4 m 2 ) constante de Stefan-Boltzmann
S ↓ (W / m 2 ) flujo de radiación de onda corta dirigida hacia el suelo
L ↓ (W / m 2 ) flujo de radiación de onda larga dirigida hacia el suelo
TS ( K )
temperatura superficial
T (K )
temperatura del aire
ϑ(K )
temperatura potencial del aire
q ( kg / kg )
humedad específica del aire
ρ (kg / m 3 )
densidad del aire
C P ( J / kgK ) calor específico del aire a presión constante
LV ( J / kg )
calor latente de vaporización
k (W / mK )
conductividad térmica
g (m / s 2 )
aceleración debida a la gravedad
f ( s −1 )
parámetro de Coriolis
P (kg / ms 2 ) presión atmosférica
U i (m / s)
componente en la dirección i de la velocidad del viento
u ′j (m / s )
componente turbulenta en la dirección j de la velocidad del viento
u ′(m / s )
componente turbulenta en la dirección x de la velocidad del viento
v ′(m / s )
componente turbulenta en la dirección y de la velocidad del viento
w′(m / s )
componente turbulenta en la dirección z de la velocidad del viento
q ′(kg / kg ) componente turbulenta de la humedad específica del aire
ϑ ′( K )
componente turbulenta de la temperatura potencial del aire
ν (m 2 / s)
viscosidad molecular cinemática
ν ϑ (m 2 / s ) difusividad térmica molecular
47
Capítulo 1
ν q (m 2 / s) difusividad molecular para el vapor de agua en el aire
Q *j (W / m 2 ) radiación neta en la dirección j
48
Capítulo 2
49
Capítulo 2
CAPÍTULO 2
OBJETIVOS
50
Capítulo 2
2.1 Introducción
El principal objetivo de este trabajo es el estudio y modelización de la isla de calor
urbana. Se prestará una atención especial a la evaluación de diferentes estrategias de
mitigación de la isla de calor y del consumo energético debido al uso de los sistemas de aire
acondicionado en períodos estivales. Para llevar a cabo estos objetivos presentamos a
continuación los pasos y la metodología utilizados.
Objetivo 1
Desarrollo de un modelo energético de edificios (BEM) acoplado a una parametrización
urbana para simulaciones del clima urbano.
Este es el primer objetivo que debemos lograr para poseer una parametrización urbana
capaz de calcular el calor antropogénico originado por los sistemas de aire acondicionado o
calefacción en las ciudades y estudiar su impacto en las diferentes variables meteorológicas.
Para esta parte del trabajo se ha preferido desarrollar un nuevo modelo energético y no hacer
uso de software avanzado, propio de los estudios de ingeniería ya que el acoplamiento de
estos programas con los modelos atmosféricos es impracticable debido a que son muy
detallados y no han sido pensados para tal fin. Una vez que el modelo energético BEM ha
sido acoplado a la parametrización urbana BEP, se han realizado diferentes simulaciones con
el fin de dar respuesta a las siguientes preguntas:
•
¿Cuál es el impacto del calor emitido por los sistemas de aire acondicionado en la
temperatura del aire externa? ¿El modelo energético BEM nos puede ayudar a
obtener una respuesta?
•
¿Los diferentes flujos de calor mejoran cuando se compara el nuevo esquema
urbano BEP+BEM frente al viejo esquema BEP?
•
¿Cuánto se ahorra en consumo energético debido al uso de los aires
51
Capítulo 2
acondicionados cambiando el albedo e introduciendo materiales aislantes en los
muros de los edificios?
•
¿Cómo depende este consumo de las condiciones meteorológicas exteriores?
La respuesta a estas preguntas se desarrolla en los dos primeros artículos presentados en el
capítulo 3.
Objetivo 2
Estudio de la representatividad de diferentes parámetros térmicos utilizados en las
parametrizaciones urbanas.
Este objetivo surge a raíz del planteamiento de que en las simulaciones a mesoescala
(con resoluciones numéricas del orden de ~ 1 km) del clima urbano nos podemos encontrar
con una gran heterogeneidad de materiales presentes en un área particular que está
representada por un solo punto en la rejilla numérica. A la hora de calcular los flujos de calor
se resuelve un balance de energía considerando, por simplificar el cálculo, un solo valor de
capacidad calorífica y conductividad térmica para el suelo (en las parametrizaciones más
simples o en zonas rurales), o tres (para calle, techos y paredes) en las urbanas más complejas.
El enfoque más usado consiste en utilizar los valores térmicos del material con una mayor
representatividad en la zona. Sin embargo, otra forma de proceder para determinar los valores
más representativos es la de promediar los distintos valores de las propiedades térmicas de
cada material teniendo en cuenta el porcentaje de área superficial que ocupa cada uno de
ellos. El objetivo es buscar una respuesta a las siguientes preguntas:
•
¿Existe alguna otra forma de calcular las propiedades térmicas representativas de
una zona urbana que mejore el cálculo de los flujos de calor sensible?
•
¿En cuánto se mejoran los resultados?
La respuesta a estas preguntas se encuentra en el tercer artículo del capítulo 3.
52
Capítulo 2
Objetivo 3
Integrar la nueva parametrización urbana BEP+BEM de forma oficial en el modelo
atmosférico WRF.
En la versión v3.2 del modelo atmosférico WRF (liberada por NCAR en Abril del
2010 y disponible online) se encuentra la nueva parametrización urbana BEP+BEM
desarrollada en esta tesis. En WRF, los esquemas urbanos BEP y BEP+BEM pueden utilizar
dos esquemas turbulentos diferentes para la parametrización de la capa límite atmosférica, el
esquema de Bougeault & Lacarrère, (1989) y el esquema de Mellor & Yamada (Janjic, 1994).
Objetivo 3.1. Inter-comparación de diferentes esquemas urbanos en WRF. Simulaciones
sobre Houston, Texas.
Una vez que la parametrización BEP+BEM fue integrada en el modelo atmosférico
WRF, se procedió a la simulación de la ciudad de Houston, Texas. En este estudio se hizo una
inter-comparación de los cuatro esquemas urbanos presentes en el modelo utilizando
diferentes estaciones de medida distribuidas por la ciudad. Tres clases urbanas derivadas de la
base de datos NLCD (National Land Cover Database) fueron definidas (comerciales o
industriales, zonas residenciales con alta y baja densidad de edificios) y los parámetros
urbanos utilizados para cada una de ellas fueron extraídos de Burian & Han, (2003). Se
pretende dar respuesta a las siguientes preguntas:
•
¿Existen diferencias importantes en los resultados al utilizar las cuatro
parametrizaciones urbanas?
•
¿Cuál es el impacto de los sistemas de aire acondicionado en la temperatura del
aire sobre la ciudad?
•
¿Es realista el cálculo del consumo energético obtenido utilizando el modelo
atmosférico y la nueva parametrización BEP+BEM?
53
Capítulo 2
La respuesta a estas preguntas se desarrolla en el cuarto artículo del Capítulo 3.
Objetivo 3.2. Evaluación de estrategias para la reducción del consumo energético y
mitigación de la isla de calor. Simulaciones sobre Madrid.
Para el cumplimiento de este objetivo se hicieron simulaciones con WRF usando el
nuevo esquema urbano BEP+BEM sobre la ciudad de Madrid. El período simulado coincidió
con la campaña DESIREX que tuvo lugar en el verano del 2008. Tres clases urbanas
derivadas de la base de datos de usos de suelo CORINE (http://www.eea.europa.eu) fueron
definidas sobre la ciudad. Una vez que la validación de las simulaciones estuvo garantizada
por comparación con las medidas, se procedió a la evaluación de diferentes estrategias de
reducción tanto de la isla de calor como del consumo energético. El objetivo es contestar a las
siguientes preguntas:
•
¿Cuál fue la magnitud máxima de la isla de calor sobre la ciudad?
•
¿En cuántos grados aumentó la temperatura del aire debido al calor
antropogénico procedente de los sistemas de aire acondicionado?
•
¿Cuál es la estrategia más efectiva para ahorrar en consumo energético y reducir
la isla de calor?
La respuesta a estas preguntas se encuentra en el quinto artículo del Capítulo 3.
Objetivo 4
Adaptar el modelo atmosférico WRF a la base de datos NUDAPT.
Cuando simulamos una ciudad, las parametrizaciones urbanas en WRF utilizan la
información definida en las distintas clases urbanas con la excepción del esquema bulk que
solo reconoce una clase urbana. De este modo, la morfología urbana representada en el
modelo está limitada al número de clases urbanas que utilicemos y por defecto a lo sumo son
tres. Sin embargo, si introducimos la información de la morfología que describe una ciudad
54
Capítulo 2
(esta información existe en la base de datos NUDAPT) como nuevas variables en los ficheros
de entrada del modelo atmosférico, las parametrizaciones urbanas utilizarán en cada punto de
la rejilla numérica la información real que describe esa zona de la ciudad. La ventaja de
proceder de esta forma es que aprovechamos toda la información morfológica que describe la
ciudad y no estamos limitados al número existente de clases urbanas.
Así se procedió de nuevo con la ciudad de Houston (Texas) repitiéndose las
simulaciones con los esquemas BEP y BEP+BEM. El objetivo es dar respuesta a las
siguientes preguntas:
•
¿Se mejora la predicción de la temperatura del aire con esta nueva forma de
proceder?
•
¿Se mejora el cálculo del consumo energético?
La respuesta a estas preguntas se encuentra en el cuarto artículo del Capítulo 3.
55
Capítulo 2
56
Capítulo 2
CHAPTER 2
OBJECTIVES
57
Capítulo 2
2.1 Introduction
The principal aim of this work is the study and modelling of the urban heat island. We
will pay a special attention to the evaluation of different strategies to mitigate the urban heat
island and reduce the energy consumption due to the use of the air conditioning systems
reduction programs in summer periods. To reach the methodology used is presented.
Objective 1
Development of a building energy model coupled to an urban canopy parameterization
for simulations of the urban climate.
This is the first aim that we must achieve to have an urban scheme able to compute the
anthropogenic heat originated in the cities from air conditioning or heating, and study its
impact on the different meteorological variables. For this part of the work we preferred to
develop a new model and not using existing software typical of engineering studies since they
are in general to detailed since they have not been built for this aim. As soon as the building
energy model (BEM) has been coupled to the urban scheme (BEP), different simulations have
been carried out in order to answer the following questions:
•
What is the impact of heat emitted from air conditioning systems on the air
temperature? Can we get this answer from BEM?
•
Does the calculation of the different heat fluxes improve when the new urban
scheme BEP+BEM is compared with the previous BEP scheme?
•
How much is the energy saving due to the use of the air conditioning systems
when the albedo is modified and insulating material is intoduced inside the
walls of the buildings?
•
How much can the meteorological conditions affect the energy consumption?
The answer to these questions is in the first and second papers of Chapter 3.
58
Capítulo 2
Objective 2
On the derivation of material thermal properties representative of heterogeneous urban
neighbourhoods for mesoscale simulations.
This aim arises because in urban climate mesoscale simulations (with a typical
resolution of ~ 1 km) we can find a great heterogeneity of materials in a particular area that is
represented by only one point in the numerical grid. To estimate the heat fluxes a surface
energy budget is solved, with only one value of heat capacity and thermal conductivity for the
surfaces (for the simplest urban schemes or for rural areas), or three values (for road, walls,
and roofs) for the most complex urban schemes. In general the values chosen correspond to
the material with a major percentage in the area. Another way to determine the most
representative thermal values is to average the different values of the thermal properties of
every material taking into account the percentage of surface area that each of them occupies.
The main questions that we try to answer in this section are:
•
Is there another way to calculate the thermal representative properties of an urban
zone that improves the calculations of the sensible heat fluxes?
•
How much are the results improved?
The answer to these questions is in the third paper of Chapter 3.
Objective 3
Integrate the new urban scheme BEP+BEM officially in the atmospheric WRF model.
In the version 3.2 of WRF model (released on April, 2010) the new urban scheme
BEP+BEM is present. In WRF, the urban parameterizations BEP and BEP+BEM can be used
with two different turbulent schemes for the parameterization of the planetary boundary layer,
the Bougeault & Lacarrère, (1989) and the Mellor & Yamada (Janjic, 1994) schemes.
Objective 3.1. Inter-comparison of different urban schemes in WRF. Simulations over
59
Capítulo 2
Houston, Texas.
As soon as the urban parameterization BEP+BEM was integrated in the atmospheric
model WRF, the model was used to simulate the atmospheric circulations over the city of
Houston, Texas. In this study an inter-comparison of the four urban schemes present in the
model was done using different measurement stations distributed over the city and belonging
to the Texas Commission on Environmental Quality. Three urban classes derived from the
NLCD (National Land Cover Database) were defined (commercial or industrial, and
residential zones with high and low density of buildings) and the urban parameters used for
each of them were extracted from Burian & Han, (2003). This objective deals with the
following questions:
•
Are there important differences in the results among the four urban schemes?
•
Which is the impact of the air conditioning systems on the air temperature?
•
Is the calculation of the energy consumption obtained using the atmospheric
model and the new urban BEP+BEM scheme realistic?
The answer to these questions is in the fourth paper of the Chapter 3.
Objective 3.2. Evaluation of strategies for energy consumption reduction and mitigation
of the urban heat island. Simulations over Madrid.
To achieve this aim, simulations with WRF using the new urban BEP+BEM scheme
were done over the city of Madrid. The simulated period coincided with the DESIREX
campaign that took place in summer of 2008. Three urban classes derived from the land use
database CORINE (http://www.eea.europa.eu) were defined. As soon as the validation of the
simulations was guaranteed by comparing the results against measurements, we proceeded to
the evaluation of different mitigation strategies of the urban heat island and reduction
programs of the energy consumption. The objective is to answer the following questions:
60
Capítulo 2
•
Which was the maximum magnitude of the urban heat island over the city?
•
How many degrees increased the air temperature due to the heat fluxes coming
from the air conditioning system?
•
Which is the most effective strategy to save in energy and how much the urban
heat island is reduced?
The answer to these questions is in the fifth paper of Chapter 3.
Objective 4
Update the atmospheric WRF model to the NUDAPT database.
When we simulate a city, the urban parameterizations in WRF use the information
defined in the different urban classes with the exception of the bulk scheme that has only one
urban class. In this way, the urban morphologies represented in the model are limited to the
number of urban classes used, which are three by default. If instead of using the urban classes,
we introduce the information of the morphology that describes the city (information that
exists in the NUDAPT database) as new variables in the input files of the atmospheric model,
the urban parameterizations will use at every point of the numerical grid the real information
that describes this area of the city. The advantage of this approach is that we take advantage
of all the morphological information that describes the city and are not limited to the existing
number of urban classes.
In this way we proceeded again with the city of Houston, Texas. The simulations with
the urban schemes BEP and BEP+BEM were repeated and compared with the old ones. Here
the objective is to answer the following questions:
•
Is the prediction of the air temperature improved by this approach?
•
Is the calculation of the energy consumption improved?
The answer to these questions is in the fourth paper of the Chapter 3.
61
Capítulo 2
62
Capítulo 3
CAPÍTULO 3
RESULTADOS
63
Capítulo 3
3.1 Un nuevo modelo energético de edificios acoplado a una parametrización urbana
para simulaciones del clima urbano-Parte-I. Formulación, verificación y análisis
sensitivo del modelo.
Salamanca, F., A. Krpo, A. Martilli, and A. Clappier, 2010a: A New Building Energy Model coupled with
an urban canopy parameterization for urban climate simulations-part I. formulation, verification, and
sensitivity analysis of the model. Theor. Appl. Climatol., 99, 331-344.
En esta sección el nuevo modelo energético de edificios BEM (Building Energy Model) se
desarrolla
e
implementa
en
la
parametrización
urbana
BEP
(Building
Effect
Parameterization). El modelo energético se compara con otros modelos usados en el análisis
térmico de edificios: CBS-MASS, BLAST (1981) y TARP (Walton, 1983). Posteriormente se
hace un análisis de la sensibilidad del modelo a diferentes parámetros, así como un primer
estudio del impacto del modelo energético en la parametrización urbana. Las validaciones
indican que el modelo energético proporciona buenas estimaciones del comportamiento físico
de los edificios y es un primer paso hacia el desarrollo de una herramienta numérica para
estudios de planificación urbana.
64
Theor Appl Climatol (2010) 99:331–344
DOI 10.1007/s00704-009-0142-9
ORIGINAL PAPER
A new building energy model coupled with an urban canopy
parameterization for urban climate simulations—part I.
formulation, verification, and sensitivity analysis
of the model
Francisco Salamanca & Andrea Krpo & Alberto Martilli &
Alain Clappier
Received: 21 October 2008 / Accepted: 21 April 2009 / Published online: 13 May 2009
# Springer-Verlag 2009
Abstract The generation of heat in buildings, and the way
this heat is exchanged with the exterior, plays an important
role in urban climate. To analyze the impact on urban
climate of a change in the urban structure, it is necessary to
build and use a model capable of accounting for all the
urban heat fluxes. In this contribution, a new building
energy model (BEM) is developed and implemented in an
urban canopy parameterization (UCP) for mesoscale
models. The new model accounts for: the diffusion of heat
through walls, roofs, and floors; natural ventilation; the
radiation exchanged between indoor surfaces; the generation of heat due to occupants and equipments; and the
consumption of energy due to air conditioning systems.
The behavior of BEM is compared to other models used in
the thermal analysis of buildings (CBS-MASS, BLAST,
and TARP) and with another box-building model. Eventually, a sensitivity analysis of different parameters, as well
as a study of the impact of BEM on the UCP is carried out.
The validations indicate that BEM provides good estimates
of the physical behavior of buildings and it is a step
towards a modeling tool that can be an important support
to urban planners.
F. Salamanca (*) : A. Martilli
Department of Environment, CIEMAT
(Center for Research on Energy, Environment and Technology),
Edificio 3, P1.9, Avenida Complutense 22,
28040 Madrid, Spain
e-mail: [email protected]
A. Krpo : A. Clappier
LPAS, Swiss Federal Institute of Technology,
Lausanne, Switzerland
1 Introduction
In the last decades, atmospheric scientists have been able to
understand the origin of the temperature differences
between an urban area and its surroundings, the so-called
urban heat island (UHI) (Oke 1987). This understanding
has been possible thanks to a series of experimental
campaigns, to the evolution of the mesoscale meteorological models, and the increase of computer power. In the
1970s and 1980s, scientists began to introduce urban
parameterizations in numerical mesoscale models to determine how cities affect the meteorological fields and the
boundary layer structure. However, these first parameterizations were still very simple and could not reproduce in
detail the dynamics of the interactions between a city and
the atmosphere. It is only during the second part of the
1990s and the beginning of the twenty-first century that
more realistic urban parameterizations appeared (Masson
2000; Kusaka et al. 2001; Martilli et al. 2002).
The models developed in this period allowed to better
understand the phenomena linked to the atmosphere over
cities and their surroundings. However, the generation of
heat within buildings and the exchanges with the exterior
were not explicitly resolved. One of the first models that
took these features into account was the one developed by
Kikegawa et al. (2003). It was successfully implemented in
an urban canopy parameterization (UCP) for mesoscale
models and clearly showed that the heat fluxes generated
by building can have an important impact on the urban
microclimate.
In this contribution, a new building energy model (BEM)
has been worked up and implemented in an UCP for
mesoscale models. It is important to remark that, due to
computational requirements, we cannot take into account
65
332
F. Salamanca et al.
all the details in the interactions between buildings and the
atmosphere. In fact, reducing complexity is particularly
important as the final goal is to link BEM with a mesoscale
meteorological model. Moreover, the current computing
capacity does not allow resolving each specific building
included in a grid cell of the meteorological model, usually
of the order of a few squared kilometers. Even though all
buildings are different, it is necessary to develop a model
that describes the general physical properties of an ideal
building, representative of the buildings included within the
grid cell, as the purpose is to investigate the interactions
between urban climate, air pollution, and energy consumption at the scale of the city and its surroundings. On the
other hand, a very simple BEM would not be capable of
describing accurately the most important interactions
between buildings and the atmosphere, and would not be
sufficient to study the interactions mentioned above, when
implemented in a mesoscale model.
For these reasons, in this work we propose a new model
that resolves explicitly:
&
&
&
&
the heat diffusion through walls, roofs and floors
the natural ventilation as well as the radiation exchanged between the indoor surfaces
the heat generation due to occupants and equipments
the energy consumption due to air conditioning systems
Buildings of several floors can be considered and the
time evolution of indoor air temperature and moisture are
estimated for each floor. Different floors can receive
different amounts of radiation and can have different
temperatures, both outdoor (e. g. for skyscrapers) and
indoor. It is logical to think that the cooling/heating loads
(energy consumption) will also be different at each level.
The links between BEM and UCP are as follows: UCP
gives to BEM the outdoor air temperature, humidity, and
radiation reaching the walls and roof for the computation of
the amount of radiation entering in the building through the
windows and the boundary condition for the calculation of
wall and roof temperatures; on the other hand, BEM gives
to the UCP the wall and roof temperature, the heat flux due
to ventilation, and the heat flux due to processes linked
with the generation of energy within the building (e. g. air
conditioning). The expected results of this study are:
&
&
To improve the capability of mesoscale models to
simulate urban canopy climate (UHI processes, etc.)
and air pollutant dispersion in the city and the
surroundings
To allow the estimation of meteorologically related
building energy consumptions (e. g. due to air conditioning in summer, or heating in winter)
In Section 2 the model formulation is described. In
Section 3, we validate BEM comparing it to well-known
models like CBS-MASS, BLAST, and TARP for three
different situations (Zmeureanu et al. 1987). The total
processed loads obtained with BEM and Kikegawa´s model
in a 25-story office building are also compared. In Section 4,
after this necessary validation, we analyze numerical results
by modifying some physical parameters to evaluate their
impact on the processed load. In Section 5, first results
about the impact of BEM in the UCP of Martilli et al.
(2002) are shown and conclusions are finally given in
Section 6. In part II of this work, BEM coupled with the
UCP is validated against meteorological measurements
from the BUBBLE campaign (Salamanca and Martilli
2009).
2 Description of BEM
The model used here is similar to that of Kikegawa et al.
(2003). In Kikegawa´s model, a building is treated as a box
and the generated cooling/heating loads are separately
calculated for sensible and latent heat components. The
heat pumped out from the building is “proportional” to
this load (more details in Section 2.5). The main differences between the two models are the computation of the
solar radiation reaching the indoor walls, the treatment of
the windows, the computation of the heat pumped out
from the building for cooling or added for heating, and
the possibility to consider several floors in a building.
The BEM developed in this paper is a box-type heat
budget model in which a building in an urban block is
treated as a pile of boxes, each box representing a
particular floor.
2.1 Dynamics and thermodynamics
In BEM, the time evolutions of the room air temperature Tr
and room air humidity qVr are estimated solving the
following equations:
QB
dTr
¼ Hin Hout
dt
lrVB
dqVr
¼ Ein Eout
dt
ð1Þ
ð2Þ
in which QB ¼ rCp VB ðJK 1 Þ and VB(m3) denote the
overall heat capacity and the total volume of the indoor
air in a floor (the reader can see more details about the
symbols used in the previous and following equations in the
Appendix). The following equations (Eqs. 3 and 4) were
used for the computation of the total sensible heat load
66
A new building energy model coupled with an urban canopy parameterization
Hin(W) and the total latent heat load Ein(W) in a floor,
respectively:
Hin ¼
X
j
X wall
Awind
hwind; j Twind; j Tr þ
Ai hwall;i Twall;i Tr þ
j
i
þð1 bÞCp rVa ðTa Tr Þ þ Af qE þ Af Pfp qhs
ð3Þ
Ein ¼ ð1 bÞlrVa ðqVa qVr Þ þ Af Pfp qhl
ð4Þ
The first and second term (on the right-hand side) in
Eq. 3 represent the heat exchange between the windows and
the indoor air and between the walls, ceiling, and pavement
and the indoor air. The third term corresponds to the sensible
heat exchange through ventilation. The fourth and the last
terms indicate the internal sensible heat generation from
equipments and occupants, respectively. The quantification
of these last terms is difficult and it is necessary to have some
information about the energy consumption provided by the
electric companies. The heat from these different processes is
added and distributed isotropically in the interior. A real
diffusion through the indoor air is not considered in the
model. The first right-hand term of Eq. 4 represents the water
vapor mixing through ventilation and the second term the
evaporation from occupants. The terms Hout(W) and Eout(W)
indicate the sensible and latent heat needed for cooling/
heating the indoor air in a floor. Remark that if there is no
human regulation of the internal temperature and humidity
these two terms are zero.
2.2 Computation of the wall temperature
In order to compute the wall temperature, the heat diffusion
equation is solved in several layers at the interior of the
materials. The transport of moisture through the walls is not
considered,
@Twall
@
@Twall
Ks
¼
ð5Þ
@x
@t
@x
where Ks (m2 s−1) is the thermal conductivity of the material,
Twall is the wall temperature. At the indoor and outdoor
surfaces, the boundary condition is defined by solving an
energy budget equation (neglecting the latent heat flux),
@Twall
1
@Twall 1
Cs HF Ks
¼
ð6Þ
Δx
@t
@x n1
4
þ H 1. The term
where HF ¼ ð1 albÞRs þ "Rl "sTwall
Cs (J K−1 m−3) is the specific heat of the layer of depth ∆x,
1
To solve numerically the equation, the wall is discretized in several
layers of depth ∆x. Here and in Eq. 6 Twall represents the temperature
of the layer close to the surface, while @T@xwall n1 represents the gradient
between the layer close to the surface and the closest internal layer.
333
and H (W m−2) is the sensible heat flux exchanged between
the surface and the air (a positive value means a gain for the
surface). It is computed as H=hwall(Tr −Twall) in the indoor
side and H=h(Tr −Twall) in the outdoor side. The term Rs is
the shortwave radiation flux incoming at the surface, Rl is
the long-wave radiation received by the surface, and finally,
alb and ε are the surface albedo and emissivity respectively.
This budget equation is solved on both sides of the wall.
2.3 Computation of the window temperature
We suppose that the differences in temperature between the
two sides of a glass are small, and, as a consequence, the
temperature of the windows is only time-dependent. In order to
compute the temperature of the glass of the window (Twind), we
suppose that the absorption is negligible (glasses without
coating or films) and the following budget equation is solved:
C
dTwind
¼f
dt
ð7Þ
where C ¼ rwind Cwind Δwind JK1 m2 , ρwind (kg m−3)is
the density of the glass, Cwind (JK−1 kg−1) is the heat
capacity
of the
glass, ∆wind (m) is the thickness of the glass,
and f Wm2 is the total flux balance of energy,
4
f ¼ "wind Rloutdoor sTwind
þ Houtdoor
4
þ "wind Rlindoor sTwind
ð8Þ
þ Hindoor
The terms Hindoor and Houtdoor are the sensible heat
fluxes, while Rlindoor and Rloutdoor are the incoming longwave radiation on each side of the window. The windows
are assumed opaque to the long-wave radiation.
2.4 Computation of the radiation
The amount of direct radiation that passes through a
window is a function of the angle of incidence and will
be computed with a polynomial approach based on Roos
(1997) and used by Karlsson and Roos (2000) and others.
The model employs a polynomial to fit the angle
dependence of the total solar energy transmittance g, based
upon the knowledge of the respective near-normal value g0.
The general form of the polynomial is gðzÞ
g0 ð1 aza ¼
0
l
g
0
bz cz Þ, where a þ b þ c ¼ 1, z ¼ q 90 , and θ0 is the
angle of incidence. When fitting to different types of
windows it was found that the above equation gives a good
fit with the following coefficients and exponents:
a ¼ 8; b ¼ 0:25=q; c ¼ 1 a b;
a ¼ 5:2 þ 0:7q; l ¼ 2;
g ¼ 5:26 þ 0:06p þ ð0:73 þ 0:04pÞq
ð9Þ
In Eq. 9, p is equal to the number of panes in the
configuration (1, 2, or 3) and q represents a ‘category’
67
334
F. Salamanca et al.
parameter, which has been given values between 1 and 10
depending on the type of window (q=4, for standard glasses).
The computation of the diffused and reflected radiation
that passes through a window can be calculated using the
albedo of that window (albwind). The albedo of a window
can be evaluated (suppose that the absorption is negligible)
equalizing the energy that crosses the glass with (1−
albwind) times the energy that reaches the window. Writing
this in mathematical form, see Fig. 1, we can say
Z 2p Z p
2
gðqÞI cos qdw dA ¼ ð1 albwind ÞFdA
ð10Þ
0
0
where I is the intensity of the radiation and F is the flux of
energy reaching the element of surface dA (it is obtained
integrating the intensity over all the possible directions).
Considering isotropy (I constant) and simplifying by the
differential area dA, the above expression becomes,
Z2
p
gðqÞ cos q sin qdq ¼ ð1 albwind Þ
2
ð11Þ
0
2.4.1 Shortwave radiation
The method used to compute the radiation reaching the
indoor surface of the walls is similar to the one adopted in the
UCP of Martilli et al. (2002). The solar energy penetrating
through the windows is assumed to be uniformly distributed on the interior surfaces. Moreover, this radiation is
reflected by the surfaces isotropically in all the directions.
The solar radiation captured by an indoor wall is the sum
of the radiation coming directly from the windows and the
radiation reflected by the other indoor walls. This shortwave radiation reaching a wall is indicated by Eqs. 13–15
(by “wall”, here and in the rest of the article, we intend all the
internal surfaces, including ceiling and pavement). For
example, for the radiation reaching a wall i (more details on
the symbols used in these equations are in the Appendix):
X
Rsi ¼ Rs þ
albj Rsj y ji
ð13Þ
j6¼i
With a simple algebraic manipulation, the albedo of the
window can be written as
albwind ¼ 1 g0 þ
UCP, this formulation is used for the direct and reflected
radiation from the other surfaces of the urban canyons.
g0
2
Zp a a
b
c
x þ l xl þ g xg sin xdx
pa
p
p
ð12Þ
0
We now have a simple expression (Eq. 12) that depends
only on two parameters (p, q) to evaluate the quantity of
radiation transmitted through the windows when the radiation
is not direct. Using a numerical method it is easy to evaluate
this expression. In the simulations presented in this work, we
have used Eq. 12 to evaluate all the shortwave radiation that
penetrates the windows. Once the module is linked to the
y ji ¼
Aj fji
¼ fij
Ai
ð14Þ
albj ¼ albwall; j 1 awind; j þ albwind awind; j
ð15Þ
Equation 13 is a linear system of six equations and six
unknowns (the radiation received by each wall) easy to
solve by matrix inversion. The functions fji represent the
view factors between wall j and wall i, and the term Aj is
the area (m2) of wall j. More details about the view factors
can be found in Sparrow and Cess (1978).
2.4.2 Long-wave radiation
The long-wave radiation reaching an indoor wall i is the
sum of the long-wave radiation emitted and reflected by the
other walls. In order to compute the radiation, the following
equations are used (more details on the symbols used in
these equations are included in Appendix):
X
X
4
4
Rli ¼
sy ji ~"j Twall;j
þ "bj Twind;j
1 "j Rlj y ji
þ
j6¼i
j6¼i
ð16Þ
~" ¼ "
j
wall;j 1 awind;j
"bj ¼ "wind awind;j
"j ¼ ~"j þ "bj :
Fig. 1 Schematic representation of a beam incoming at a window
with an angle of incidence θ
ð17Þ
This is, once again, a linear system of six equations and
six unknowns easy to solve (the incoming long-wave and
68
A new building energy model coupled with an urban canopy parameterization
short-wave radiation at the outdoor surfaces are coming
from the mesoscale model).
335
(c) T* is smaller than the target temperature minus the
comfort range, i.e. T* <Ttarget −∆T. With a similar
procedure to that of the previous paragraph,
2.5 Mathematical model of the air conditioning system
In BEM the indoor air temperature and humidity can be
controlled with the help of the air conditioning system. We
can decide when the air conditioning is working and when
it is not. Kikegawa´s model (Kikegawa et al. 2003) is quite
different because it supposes that the processed load Hout
and Eout in Eqs. 3 and 4 are proportional to Hin and Ein,
respectively (Hout =φpHin and Eout =φpEin).
In our model the same method is used for the
computation of Hout and Eout, and, hence, in the following
only the computation of Hout will be explained. In the
model the air conditioning system (here and in the
following we use the term “air conditioning”, even though
heating can also be obtained with other systems) has a
target temperature Ttarget and a gap of comfort ∆T fixed that
the user can define.
If the air conditioning is not in use, then Hout =0. If it is,
a first guess of the temperature at time n+1, called T*, is
computed as follows (it is the discretization of Eq. 1 by
setting Hout =0),
T ¼
Δt n
H þ Tn
QB in
ð18Þ
At this point there are three possibilities:
(a) T* lies within the comfort range, i.e.T Ttarget ΔT , then Houtn =0, and T n+1=T*.
(b) T*is bigger than the target temperature plus the
comfort range, i.e. T > Ttarget þ ΔT . In this case
Houtn is calculated as:
n
Hout
¼
Hinn
QB Ttarget þ ΔT T n
Δt
ð19Þ
ð20Þ
Once Houtn is known, the temperature at time step n+1
is estimated as
T nþ1 ¼
Δt n
n
þ T n
Hin Hout
QB
n
n
Hout
¼ Hinn dQB ) Hinn Hout
¼ dQB > 0
Δt
ð22Þ
and the temperature at time n+1 is T nþ1 ¼ QB Hinn n
Þ þ T n.
Hout
With this method, the indoor temperature always lies
within a range of comfort (defined by the user), and the
cooling/heating power will never be higher than a fixed
value δ that depends on the properties of the air
conditioning system. The same treatment is done with
respect to the latent heat load Eout. Finally, we can calculate
the total processed load Hout +Eout.
3 Verification of BEM (without the coupling
with the UCP)
A combination of analytical, inter-program, and empirical
testing procedure has been used for the verification and
validation of the building energy model. The verification
and the inter-program validation were made comparing the
results of BEM against those obtained by Zmeureanu et al.
(1987) with the models CBS-MASS, BLAST (BLAST-3.0,
1981), and TARP (Walton 1983). The simulations with
BEM were done for a building with five floors. The results
refer to the third floor (intermediate floor).
3.1 Verification
n n
H H However, if outQB in > d(δ being the maximum power
of cooling/heating (Ks−1) of the air conditioning system,
which is a fixed value dependent on the air conditioning
n
system) Eq. 19 is not used and Hout
is calculated as:
n
n
Hout
¼ Hinn þ dQB ) Hinn Hout
¼ dQB < 0:
QB Ttarget ΔT T n
ð21Þ
Δt
n n
H H If outQB in > d, Eq. 21 is not used, and Houtn is
computed as:
n
Hout
¼ Hinn The verification is applied to a room 6.0×6.0×3.6 m3 on the
intermediate floor with four exterior walls and no windows.
The indoor air is considered dry and the main assumptions
are: no solar radiation, no sensible/latent heat generated by
equipments and occupants, and constant long-wave radiation incoming at the exterior walls. More details about the
inputs are presented in Table 1.
3.1.1 Variation of inside surface temperature of a wall due
to a step change in outdoor air temperature
Initially, the temperature of the walls and room air are
assumed to be 20°C. Then, while the room air temperature
is kept constant at 20°C (Hin =Hout), the outdoor air
temperature drops suddenly to 0°C (∆T0 =20°C). No air
69
336
Table 1 Physical parameters
used for the simulation in the
analytical validation
F. Salamanca et al.
Parameters
Settings
Exterior walls
Intermediate walls (ceilings and floors)
Ground wall (Dirichlet b. c. )
Constant surface wall coefficient (indoor and outdoor)
Volumetric ventilation rate
Physical properties used for brick
Conductivity
Density
Specific heat
Emissivity
0.28 m brick
0.28 m brick
0.28 m brick
8 WK–1 m–2
3.6 m3 m–2 h–1
infiltration (β=1, in Eq. 3) is considered in this case. The
temperature of the inside surface of the wall is analyzed and
the comparison shows that results from BEM are in good
agreement with analytical solutions and CBS-MASS
(Fig. 2).
0.73 WK–1 m–1
1.84×103 kg m–3
900 J kg–1 K–1
0.9
3.1.2 Variation of room air temperature for a step change
in outdoor air temperature
Initially, the temperature of the walls and room air are both
equal to 20°C. The variation of the room air temperature,
subject to a sudden drop of outdoor air temperature to 0°C
(∆T0 =20°C) is analyzed. The effect of air infiltration (β=0)
and internal mass is studied. We impose that the temperature of internal mass (ceilings and floors on the intermediate floors) is constant and equal to their respective room
air temperature (Tim =TR) in the building. The results
(Fig. 3) indicate good agreements between BEM, analytical
solutions, and CBS-MASS. It is interesting to note that the
evolution of the indoor air temperature is different on each
floor (Fig. 4). On the top floor, the cooling of indoor air is
faster than on the other floors because the roof is exposed to
the cold outdoor air. In contrast, on the first floor the
cooling is slower than on the other floors because we have
imposed a net flux equal to zero at the lowest layer in the
ground wall (Dirichlet boundary condition).
3.2 Inter-program validation
Fig. 2 Variation of the inside surface temperature for a 0.28 m deep
brick wall due to a step change in outdoor air temperature: a analytical
solution against CBS-MASS (From Zmeureanu et al. 1987); b
variation obtained with BEM
The inter-program validation deals with the comparison
between the estimation of the space thermal loads provided
by BEM against the predictions of three well-known
programs in the thermal analysis of buildings: BLAST,
TARP, and CBS-MASS. The comparison is performed in a
winter design day (Table 2) for an intermediate floor office
space 30 × 30 × 3.6 m3, with four exterior walls and
windows. The main characteristics used in this space are
presented in Table 3. It is important to point out that in our
simulation the solar radiation incoming at each intermediate
floor is the same. In this test BEM was not linked to the
UCP (Martilli et al. 2002) because, if it was, distinct floors
would receive different radiation fluxes due to shadowing
effects induced by neighboring buildings. For this reason
the results of an intermediate floor in the test are almost
70
A new building energy model coupled with an urban canopy parameterization
337
Table 2 Weather data for the inter-program validation in a winter
design day
Hour (h)
Fig. 3 Variation of the room air temperature due to step change in
outdoor air temperature (internal mass and air infiltration are considered):
a analytical solution against CBS-MASS (From Zmeureanu et al. 1987);
b variation obtained with BEM
independent of the height of the building. Thus, the largest
differences only occur between the top or the ground floor
and the intermediate floors. The top floor exchanges more
energy (solar radiation and heat conduction through its
Fig. 4 Variation of the room air temperature in different floors
obtained with BEM (as in Fig. 3)
Outdoor
temperature (ºC)
Direct normal
radiation (Wm–2)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
−18.05
−18.80
−19.40
−19.85
−20.00
−19.70
−18.95
−17.60
−15.65
−13.40
−10.85
−8.45
−6.65
−5.45
−5.00
−5.45
−6.50
−8.15
–
–
–
–
–
–
–
–
398.1
685.9
794.5
833.7
830.0
781.2
652.7
301.9
–
–
19
20
21
22
23
24
−10.10
−12.05
−13.70
−15.20
−16.40
−17.30
–
–
–
–
–
–
roof) than the other floors. On the other hand, on the first
floor the flux exchanged through the ground (generally in
contact with the soil) is different when compared against
the intermediate floors. Figure 5 shows the results of the
space thermal load (–Hout) necessary to maintain the indoor
air temperature constant (20ºC in this case) as estimated by
the BLAST, TARP, and CBS-MASS programs against
BEM. Dry air was considered (E in = Eout = 0) in the
simulation. The comparisons show that BEM provides
estimations of the heating load close to those of the other
models. The small differences observed may also be a
consequence of the uncertainties on the values of some
parameters (incoming long-wave radiation at the outdoor
surfaces and convective heat transfer coefficient at the
exterior wall) that were not explicitly mentioned in
Zmeureanu et al. (1987). Moreover, the radiation reflected
by the ground and other buildings, and incoming at the
exterior walls, is not considered in BEM and it is not
clear if it was taken into account by Zmeureanu et al.
(1987).
71
338
F. Salamanca et al.
Table 3 Physical parameters used for the simulation in the interprogram validation
Parameters
Settings
Exterior walls
0.10 m concrete
0.10 m insulator
0.02 m gypsum board
0.10 m concrete
0.10 m insulator
0.02 m gypsum board
0.22 m concrete
8 WK–1 m–2
Intermediate walls (ceilings and floors)
Ground wall (Dirichlet b. c.)
Constant surface wall coefficient
(indoor and outdoor)
Constant surface window coefficient
(indoor and outdoor)
Air infiltration
Volumetric ventilation rate
Glazing-to-wall ratio
(double standard glazing)
Internal heat gains
Room air temperature
Physical properties used for the materials
Emissivity
Concrete
Conductivity
Density
Specific heat
Albedo
Insulator
Conductivity
Density
Specific heat
Gypsum board
Conductivity
Density
Specific heat
Albedo
2.8 WK–1 m–2
β=0
3.6 m3 m–2 h
0.5
started on a Sunday, during the Pacific Ocean anticyclone
and under typical summer-day conditions. In our validation
of the air conditioning system we have compared the total
processed loads (Hout +Eout) for the building-O (the dimensions considered in each floor were47.0×47.0×3.78 m3)
obtained by BEM with the one generated by Kikegawa´s
model. Even though the calculation of the processed load is
rather different in the two models, the results are quite
similar (Fig. 8). Hence, it is possible to say that BEM is
capable of capturing the most important mechanisms
governing heat generation within buildings and exchanges
with the exterior. While comparing the two models, we
forced the exterior meteorological variables (temperature
and humidity) to the measured values. The aim of this test
was the validation of the air conditioning model, and not
30 Wm–2 between
the 9:00 to 17:00
20°C
0.9
1.73 WK–1 m–1
2.35×103 kg m–3
880 J kg–1 K–1
0.2
0.057 WK–1 m–1
13 kg m–3
840 J kg–1 K–1
0.14 WK–1 m–1
760 kg m–3
800 J kg–1 K–1
0.7
3.3 Comparison of BEM against other box model
to validate the air conditioning system
This last validation deals with the comparison between
BEM and the model of Kikegawa et al. (2003) using data
collected in a campaign over Tokyo (Japan). The measurement data were acquired in August, 1998 for a 25-story
office building (building-O) located at the center of a
business area (Ootemachi). The measurement includes the
continuous acquisition of meteorological data, Fig. 6, from
the rooftop of the building of approximately 100 m height.
The simulation was initialized at 0000 LST, 2 August and
terminated at 2400 LST, 5 August 1998. The simulations
Fig. 5 Comparison of the thermal load for an office space on a winter
design day in an intermediate floor: a estimations of CBS-MASS,
BLAST, and TARP (From Zmeureanu et al. 1987); b estimations
obtained with BEM
72
A new building energy model coupled with an urban canopy parameterization
339
Fig. 6 Temporal series of the
outdoor temperature (left) and
specific humidity (right) used in
the period of simulation and
measured at the top of the
building-O
the study of the interactions between BEM and the
atmosphere.
Temporal variations of φp and qE in Eq. 3 used in the
simulation are shown in Fig. 7. The parameters of typical
office buildings were adopted (Table 4) for the structures
and the air conditioning systems. In Fig. 8 one can see the
total (sensible and latent) cooling energy for the building-O
(obtained adding up over every level, 25 floors) computed
with BEM and Kikegawa´s model. Observe that the BEM’s
air conditioning model is able to reproduce results similar
to those obtained by Kikegawa.
4 Sensitivity of the processed load to different physical
processes
As explained, the air temperature and humidity inside a
building, and the energy needed to control them (through
air conditioning or heating), are influenced by several
physical processes (the different terms on the right-hand
sides in Eqs. 3 and 4). In this section, we present a series of
simulations with the goal of analyzing the sensitivity of the
processed load (e. g. the energy needed to control air
temperature and humidity) to these different processes.
The simulations were performed fixing Hout =Hin for
each floor. The following results (Fig. 9) are obtained by
adding the total load of the different floors. The same
building parameters and conditions used in Section 3.2
were considered. In the following, by “base case” simulation we will refer to the one described in Section 3.2.
The first test deals with the impact of the radiation
through the windows. Two simulations were performed, by
modifying the base case: one simulation without windows
(“no window” case), where no radiation is entering the
building (αwind,j =0, in Eq. 15), and the other one with
the internal albedo of the windows equal to 1, once the
Fig. 7 Time-dependent parameters φp (right) and qE (left) used
in weekdays
73
340
Table 4 Parameters used in the
validation of BEM against
Kikegawa´s model
F. Salamanca et al.
Parameters
Settings
Exterior vertical walls
0.11 m concrete
0.05 m insulator
0.11 m concrete
0.22 m concrete
0.33 m concrete
1.07 m soil
8 WK–1 m–2
2.8 WK–1 m–2
0900-1900 LSTa
26.0ºC
50.0%
5.0 m3 m–2 h–1
60%
5 m2/person
54.7 W/person
64.0 W/person
30%
30%
Intermediate walls (ceilings and floors)
Ground wall (Dirichlet b. c.)
a
Pre-cooling starts from 0800
LST
Constant surface wall coefficient (indoor and outdoor)
Constant surface window coefficient (indoor and outdoor)
Duration of air conditioning on weekdays
Target temperature of room cooling
Target relative humidity of room cooling
Volumetric ventilation rate per unit floor area
Thermal efficiency of the total heat exchanger (β)
Floor area per occupant
Sensible heat generation from an occupant (qhs)
Latent heat generation from an occupant (qhl)
Insolation transmittance through the windows (windows with blinds)
Glazing-to-wall ratio
Parameters own of BEM
Comfort range of temperature
Power of cooling/heating
Comfort range of humidity
Power of drying/moistening
Physical properties used for the materials
Emissivity
Concrete
Conductivity
Volumetric heat capacity
Albedo
Insulator
Conductivity
Volumetric heat capacity
Soil
Conductivity
Volumetric heat capacity
Fig. 8 Comparison of the total processed load in building-O obtained
with BEM and with the Kikegawa´s model
0.1 K
10–3 K s–1
10–3 kg kg–1
10–6 (kg kg–1)s–1
0.9
1.39 WK–1 m–1
1.93 106 J m–3 K–1
0.2
0.04 WK–1 m–1
0.06×106 J m–3 K–1
1.00 WK–1 m–1
1.74×106 J m–3 K–1
radiation is inside the floor (“total trapping” case). The aim
of this second case is to simulate the impact of the internal
walls in the building (rooms) that may prevent the
radiation from exiting windows, and trap the totality of
the solar radiation entering the building. The results
(Fig. 9a) show that in this winter case, the absence of
windows reduces the energy consumption during the night.
The reason is that the glasses have a higher heat capacity
than the walls and the heat flux exchanged with the air is
smaller than the flux exchanged through the walls. In
contrast, during the day, the absence of windows increases
the energy consumption because there is no radiation
penetrating the floor and it is more difficult to maintain a
warm indoor temperature. The effect due to considering
indoor walls in a floor (total trapping) is small and the
decision not to account for them is justified.
74
A new building energy model coupled with an urban canopy parameterization
341
Fig. 10 Comparison of the thermal load over various floors in an
office building when the incoming radiation is different for each floor
The second test (Fig. 9b) was performed to investigate
the impact of natural ventilation and the heat released by
people and equipments. One simulation was done without
people or equipments (“no people” case, φp =0 and qE =0 in
Eq. 3), and another with no ventilation (“no ventilation”
case, β=1 in Eq. 3). For this winter simulation, the impact
of people and equipments is of about 150 kW during
daytime (during night time people were absent also in the
base case). It is also interesting to notice the importance of
ventilation: for this winter case, during the night, the lack of
ventilation results in a decrease in energy consumption,
while during daytime, energy is required to cool down the
air in the building2 heated up by the radiation, and internal
sources (people and equipments).
The third test (Fig. 9c) was carried out to study the effect
of the convective heat coefficients at the external surfaces.3
Usually, these coefficients are estimated as a function of
wind speed. However, there is still a significant uncertainty
in the determination of such a relationship (see Martilli et
al. 2002, Masson 2000). One simulation was carried out
with a value smaller (3 WK−1 m−2) than in the base case
(8 WK−1 m−2) and another with a higher value (15 WK−1
m−2). As one can observe in the graph, the processed loads
are sensitive to these coefficients, with a maximum
variability of about 50 kW.
Fig. 9 Estimations of the total thermal load for the office building in
different situations: a comparison to study the impact of the radiation
through the windows; b comparison to study the impact of the natural
ventilation and the heat released by people and equipments; c
comparison to study the impact of the convective heat transfer
coefficients
2
It must be remembered here that the ventilation has an impact not
only on air temperature, but also on indoor air quality (for example, it
helps to disperse pollutants emitted indoor). The optimal ventilation
must then, takes into account both effects.
3
The convective heat coefficient h is used to estimate the sensible
heat H exchanged between the external wall surface and the
atmosphere, using the formula H=h (Ta–Twall), where Twall is the
temperature of the external surface of the wall, and Ta is the outdoor
air temperature. H enters in the surface energy budget at the external
surface and gives the b.c. for the heat diffusion equation in the wall.
75
342
Fig. 11 Comparison of the heat exchanged through natural ventilation
for the office building over different floors (no regulation of the indoor
air temperature)
Finally, a simulation with a different outdoor input
radiation for every floor was conducted. A five floor
building (15 m high, with floors 3 m high), in a street
15 m wide, (H/W=1) was considered to compute the
outdoor radiation. Using the parameterization of Martilli et
al. (2002), the solar radiation reaching the walls (accounting for shadowing and reflections) was computed for every
floor and for a N–S and W–E street orientation. Four
radiations (for North, South, West, and East walls) were
obtained for each floor. Such values were then used as an
input to BEM to compute the processed load in the same
conditions than in the base case. As one can see in Fig. 10,
when the incoming radiation fluxes are different in every
level, the load is different for every floor. In particular, the
fifth floor loses more energy during night time and needs
more energy to keep the temperature constant. On the other
F. Salamanca et al.
hand, during daytime, the upper floors receive more solar
radiation (less shadowing) than the lower floors and need
less energy to keep the temperature constant (for this winter
case).
The last test was performed with the same configuration,
but without any control over the indoor temperature Hout =
0. The temperature of the different floors could fluctuate
freely in response to the different forcings. This affects the
exchanges of heat between the indoor and outdoor air
through ventilation. In fact, as shown in Fig. 11, having
different temperatures on each floor, the heat fluxes due to
ventilation are also different.
These last two examples show that it is important to
consider the presence of floors in the building in the
estimation of energy consumption, as well as in the
calculation of the heat exchanged between the indoor and
the outdoor air.
5 First results about the impact of BEM in the UCP
In this last section, we present preliminary results about the
impact of BEM in the UCP of Martilli. The UCP-BEM
scheme has been coupled to the mesoscale model FVM
(Clappier et al. 1996) and simulations in a vertical column
(neglecting horizontal derivatives) have been carried out in
an ideal middle latitude city in a summer day. The ideal city
is composed of cubical buildings 15 m high (H/W=1). To
study the feedback between the air conditioning systems
and the atmosphere, two different simulations were carried
out using the same building parameters. In the first
simulation, the heat extracted by the air conditioning
systems is directly released into the atmosphere, while in
the second it is not (i.e., the feedback between the building
Fig. 12 Comparison of the difference in the energy consumption and the outdoor temperature
for 4 days of simulations. In the
first case the heat is released
directly into the atmosphere and
in the second one the heat is
released into a sewage or in the
soil
76
A new building energy model coupled with an urban canopy parameterization
with the air conditioning and the atmosphere is not taken
into account). Both simulations were carried out without
considering people or equipments ( 8 p =0 and qE =0 in
Eq. 3), with 30% of windows and without any natural
ventilation (β=1 in Eq. 3). Standard values of the air
conditioning system (the air conditioning was working
from 8:00 to 19:00 every day, and the target temperature
was 25ºC) and building materials were used. In Fig. 12, one
can observe the temporal evolution of the difference in the
air temperature on top of the buildings (where heat is
released) between the two simulations ∆T (°C) during a 4day period. Moreover, the temporal evolution of the
difference in energy consumption (W), as well as the total
daily variation in energy consumption ∆EC (kWh) (here
1 kWh=3.6×106 J) per building, have been computed. The
results show that when the air conditioning is working, the
heat released into the atmosphere can increase the outdoor
air temperature by 2 to 3°C. It is important to mention that
when the outdoor air temperature increases, the energy
necessary to maintain the indoor temperature within the
comfort range also increases. Even though the results are
not conclusive in these simulations (the atmospheric
heating may be overestimated because the horizontal
advection is not accounted for), one can clearly see that
the impact of the air conditioning systems on the urban
atmosphere is not negligible. Finally, the relative difference
(∆Ec/Ec) has been computed to evaluate the feedback
effects. In our 4-day simulation, the corresponding values
were 6.33%, 7.88%, 8.42%, and 9.53%, respectively. The
results indicate that an increase in air temperature of about
2°C corresponds to an increment in energy consumption of
approximately 7–8%.
In conclusion, this work is a first step towards a
modeling tool that can account for the complex interactions
between urban climate, air pollutant dispersion, and the
energy demand of buildings. Such a tool can be an
important support to urban planners.
Acknowledgements The authors wish to thank CIEMAT and LPASEPFL for the doctoral fellowships held by Francisco Salamanca and
Andrea Krpo, respectively. We also thank Y. Kikegawa for providing
important data for the validation. This work has been funded by the
Ministry of Environment of Spain.
Appendix
List of symbols
albwall,j
Af
Awall
i
Awind
j
Cp
hwall,i
hwind,j
l
Ta
Tr
Twall,i
Twind,j
P
qE
6 Conclusions
qhl
The verification indicates that BEM has accurately simulated
the basic heat transfer phenomenon. The inter-program
validation provides important information about the accuracy of BEM compared with other well-known computer
programs used in the thermal analysis of buildings. These
results show that BEM is able to capture the most important
mechanisms governing heat generation within buildings and
exchanges with the exterior. It is simpler and less CPU
expensive than other building energy models and can be
easily coupled with an UCP for mesoscale models. Moreover, BEM is able to reproduce the effects of the air
conditioning systems. Finally, the sensitivity test (Section 4)
shows the importance of considering different floors. A
more detailed validation of BEM in the UCP of Martilli is
carried out in Part II of this work (Salamanca and Martilli
2009), using meteorological data recorded during the
BUBBLE campaign over Basel (Switzerland).
343
qhs
qVa
qVr
Rlj
Rs
Rsj
Va
αwind,j
β
εwall,j
εwind
albedo of the indoor surface of the wall j
floor area (m2)
surface area of the wall i (m2)
surface area of window in the wall j (m2)
specific heat of air (J K−1 kg−1)
convective heat transfer coefficient between the
indoor air and the wall i (WK−1 m−2)
convective heat transfer coefficient between the
indoor air and the window in the wall
j (WK−1 m−2)
latent heat of evaporation (J kg−1)
outdoor air temperature (K)
indoor air temperature (K)
indoor surface temperature of the wall i (K)
temperature of the window in the wall j (K)
peak number of occupants per floor area
(person m−2)
sensible heat gain from equipments per floor
area (W m−2)
latent heat generation from the occupants
(W person−1)
sensible heat generation from the occupants
(W person−1)
specific humidity of the outdoor air (kg kg−1)
specific humidity of the indoor air (kg kg−1)
total long-wave radiation flux received by the
wall j (W m−2)
solar radiation energy crossing the windows
received directly by the indoor walls (W m−2)
total shortwave radiation flux received by the
wall j (W m−2)
total ventilation rate (m3 s−1)
% of window in the wall j
thermal efficiency of the total heat exchanger,
0b1
emissivity of the indoor surface of the wall j
emissivity of the windows
77
344
φP
ρ
σ
F. Salamanca et al.
ratio of hourly occupants to P, 0 8 p 1
air density (kg m−3)
Stefan-Boltzmann constant (W m−2 K−4)
References
BLAST-3.0-1981. The Building Loads Analysis and System Thermodynamics Program, Users Manual, U. S. Army Construction
Engineering Research Laboratory, Champaign, Illinois, March.
Clappier, A., Perrochet, P., Martilli, A., Muller, F. and Krueger, B. C.
1996. A new nonhydrostatic mesoscale model using a CVFE
(control volume finite element) discretisation technique, in P. M.
Borell et al (eds.), Proceedings, EUROTRAC Symposium ’96,
Computational Mechanics Publications, Southampton, pp. 527–531
Karlsson J, Roos A (2000) Modelling the angular behaviour of the total
solar energy transmittance of windows. Sol Energy 69:321–329
Kikegawa Y, Genchi Y, Yoshikado H, Kondo H (2003) Development
of a numerical simulation system toward comprehensive assessments of urban warming countermeasures including their impacts
upon the urban buildings energy-demands. Appl Energy 76:449–
466
Kusaka H, Kondo H, Kikegawa Y, Kimura F (2001) A simple singlelayer urban canopy model for atmospheric models: comparison
with multi-layer and slab models. Bound-Lay Meteorol 101:329–
358
Martilli A, Clappier A, Rotach MW (2002) An urban surface
exchange parameterization for mesoscale models. Bound-Lay
Meteorol 104:261–304
Masson V (2000) A physically based scheme for the urban energy
budget in atmospheric models. Bound-Lay Meteorol 94:357–397
Oke, T. R. 1987, The surface energy budget of urban areas, in
Modeling the Urban Boundary Layer, edited by the American
Meteorological Society, 1–52.
Roos A (1997) Optical characterisation of coated glazings at oblique
angles of incidence: measurements versus model calculations.
J Non-Cryst Solids 218:247–255
Salamanca F, Martilli A (2009) A new building energy model coupled
with an urban canopy parameterization for urban climate
simulations—Part II. Validation with one dimension off-line
simulations. Theor Appl, Climatol. doi:10.1007/s00704-0090143-8
Sparrow EM, Cess RD (1978) Radiation Heat Transfer. Brooks/Cole,
Belmont, p 366
Walton, G. N. 1983. Thermal Analysis Research Program (TARP)
Reference Manual, U. S. Department of Commerce, National
Bureau of Standards, National Engineering Laboratory, Washington, DC, March.
Zmeureanu R, Fazio P, Haghighat F (1987) Analytical and interprogam validation of a building thermal model. Energy Build
10:121–133
78
Capítulo 3
79
Capítulo 3
3.2 Un nuevo modelo energético de edificios acoplado a una parametrización urbana
para simulaciones del clima urbano-Parte-II. Validación con simulaciones (off-line) en
una dimensión vertical.
Salamanca, F., and A. Martilli, 2010: A New Building Energy Model coupled with an urban canopy
parameterization for urban climate simulations-part II. Validation with one dimension off-line
simulations. Theor. Appl. Climatol., 99, 345-356.
En esta sección se presentan los resultados de varias simulaciones en una dimensión vertical
(off-line) utilizando el modelo energético BEM acoplado a la parametrización urbana BEP.
Se calculan diferentes flujos (calor sensible/latente y radiación neta) y se comparan con
valores observados recogidos durante la campaña meteorológica BUBBLE (2002) que tuvo
lugar en la ciudad suiza de Basel. Gracias a que con el nuevo modelo energético se calcula la
evolución de la temperatura y humedad interiores, esta campaña es particularmente
interesante porque se tomaron medidas de la temperatura interior en algunos edificios. Los
resultados obtenidos muestran la importancia de los flujos resueltos en el modelo energético.
Las diferentes simulaciones muestran que el calor proveniente de los sistemas de aire
acondicionado tiene un importante impacto en la temperatura exterior y que debería tenerse en
cuenta en las simulaciones del clima urbano a mesoescala.
80
Theor Appl Climatol (2010) 99:345–356
DOI 10.1007/s00704-009-0143-8
ORIGINAL PAPER
A new Building Energy Model coupled with an Urban
Canopy Parameterization for urban climate
simulations—part II. Validation with one
dimension off-line simulations
Francisco Salamanca & Alberto Martilli
Received: 21 October 2008 / Accepted: 21 April 2009 / Published online: 13 May 2009
# Springer-Verlag 2009
Abstract Recent studies show that the fluxes exchanged
between buildings and the atmosphere play an important role
in the urban climate. These fluxes are taken into account in
mesoscale models considering new and more complex Urban
Canopy Parameterizations (UCP). A standard methodology to
test an UCP is to use one-dimensional (1D) off-line
simulations. In this contribution, an UCP with and without a
Building Energy Model (BEM) is run 1D off-line and the
results are compared against the experimental data obtained in
the BUBBLE measuring campaign over Basel (Switzerland)
in 2002. The advantage of BEM is that it computes the
evolution of the indoor building temperature as a function of
energy production and consumption in the building, the
radiation coming through the windows, and the fluxes of heat
exchanged through the walls and roofs as well as the impact of
the air conditioning system. This evaluation exercise is
particularly significant since, for the period simulated, indoor
temperatures were recorded. Different statistical parameters
have been calculated over the entire simulated episode in
order to compare the two versions of the UCP against
measurements. In conclusion, with this work, we want to
study the effect of BEM on the different turbulent fluxes and
exploit the new possibilities that the UCP–BEM offers us, like
F. Salamanca : A. Martilli
Research Centre for Energy,
Environment and Technology, CIEMAT,
Avenida Complutense 22,
28040 Madrid, Spain
F. Salamanca (*)
CIEMAT,
Edificio 03, P1.9, Avenida Complutense 22,
28040 Madrid, Spain
e-mail: [email protected]
the impact of the air conditioning systems and the evaluation
of their energy consumption.
1 Introduction
The Urban Canopy Parameterization (UCP) of Martilli et al.
(2002) simulates the impact of urban buildings on airflow in
mesoscale atmospheric models. This scheme takes into
account the impact of the urban surfaces on wind,
temperature, and turbulent kinetic energy (TKE), but does
not explicitly resolve the generation of energy within the
buildings and its transfer to the atmosphere. Since this effect
can significantly modify the urban energy budget, Salamanca
et al. (2009) developed a Building Energy Model (BEM) that
was implemented in the urban parameterization (UCP–BEM)
of Martilli. Thanks to this improvement, a detailed study of
the impact of cities on the urban climatology can be
conducted. However, this parameterization needs to be
validated against meteorological observations in order to
judge the reliability of the results and its predictions. In the
last years, several measurement campaigns have been carried
out to evaluate different urban schemes; see Masson et al.
(2002) and Best et al. (2006) for example. These necessary
campaigns help us to understand the physical processes that
take place in the urban atmosphere and to validate the
accuracy of our schemes. All the experience and knowledge
acquired with these studies could be applied to evaluate new
strategies of city development to minimize the intensity of
the Urban Heat Island phenomenon (UHI). Following this
idea, the goal of the present work is twofold: on the one
hand, the recent UCP–BEM scheme is evaluated against the
surface energy balance fluxes measured in the BUBBLE
(http://www.unibas.ch/geo/mcr/Projects/BUBBLE/) experi-
81
346
F. Salamanca, A. Martilli
ment and compared with the results obtained with the
previous version of UCP of Martilli; on the other hand, the
new possibilities that the UCP–BEM scheme offers are
evaluated. In particular, with this new scheme, the impact of
the air conditioning systems on the atmosphere can be
evaluated, and the energy consumption can be calculated for
different situations. The UCP–BEM parameterization has
been coupled to the mesoscale model (FVM, Clappier et al.
1996), and for this work, the 1D off-line version (see
Section 3.1 for details) is used.
In Section 2, a short description of the theoretical context
is explained. In Section 3, the 1D off-line configuration and
the fundamental characteristics of the simulated urban area
are described. In Section 4, results obtained with the two
schemes are compared. And finally, in Section 5, the heat
emission and the energy consumption of the air conditioning systems in different situations are analyzed. Conclusions and future research are given in Section 6.
within the buildings due to human presence). In the
model, the anthropogenic latent heat interacts with the
outdoor air through the natural ventilation of the buildings, while the sensible heat is exchanged with the
atmosphere through natural ventilation and by heat
diffusion through the walls. In the model, all these
anthropogenic heats (jointly with the heat generated by
the air conditioning systems) are added to the terms QLE
and QH of the Eq. (1). In the two urban schemes, the
different urban fluxes can be addressed independently, and
the heat stored in the urban fabric is calculated using the
energy balance equation,
2 Approach
3 Applied framework
The urban surface energy balance, defined by (Oke 1988),
plays a fundamental role in this work. Assuming no
horizontal advection, the relevant energetic exchange
processes can be written as:
3.1 The 1D off-line configuration
Qnet ¼ QH þ QLE þ Qstor
Qstor ¼ Qnet QH QLE
ð2Þ
A standard way to validate an urban parameterization is
to compute the different fluxes at different heights and to
compare them with the measurements. An example is
sketched in Fig. 1.
A peculiarity of the UCP used in this study is that it is
multilayered. To run it off-line, then the simulation is
ð1Þ
where the term Qnet is the net all wave radiation, QH the
sensible heat flux, QLE the latent heat flux, and finally Qstor
the net heat stored in the urban area. It is important to
mention that in an urban area the heat fluxes (QH and QLE)
are not only the result of the partitioning of the net
radiation, as it happens for natural surfaces, but they have
a component due to human activities (anthropogenic heat).
In the UCP of Martilli et al. (2002), this was not
considered1, while in the new UCP–BEM scheme, a
fraction of the total anthropogenic heat is taken into
account (the heat released by the traffic and industry are
not considered in the new UCP–BEM scheme). The latent
and sensible anthropogenic heat is considered in this new
UCP–BEM separately (more details in Salamanca et al.
2009), and includes all additional energy produced by
human activities within the buildings; the latent heat
generated by people, the sensible heat generated by
machines and people, and the heat generated by the air
conditioning systems (this last heat is injected directly
into the atmosphere, while the other fluxes are released
1
To be precise, in the UCP of Martilli, it is possible to fix the internal
building temperature, which accounts in some indirect and very rough
way for the anthropogenic heat. But this technique is not precise, and
does not allow any estimation of energy consumption.
Fig. 1 Schematic picture of the heat fluxes considered in an urban
environment
82
A new Building Energy Model coupled with an Urban Canopy — part II
performed on a vertical column (1D), ranging from street
level to the height of the forcing (32 m in this case), with a
vertical resolution of 2 m. The model calculates the vertical
profile of several variables (temperature, wind, humidity,
TKE) and turbulent fluxes from the forced altitude down to
ground. The model uses a k−l turbulence closure scheme,
hence the vertical turbulent fluxes w0 ξ0 (ξ stands for any
scalar variable) are computed using the K-theory, as,
w0 ξ0 ¼ K
@ξ
:
@z
ð3Þ
The computation of the turbulent transfer coefficients K
leads to the calculation of a prognostic equation for the
TKE (Bougeault and Lacarrère 1989), as it is explained in
Martilli et al. (2002). The only difference, compared to the
formulation presented in Martilli et al. (2002), is in the
estimation of the length scales. Since it is impossible to
estimate the values of lup because the whole planetary
boundary layer (PBL) is not resolved, it is assumed that the
relevant length scale is the height above ground.
The forcing was applied to air temperature, humidity,
pressure, wind components, long-wave downward radiation, and solar radiation.
In summary, the sensible and latent heat fluxes are
estimated at different heights from (details of the symbols
in Appendix)
QH ¼ ρCP w0 θ0 ¼ ρCP K
QLE ¼ ρLv w0 q0 ¼ ρLv K
@θ
@z
ð4Þ
@q
;
@z
ð5Þ
while the net radiation Qnet is a weighted average of the net
radiation at each surface (walls, roof, street, and vegetated
part).
It should not be forgotten that the energy balance (Eq. 1),
which represents a budget of the different fluxes representative of an urban zone, never describes local values.
3.2 Urban area characteristics and measurements
The data used in this work were collected at the main urban
surface site of the BUBBLE experiment (Sperrstrasse). It is
located in a heavily built-up part of the city of Basel,
Switzerland (European urban, mainly residential three- to
four-story buildings in blocks, flat commercial and light
industrial buildings in the backyards). The parameters of
the city surface are summarized in Table 1 (more details in
Christen 2005) and the thermal properties of the materials
in Table 2. The measurement setup consists of a tower (see
Fig. 2) inside a street canyon reaching up to 32 m
(approximately 2.2 times the mean roof height of the urban
347
Table 1 Morphometric parameters and surface characteristics of the
city surface of the city of Basel for a circle of 250 m around the tower
at Basel-Sperrstrasse
Mean building height
ZH
14.6m
Population density
ρinhab
Frontal aspect ratio
Plan aspect ratio of buildings
Plan aspect ratio of vegetation
Complete aspect ratioa
1F
1p
1V
1C
Between 200 and 300
inhabitants/ha
0.37
0.54
0.16
1.92
Plan area ratio of impervious
non-building surfaces
Roof materials
1I
0.30
–
Building materials
–
45% tiles, 50% gravel, 5%
corrugated iron
Plaster, concrete, brick
a
This term is the total surface of a building in contact with the outdoor
air divided by the area of a unit urban cell
surface). The intensive operation period (IOP) analyzed was
between June 10th and July 10th 2002. The overall
framework and the experimental activities during BUBBLE
are documented in Rotach et al. (2005). Measurements
were taken at different heights in and above the street
canyon with a 10-min average time resolution.
4 Results
During the IOP, the mean air temperature was 20°C and the
mean precipitation was 65 mm (more details in Christen
2005). The simulations were carried out during the overall
Table 2 Thermal properties of the building materials for roofs, walls,
and roads used in the simulations corresponding to the BaselSperrstrasse site
Roof layer
d
C
1
Road layer
d
C
1
Wall layer
d
C
1
1
2
3
4
0.02
1.128
0.614
0.02
0.276
0.129
0.02
0.382
0.090
0.04
1.745
0.984
0.010
1.940
0.750
0.040
1.940
0.750
0.025
1.550
0.934
0.975
1.350
0.275
0.01
1.778
1.070
0.03
1.780
1.076
0.08
1.764
1.071
0.02
1.779
0.651
Layer sequence: 1 is nearest to the surface. Here, d is the thickness of
layer (m), C is the heat capacity of the layer (MJ m−3 K−1 ), and 1 is
the thermal conductivity (W m−1 K−1 )
83
348
F. Salamanca, A. Martilli
Fig. 2 The street canyon from above (Sperrstrasse). The tower
reaches approximately 2.2 times the mean roof height of the urban
surface (photography obtained from the internet page of the BUBBLE
experiment)
IOP with four different setups. In the first one, only the
UCP of Martilli (ucp case) was considered. The second
simulation was carried out considering the UCP–BEM
scheme (ucp–bem case), but without considering the effects
of the air conditioning systems (they were turned off in the
model). Finally, the last two simulations were carried out
using the same UCP–BEM parameterization but with the
air conditioning systems running in two different ways; for
the first one, the air conditioning was working 24 h a day
(ucp–bemac case), while for the second one, air conditioning was working from 8:30 to 18:30 h every day (ucp–
Table 3 Input parameters and
variables considered in the four
different simulations that were
carried out
a
In the Urban Canopy Model,
the roughness of the roofs and
the streets is taken into account
to compute the exchange of the
heat fluxes between the surfaces
and the atmosphere
b
Typical air conditioning
systems for office buildings
have values of COP between 2
and 5; see Ashie et al. (1999)
bemac* case). A mathematical description of the modeling
of the air conditioning systems is explained in Salamanca et
al. (2009). During the IOP, sensors of temperature were
installed in the stairwells of some buildings. These
measurements suggest that the target temperature of the
air conditioning systems was close to 24°C (Voogt,
personal communication). A detailed summary of the
setting used in the four simulations is described in Table 3.
The heat production by an adult when he is resting is
about 70 W, when working (office work) is about 110 W,
and when occupied (walking, driving, domestic work, etc.)
is about 300 W (Oke 1987). Averaging these quantities and
assuming 8 h daily for each activity gives a heat production
per person of 160 W. Furthermore, the water lost by
evaporation during a day by an adult is about 0.8 kg. This
quantity corresponds to approximately 22.7 W of latent
heat per person. Using these values and the fact that in
Basel-Sperrstrasse the population density is about
250 inhabitants/ha (Christen 2005), it is possible to
estimate the total sensible and latent human heat generated
in this zone. In the simulations, a constant ratio of
occupants of 0.0116 persons/m2 of floor and a sensible
heat flux from equipment of 7.4 W/m2 of floor were
considered. The values of sensible heat flux from equipment were chosen to be coherent with the estimation of an
anthropogenic heat emission of 20 W/m2 of land for the
Basel-Sperrstrasse site (Christen 2005).
4.1 Sensible heat
Two periods are analyzed in this section. The first period
goes from 14th of June 2002 (165 Julian day) to 23 rd of
June 2002 (174 Julian day), both inclusive, and the second
one from 30th of June (181 Julian day) to 14th of July 2002
(195 Julian day). Most of the days in the first period were
sunny, and in some days the temperature reached up to
35°C. In the second period, cloudy skies and lower
temperatures were more frequent. In Fig. 3, one can see
the results obtained for the sensible heat flux (w0 θ0
Case
ucp
ucp–bem
ucp–bemac
ucp–bemac*
Z0 (m)a
Indoor surface wall temperature fixed
Natural ventilation
Number of floors considered in a building
Coefficient of performance
(COP)b
Target temperature of room cooling
Comfort range of temperature
Sensible heat generated by a person
Latent heat generated by a person
0.0005
20°C
No
–
0.0005
Not fixed
Yes
4
0.0005
Not fixed
No
4
0.0005
Not fixed
No
4
–
–
–
–
–
–
–
–
160 W
22.7 W
3.5
23.5°C
±0.5°C
160 W
22.7 W
3.5
23.5°C
±0.5°C
160 W
22.7 W
84
A new Building Energy Model coupled with an Urban Canopy — part II
kinematic heat flux) calculated in the four different cases
against the measurements at different heights (to facilitate
the clarity in the plots and to avoid noise induced by the
intermittent presence of clouds, only three selected days for
the first period and two selected days for the second are
shown, and hourly mean values are used). Above the roofs,
the ucp–bem’s cases show better fits than the ucp case
(Fig. 3c–d). In the IOP, the air conditioning systems were
349
working (this can be deduced from the indoor air
temperature measurements showing a little variation of
temperature during the day) and these systems produce an
increase of sensible heat fluxes into the atmosphere. To
quantify the differences between the simulation and the
measurements, we computed the root mean square error
(RMSE) for the sensible heat flux QH for the four cases
during the two periods of simulation (hourly mean values
Fig. 3 Sensible heat fluxes obtained with the two parameterizations (UCP and UCP–BEM) in four different situations against the measurements
for the two periods analyzed: a–b at 32 m, c–d at 18 m, and e–f at 4 m from the ground
85
350
Table 4 Performance statistics
(RMSE) for sensible heat fluxes
(W/m2) at different heights for
the four cases simulated at the
Basel-Sperrstrasse site (first
period)
F. Salamanca, A. Martilli
Case
QH
QH
QH
QH
QH
QH
QH
QH
QH
(32 m)
(night-time) (32 m)
(daytime) (32 m)
(18 m)
(night-time) (18 m)
(daytime) (18 m)
(4 m)
(night-time) (4 m)
(daytime) (4 m)
were considered). Moreover, night-time and daytime values
were also calculated. A value was considered a night-time
value when the observed net radiation was negative.
Otherwise, the value was considered a daytime value. Here,
the RMSE is defined as:
"
#1
N 2 2
1 X
RMSE ¼
Vj V0
ð6Þ
N j¼1
where Vj and V0 are simulated and observed values,
respectively.
4.1.1 First period
RMSE results (see Table 4 and Fig. 3a) at 32 m show that
the inclusion of BEM improves the results compared to the
standard ucp. For the first period, the best result is obtained
in the ucp–bem (no air conditioning) case when entire days
(day and night-time) are considered. During the night, the
best fit (RMSE night-time) is obtained when the air
conditioning is used only during daytime (ucp–bemac*
case), while for the ucp–bem simulation (no air condition-
Table 5 Performance statistics
(RMSE) for sensible heat fluxes
(W/m2) at different heights for
the four cases simulated at the
Basel-Sperrstrasse site (second
period)
Case
QH
QH
QH
QH
QH
QH
QH
QH
QH
(32 m)
(night-time) (32 m)
(daytime) (32 m)
(18 m)
(night-time) (18 m)
(daytime) (18 m)
(4 m)
(night-time) (4 m)
(daytime) (4 m)
Days
165–174
ucp–bemac
ucp–bemac*
ucp–bem
ucp
59.71
46.69
68.61
53.30
35.75
64.28
30.94
16.85
38.90
55.66
16.42
73.71
53.62
37.38
64.00
30.80
16.21
38.92
39.20
23.24
48.58
56.62
29.19
71.76
30.78
15.55
39.12
46.35
20.77
59.71
77.94
51.85
94.19
31.53
17.32
39.58
ing), we observed the best RMSE during daytime. In fact,
during the day, when the air conditioning is in use, the
sensible heat is slightly overestimated at this height. As
indicated in Fig. 3c, larger differences are observed 18 m
above the ground (the mean building height was 14.6 m).
Here, the best RMSE result (considering all the days) is
obtained in the ucp–bemac case, the second better in the
ucp–bemac*, the third in the ucp–bem, and the worst fit is
generated by the ucp simulation.
During daytime, the best fit is obtained with the ucp–
bemac* scheme and during night-time with the ucp–bem.
Figure 3e indicates that near the ground the effect of the air
conditioning systems is negligible. The RMSE parameters
confirm this hypothesis (there are no important differences
between the four cases simulated). In fact, in the model, the
heat is released into the atmosphere by an air conditioning
system located on the roof of the buildings, which might
explain the small difference obtained near the ground
within the urban canopy. It could be interesting to study
the effect of air conditioning systems located at different
heights on the facade of buildings. In fact, in that case, heat
would be directly released within the urban canyons.
Days
181–195
ucp–bemac
ucp–bemac*
ucp–bem
ucp
44.63
37.43
49.92
37.65
25.88
45.28
27.41
15.62
34.36
39.64
21.84
49.97
34.99
23.96
42.12
27.33
15.32
34.36
35.35
25.80
41.75
41.15
19.56
52.94
27.15
14.91
34.24
36.09
14.73
47.12
54.17
30.77
67.94
27.87
16.34
34.75
86
A new Building Energy Model coupled with an Urban Canopy — part II
351
4.1.2 Second period
For this period (see Table 5 and Fig. 3b), the best fit at 32 m
is obtained again in the ucp–bem case when complete days
are considered. Values close to ucp–bem are generated by
the ucp and ucp–bemac* cases. Unexpectedly, the best
night-time value is obtained with the ucp scheme, and
finally, during daytime the lower RMSE is computed with
the ucp–bem (no air conditioning) case. At 18 m (Fig. 3d),
the best results are obtained when the air conditioning
systems are in use (ucp–bemac and ucp–bemac* cases),
with a large difference between the new schemes and the
traditional ucp. During night-time, the best fit is obtained in
the ucp–bem case, followed closely by the ucp–bemac*
one. During daytime, the best adjustment is generated with
the ucp–bemac* scheme. Finally, at 4 m above the ground
(Fig. 4f), there are no important differences between the
four cases simulated.
In the above comparison between modeled and measured sensible heat fluxes, the following points must be
taken into account:
&
There is a general tendency for the ucp–bemac and
ucp–bemac* simulations to have worse RMSE than ucp
at 32 m, but better at 18 m. Since there are no sources
or sinks of energy between 18 and 32 m in the model,
the computed values are similar at the two heights. The
difference should therefore derive from the measurements. To confirm this hypothesis, the difference
between the sensible heat fluxes measured at 32 and
18 m have been plotted (Fig. 4). As one can see,
differences in the measured values between the two
heights are around 50–100 W/m2, and may be due to
some horizontal advection effect. Since in the model
horizontal advection is not taken into account, we think
&
&
&
that the 18 m measurements are more significant for the
validation of the model.
The ucp (old version) maintains the internal temperature inside the building constant, but does not take
into account the energy consumption needed to keep
it constant. Therefore, this model cannot reproduce
the complete impact generated by anthropogenic
heating. The fluxes computed by ucp, then, do not
result from a complete representation of the physics
of the system.
The ucp–bem without air conditioning does not control
the temperature inside the building. The variation
of the internal temperature modeled by ucp–bem is
much higher than the measured variation (close to 1°C
around 24°C). So, even if the RMSE of the sensible
heat flux at 18 m is comparable to those computed by
ucp–bemac and ucp–bemac*, the sensible heat fluxes
computed by ucp–bem are not a complete representation of the physics of the system.
It is interesting to observe (Fig. 3) that, at 18 m during
night-time, ucp has the lowest sensible heat flux, close
to zero, while ucp–bem, ucp–bemac, and ucp–bemac*
all have a clear positive sensible heat flux, in agreement
with the measurements. The nocturnal positive sensible
heat flux is a crucial feature to model the nocturnal
Urban Heat Island.
In conclusion, it is possible to say that the fact of
considering the generation of heat within the buildings, and
in particular the effect of the air conditioning, improves the
estimation of the sensible heat fluxes in the city, not only
because the statistical parameters are better than those of
the old ucp but also because the physics of the system is
better represented. This is a very important point since it
increases the confidence in the predictive capability of the
model.
4.2 Net radiation
Fig. 4 Sensible heat flux at 32 m minus sensible heat flux at 18 m of
height. The lower axis corresponds to the first period and the upper
axis to the second period
The differences between the four simulations for the net
radiation (Fig. 5 and Table 6) are small when the complete
days are considered. However, for the first period, it is
interesting to observe that during night-time the best results
are obtained with the ucp–bem schemes, and the simulation
that matches the measurements most closely is the ucp–
bemac*. During daytime, the best fit is obtained in the ucp
case in the two periods followed closely by the ucp–bemac
schemes.
The underestimation of the net radiation during daytime
for the two periods (Fig. 5) could be a consequence of
overestimating the upward long-wave radiation in the four
schemes which, in turn, might be caused by an overestimation of the roof surface temperature. The roof surface
87
352
F. Salamanca, A. Martilli
Fig. 5 Net radiation for the four different schemes against measurements: a 3 days are shown for the first period, and b two selected days are
shown for the second period
temperature is very sensitive to the roof’s roughness length,
a parameter for which there is little information.
4.3 Latent heat
The RMSE (Table 6) for the latent heat shows that in the
two periods there are no significant differences between the
four schemes when complete days are considered. It was
not necessary to evaluate the daytime and night-time RMSE
parameters because the plots (not shown) do not reveal
significant differences.
ejected into the atmosphere by an air conditioning system is
calculated as (more details about the symbols in Appendix):
COP þ 1
ðHsout þ Hlout Þ:
ð7Þ
$Hs ¼
COP
If we sum the heat fluxes of all the buildings in the grid cell
and divide by the corresponding area, we obtain the heat flux
ejected into the atmosphere per unit of land area. Ten
simulations are presented in this section, which were carried
out considering only the cases with air conditioning systems:
&
5 Waste heat emission and energy consumption
In this last section, the waste heat emission of the air
conditioning systems and the energy consumption are
evaluated. The sensible heat ∆Hs (W) (in the QH term)
Table 6 Performance statistics (RMSE) for different heat fluxes
(W/m2) at 32 m of height for the four cases simulated at the BaselSperrstrasse site
Case
ucp–bemac
Days 165–174 (first period)
36.95
Qnet
Qnet
7.02
(night-time)
49.51
Qnet (daytime)
QLE
25.29
Days 181–195 (second period)
Qnet
36.00
Qnet
12.22
(night-time)
47.61
Qnet (daytime)
QLE
23.67
ucp–bemac*
ucp–bem
ucp
37.79
5.09
39.94
6.62
34.42
15.89
50.85
25.26
53.63
25.46
44.21
25.35
36.26
12.73
37.47
14.92
33.32
10.71
47.85
23.61
49.03
23.82
44.19
23.82
&
&
To evaluate the sensitivity of the model to the target
temperature imposed inside the buildings, we carried
out two simulations with the target temperature decreased by 1°C (bemac −1in and bemac* −1in cases),
and two with the target temperature increased by 1°C
(bemac +1in and bemac* +1in).
To study the impact of the air conditioning system on
the outdoor temperature, and consequently the
corresponding increase in energy consumption, two
simulations were carried out increasing the outdoor
temperature by 1°C (bemac +1out and bemac* +1out
cases) at the forcing height.
Finally, two more simulations (bemac-insulating and
bemac*-insulating) were considered by increasing the thickness of the insulating material (thermal conductivity 1=
0.09 W/m K) at the roof of the buildings from 2 to 6 cm.
All these simulations were performed for the two entire
periods (from 165 to 174 Julian days for the first period and
from 181 to 195 Julian days for the second period). In
Figs. 6 and 7, one can see the heat ejected into the
atmosphere (only the results for the three above selected
days are plotted for the first period and the two selected
days for the second). The time average hΔHs i during
10 days (first period) and during 15 days (second period)
(10-min time resolution) gives, in the bemac cases (see
Tables 7 and 8), heat fluxes near to 100 (W/m2) and to 50
88
A new Building Energy Model coupled with an Urban Canopy — part II
353
Fig. 6 Sensible heat ejected into the atmosphere by the air conditioning systems corresponding to the first period (only 3 days are shown): a
bemac cases and b bemac* cases
(W/m2) of land, respectively. On the other hand, we
obtained close to 160 (W/m2) for the first period and to
90 (W/m2) of land for the second in the bemac* schemes
(for the bemac* cases, the average was computed considering only the working time by day). Observe that the waste
heat released into the atmosphere when the target temperature is lowered by 1°C is higher compared to when the
outdoor temperature is increased by 1°C (similar differences are observed in the two periods). On the other hand,
the heat ejected into the air is decreased considerably when
the thickness of the insulating material is increased, and an
energy saving of 7–11% was observed.
The cooling energy consumption EC (W) for an air
conditioning system and the total consumption ΔEC (J) for
a period of time can be calculated as:
EC ¼
1
ðHsout þ Hlout Þ;
COP
ð8Þ
and
Z
$EC ¼ EC dt:
ð9Þ
Period
of
simulation
It is also interesting to estimate the total consumption
between the different simulations. One can see in Tables 7 and
8 the results for the total consumption by square kilometer of
city and by day (here 1 kWh=3.6×106 J). The saving in
energy consumption due to an increased thickness of the
insulation material approaches 7–11% for both periods. In
contrast, when the target temperature was decreased by 1°C,
we observed an increase (negative values in Tables 7 and
8 indicate an increase in energy consumption) in the
consumption of nearly 9% for the first period and 14% for
the second. Eventually, the consumption increased by 3% in
Fig. 7 Sensible heat ejected into the atmosphere by the air conditioning systems corresponding to the second period (only 2 days are shown): a
bemac cases and b bemac* cases
89
92.63
4.95
–
bemac
84.20
4.51
8.87
bemac
+1in
86.40
4.61
6.91
bemacinsulating
95.76
5.11
−3.34
bemac
+1out
101.17
5.40
−9.12
bemac
−1in
161.00
3.69
–
bemac*
147.00
3.38
8.45
bemac*
+1in
144.19
3.31
10.19
bemac*insulating
165.00
3.80
−3.08
bemac*
+1out
Days 181–195
h$Hs i (W/m2) of – land
ΔEC (105 kWh/km2day) of – land
Saving energetic (%)
h$Hs i (W/m2) of – land
ΔEC (105 kWh/km2 day) of – land
Saving energetic (%)
Case
50.18
2.81
–
bemac
42.92
2.47
11.88
bemac
+1in
46.16
2.51
10.64
bemacinsulating
53.08
2.95
−4.95
bemac
+1out
57.79
3.17
−13.03
bemac
−1in
89.60
2.06
–
bemac*
76.20
1.76
14.76
bemac*
+1in
79.97
1.84
10.76
bemac*insulating
94.80
2.18
−5.70
bemac*
+1out
Table 8 Results of the energy consumption and the heat ejected into the atmosphere corresponding to the ten cases simulated at the Basel-Sperrstrasse site (second period)
Days 165–174
h$Hs i (W/m2) of – land
ΔEC (105 kWh/km2 day) of – land
Saving energetic (%)
h$Hs i (W/m2) of – land
ΔEC (105 kWh/km2 day) of – land
Saving energetic (%)
Case
Table 7 Results of the energy consumption and the heat ejected into the atmosphere corresponding to the ten cases simulated at the Basel-Sperrstrasse site (first period)
103.15
2.37
−15.00
bemac*
−1in
173.94
3.99
−8.34
bemac*
−1in
354
F. Salamanca, A. Martilli
90
A new Building Energy Model coupled with an Urban Canopy — part II
the first period, and by almost 5% in the second, when the
outdoor temperature was increased by 1°C.
6 Conclusions
In this work, an urban canopy parameterization (ucp–bem
(ac)) coupled with a building energy model has been
compared with its counterpart without the building energy
model (ucp) and evaluated against measurements obtained
in the BUBBLE campaign. This work shows that the new
scheme ucp–bem(ac) is able to reproduce satisfactorily the
urban fluxes, and that it reproduces the physics of the
system better than ucp. Since phenomena like the Urban
Heat Island, and in general the structure of the Urban
Boundary Layer, are dependent on the urban fluxes, it is
expected that the inclusion of this scheme in a mesoscale
model will improve the capability of that model to
reproduce these phenomena.
Moreover, and most important, the scheme is able to
compute the heat ejected into the atmosphere by the air
conditioning systems and in general the energy consumption
linked to meteorological variables. Although further tests in
2D and 3D are needed, it is possible to say that the impact of
the air conditioning systems is not negligible and should be
taken into account in the mesoscale models to determine the
outdoor temperature in big cities in summer conditions. The
heat flux due to air conditioning, in fact, can be between 50
and 160 W/m2 in average (with peaks of up to 250 W/m2
in the hottest hours of the day) depending on meteorological conditions and time of use of the system.
Furthermore, the scheme has been used to test the
sensitivity of the energy consumption to different parameters. An increase in the thickness of the insulation materials
could reduce the consumption by about 10%. On the other
hand, the reduction in the target temperature by 1°C
increases the consumption by nearly 10–15%. Finally, an
increase in outdoor temperature by 1°C increases the
consumption by 3% to 5%. This relationship between air
temperature and energy consumption (similar to what was
found by Kikegawa et al. 2003, by analyzing the correlation
between data on energy consumption and measured air
temperature for Tokyo) highlights the importance of using a
coupled system. In fact, the feedbacks between the
following three points must be considered:
&
&
&
the air temperature in a city depends on the sensible
heat fluxes released into the atmosphere,
part of the sensible heat fluxes depend on the energy
consumption,
energy consumption depends on the air temperature.
Due to these feedbacks, the estimation of the impact of a
change on the target temperature, or the insulation,
355
mentioned above, may also be underestimated. The
inclusion of ucp–bem(ac) in a mesoscale model will allow
to account for all these feedbacks.
It is also interesting to mention that a variation in
sensible heat fluxes of the order of those estimated in this
work due to the air conditioning (50–160 W/m2) may have
a significant impact on the pollutant dispersion and also on
cloud formation. Potentially a further feedback can exist,
since short- and long-wave radiations are affected by
aerosols and clouds. Again, having the scheme implemented
in a mesoscale model will allow us to account for this impact.
Although these last considerations are not conclusive
(more realistic simulations are needed), we can say that the
new scheme is able to estimate the urban fluxes, it is a good
tool to test new energy consumption reduction strategies,
and it can help to better understand the Urban Heat Island
phenomenon in big cities.
Acknowledgements We are particularly grateful to Andrea Krpo of
the EPFL for the implementation of BEM in the UCP. The authors
wish to thank CIEMAT for the doctoral fellowships held by Francisco
Salamanca. We also thank Andreas Christen of the University of
British Columbia for the important explanations about the input data
used in the simulations. Moreover, we want to thank James Voogt of
the University of Western Ontario who provided us the data about the
indoor air temperature in some buildings obtained during the
BUBBLE campaign, and finally Scott Krayenhoff of the University
of British Columbia who sent us the thermal parameters corresponding
to Basel-Sperrstrasse. This work has been funded by the Ministry of
Environment of Spain.
Appendix
List of symbols
CP (Jkg−1
K−1)
COP
Hsout (W)
Hlout (W)
Lv (Jkg−1)
q (kgkg−1)
w (ms−1)
θ (K)
ρ (kgm−3)
specific heat of the air at constant pressure
energy efficiency (coefficient of
performance)
sensible heat pumped out for cooling per
building
latent heat pumped out per building
latent heat of vaporization
specific humidity
vertical component of the wind speed
potential temperature
density of the air
References
Ashie Y, Thanh Ca V, Takashi A (1999) Building canopy model for
the analysis of urban climate. J. Wind Eng. Ind. Aerodyn.
81:237–248
91
356
Best MJ, Grimmond CSB, Villani MG (2006) Evaluation of the urban
tile in MOSES using surface energy balance observations.
Boundary-Layer Meteorology 118:503–525
Bougeault P, Lacarrère P (1989) Parameterization of orographyinduced turbulence in a mesobeta-scale model. Mon. Wea. Rev.
117:1872–1890
Christen, A. 2005. Atmospheric turbulence and surface energy
exchange in urban environments. Results from the Basel Urban
Boundary Layer Experiment (BUBBLE)—(gleichzeitig Diss.
Phil.-Nat.-Fak. Univ. Basel 2005)—ISBN 3-85977-266-X
Clappier A, Perrochet P, Martilli A, Muller F, Krueger BC (1996) A
new nonhydrostatic mesoscale model using a CVFE (Control
Volume Finite Element) discretisation technique. In: Borell PM
et al. (eds) Proceedings, EUROTRAC Symposium ’96,
Computational Mechanics Publications, Southampton, pp.
527–531
Kikegawa Y, Genchi Y, Yoshikado H, Kondo H (2003) Development
of a numerical simulation system toward comprehensive assessments of urban warming countermeasures including their impacts
upon the urban buildings energy-demands. Appl Energy 76:449–
466
F. Salamanca, A. Martilli
Martilli A, Clappier A, Rotach MW (2002) An urban surface
exchange parameterization for mesoscale models. BoundaryLayer Meteorology 104:261–304
Masson V, Grimmond CSB, Oke TR (2002) Evaluation of the Town
Energy Balance (TEB) scheme with direct measurements from
dry districts in two cities. J. Appl. Meteorol. 41:1011–1026
Oke TR (1987) Boundary layer climates, 2nd edn. Methuen, London,
p 435
Oke TR (1988) The urban energy balance. Prog. Phys. Geogr. 12:471–
508
Rotach MW, Vogt R, Bernhofer C, Batchvarova E, Christen A,
Clappier A, Feddersen B, Gryning SE, Martucci G, Mayer H,
Mitev V, Oke TR, Parlow E, Richner H, Roth M, Roulet YA,
Ruffieux D, Salmond J, Schatzmann M, Voggt J (2005)
BUBBLE—an urban boundary layer meteorology project. Theor
Appl Climatol 81:231–261
Salamanca F, Krpo A, Martilli A, Clappier A (2009) A new
Building Energy Model coupled with an Urban Canopy
Parameterization for urban climate simulations—Part I. Formulation, verification and a sensitive analysis of the model. Theor
Appl Climatol doi:10.1007/s00704-009-0142-9
92
Capítulo 3
93
Capítulo 3
3.3 Derivación de las propiedades térmicas de un material representativo de un área
heterogénea de la ciudad.
Salamanca, F., E. S. Krayenhoff, and A. Martilli, 2009: On the Derivation of Material Thermal Properties
Representative of Heterogeneous Urban Neighbourhoods. Journal of Applied Meteorology and
Climatology, 48, 1725-1732.
En esta sección se analiza el cálculo de calor sensible intercambiado por las superficies de
una zona urbana con la atmósfera. A la hora de calcular el flujo de calor sensible para un
determinado tipo de superficies (horizontales o verticales) sólo podemos utilizar las
propiedades térmicas de un determinado material que represente un punto de la rejilla
numérica. En el trabajo presentado en esta sección analizamos dos formas estándar de calcular
las propiedades térmicas del material representativo de una zona urbana y proponemos una
nueva que mejora notablemente el cálculo del flujo de calor sensible intercambiado con la
atmósfera. La idea es derivar las propiedades térmicas del material representativo de la zona
suponiendo que el flujo de calor sensible calculado con el nuevo “material” sea igual a la
suma de los calores sensibles intercambiados con la atmósfera por cada uno de los distintos
materiales presentes.
94
AUGUST 2009
NOTES AND CORRESPONDENCE
1725
On the Derivation of Material Thermal Properties Representative of
Heterogeneous Urban Neighborhoods
F. SALAMANCA
Research Centre for Energy, Environment and Technology (CIEMAT), Madrid, Spain
E. S. KRAYENHOFF
Department of Geography, University of British Columbia, Vancouver, British Columbia, Canada
A. MARTILLI
Research Centre for Energy, Environment and Technology (CIEMAT), Madrid, Spain
(Manuscript received 18 December 2008, in final form 11 March 2009)
ABSTRACT
An important question arises when modeling a heterogeneous landscape (e.g., an urbanized area) with a
mesoscale atmospheric model. The surface within a grid cell of the model (which has a typical dimension of
one or more kilometers) can be composed of patches of surfaces of different character. The total sensible heat
flux in the grid cell, then, is the aggregate of the heat fluxes from each individual surface, each one with a
unique thermal response arising from its thermal properties, among other factors. Current methods to estimate the sensible heat flux consider only one (in the case of flat terrain) or three (roof, walls, and ground, for
urban areas) active surfaces with thermal properties that are ideally representative of the materials present in
the grid cell. The question is then how to choose the representative thermal properties such that the heat flux
computed by the model most closely approximates the aggregate of the fluxes from the different patches. In
this work a new way to average building material thermal properties for urban canopy parameterizations is
presented, and a suite of idealized numerical simulations demonstrates its superiority to two more standard
averages. Moreover, this novel approach points to a new way of determining physical properties that are
representative of heterogeneous zones.
1. Introduction
In recent decades, the number of urban meteorological and dispersion modeling publications has grown
rapidly as a result of increased computational speed.
These studies help us to understand better the atmospheric and environmental effects of urbanization (urban heat islands, pollution, etc.), and they aid in the
search for mitigation strategies. However, contemporary
increases in computational speed are no panacea [see a
recent review of urban modeling by Martilli (2007)]. A
mesoscale model needs a horizontal domain of tens/
hundreds of kilometers to simulate mesoscale circula-
Corresponding author address: F. Salamanca, Research Centre
for Energy, Environment and Technology (CIEMAT), Avenida
Complutense 22, 28040 Madrid, Spain.
E-mail: [email protected]
DOI: 10.1175/2009JAMC2176.1
Ó 2009 American Meteorological Society
tions. Yet, for computational reasons the resolution
cannot be better than a few kilometers, and much of the
heterogeneity in the urban landscape cannot be explicitly represented in the model.
Several urban parameterizations (e.g., Masson 2000;
Kusaka et al. 2001; Martilli et al. 2002; Kanda et al. 2005)
have been developed to communicate the mean thermal
and dynamic effects of the city to the mesoscale model.
Urban parameterizations have notably improved numerical results as they have developed, though their
complexity and their computational demands have also
increased. A key problem is the assignment or derivation of averaged physical properties that optimally
represent or define each local-scale urban neighborhood, ‘‘urban climate zone’’ (Oke 2006), or urban model
grid cell. These averaged physical properties are very
important because they feed the urban parameterization; as a consequence, they are directly responsible for
95
1726
JOURNAL OF APPLIED METEOROLOGY AND CLIMATOLOGY
the quality of the numerical results. An important issue,
then, is how we can obtain averaged material properties
that represent the same physical interaction with the
atmosphere as a given heterogeneous ensemble of materials in a city (note that we use the word ‘‘average’’
throughout this paper to mean ‘‘weighted’’ or ‘‘geometrically combined’’). In this work, we will focus on the
modeled accuracy of the sensible heat flux, because it is
the most relevant interaction between (dry) urban surfaces and the atmosphere. This work is a first exploration of this question.
In section 2, three different formulations for averaging thermal properties of materials are presented. The
numerical framework is explained in section 3. The
comparison between the different proposals is carried
out in section 4. Conclusions and future directions are
discussed in section 5.
VOLUME 48
perature of the averaged material. So, clearly the
thermal properties of the averaged material should be
such that the previous relation [Eq. (2)] is satisfied.
Therefore, the surface temperature of the averaged
material is equal to the weighted sum of the individual
surface temperatures.
Within a layer of homogeneous material, the temperature T satisfies the heat diffusion equation:
›T
› l ›T
5
,
›t
›z c ›z
(3)
where l (W m21 K21) is the thermal conductivity and
c (J m23 K21) is the volumetric heat capacity of the material. So, the question is how to determine the physical
properties l, c, and the thickness d of the new (averaged)
material such that
N
2. Theoretical framework
T(z 5 d, t) 5
In an urban neighborhood of interest, the walls and
roofs of the buildings are generally built with a selection
of different materials. Because we are interested in interactions with the atmosphere, the averaged material
properties for walls and roofs used in urban parameterizations should give the same total sensible heat flux
as the weighted sum of the sensible heat fluxes from the
individual materials considered.1 So, in a general sense,
indicating with aj the area fraction of the material j in
the zone and indicating with Hj the sensible heat flux
from the material j to the atmosphere, we are seeking a
set of averaged material thermal properties that would
yield a sensible heat flux H such that
N
approach a: d 5
c5
å
aj h(T j
j51
l5
(1)
T A ),
(2)
approach b: d 5
and
å
aj cj ;
j51
(5a)
å
aj dj ,
j51
l5
å
aj dj lj
j51
N
,
and
å
aj dj
j51
N
c5
å
aj dj cj
j51
N
;
and
(5b)
å
aj dj
j51
where TA is the air temperature, Tj is the surface temperature of the jth material, and T is the surface tem-
N
approach c: d 5
1
We want to stress here that we are not proposing to average
together materials from roofs with those from walls. Rather, the
averages are among roof materials and wall materials, to arrive at
average values for each of these two surfaces. Urban canopy parameterizations, in fact, resolve different budgets for walls and
roofs, because the radiation and dynamics behave differently for
vertical and horizontal surfaces.
å
aj lj ,
j51
N
N
T A) 5
å
aj dj ,
j51
N
Assuming that the convective heat exchange coefficients
h are equal for each different material, this translates into
h(T
N
N
å
aj H j .
j51
(4)
where dj represents the thickness of the jth material
present in the urban zone. The physical properties (d, c,
and l) of the averaged material could be determined in
many ways, but three approaches will be analyzed here—
N
H5
å
aj T j (z 5 dj , t),
j51
l5
å
aj dj ,
j51
cpZD2
P
N
c5
å
a j cj ,
j51
and
(5c)
—where lj is the thermal conductivity of the jth material
and cj is its volumetric heat capacity. In the third approach [Eq. (5c)], P is a time period of 1 day (because in
96
AUGUST 2009
1727
NOTES AND CORRESPONDENCE
this context the diurnal cycle typically dominates) and
ZD is a characteristic depth of the averaged material.
The first average [Eq. (5a)] is the standard approach. It
weights the parameters with the area fractions aj of the
different materials present in the urban zone. The second
average [Eq. (5b)] is similar to the first one but weights
layer-integrated thermal conductivity (dj lj; W K21) and
heat capacity (dj cj; J K21 m22) as opposed to weighting
lj and cj alone without accounting for the variation in dj.
By including the dj in the numerator, this last approach
assumes that material thicknesses are small relative to the
damping depth associated with the period of the dominant forcing (i.e., it assumes that the time scale for thermal adjustment across the layer, which depends on lj, cj,
and dj, is small relative to the period of the forcing).
The third average [Eq. (5c)] is significantly different.
A justification for the last average proposed is as follows.
Assuming that at the internal side of a material the
temperature exhibits a periodic signal with period P
and that the amplitude increases as the outdoor side
(where heat exchanges with the atmosphere occur) is
approached, a particular solution of the diffusion equation [Eq. (3)] can be written as
!
!
z
2pt
z
1
sin
,
T j (z, t) 5 T 0 1 DT j exp
ZDj
P
ZDj
(6)
where T0 is a reference temperature, DTj is the amplitude of the sinusoidal (within the material), z is the
distance from the interior boundary of the material, and
ZDj 5 [(ljP)/(cjp)]1/2 is a characteristic depth of the
material (when the amplitude of a signal is damped, ZDj
is known as the ‘‘damping depth’’). Using the same
reasoning for the averaged material, Eq. (4) can be
written as
TABLE 1. Thermal properties of the different materials
(Clarke et al. 1991).
Materials
l (W m21 K21)
c (MJ m23 K21)
1: Metal
2: Softwood
3: Concrete
4: Brick
72.0
0.14
0.87
0.71
8.73
1.11
1.52
1.26
which is independent of the amplitude DT. With this
approach, the physical parameters d and c of the averaged material can be chosen freely but the thermal
2
conductivity [l 5 (cpZD
)/P] is fixed through Eq. (8) and
the consideration that the averaged temperature is described by the left-hand side of Eq. (7). The assumptions
and simplifications used in the above derivation are
never completely fulfilled in any real situation. However, the goal is to find an averaged set of material
properties (d, c, and l) that is able to satisfy Eq. (2), and
we conduct numerical simulations solving Eq. (3) (because no analytical solutions of the diffusion equation
exist for most real situations) to evaluate the physical
appropriateness of our approach. If the numerical results indicate that the third average [Eq. (5c)] better
describes the interaction between urban surfaces and
the atmosphere, the above hypotheses are supported. In
essence, the new approach consists of calculating an
average from an analytical solution of the heat conduction equation and then comparing the results with
those obtained numerically, with less stringent and more
realistic assumptions. Thus, a large number of simulations are carried out, and for each proposed average in
Eq. (5) the surface temperature T is compared with the
‘‘correct’’ average temperature
N
d
2pt
d
1
sin
T 0 1 DT exp
ZD
P
ZD
"
!
!#
N
dj
dj
2pt
sin
. (7)
aj T 0 1 DT j exp
5
1
P
ZDj
ZDj
j51
å
Considering the case in which each of the different
materials has the same amplitude DTj, Eq. (7) can be
reduced to
ZD 5
2
d
N
å
dj
!
dj
!3 ,
6 aj exp
7
sin
6 j51
ZDj
ZDj 7
6
7
arctan6 N
!
!7
6
dj
dj 7
4
5
aj exp
cos
Z
Z
j51
Dj
Dj
å
(8)
å
aj T j
j51
(i.e., that temperature that will result in the correct
sensible heat flux to the atmosphere).
3. Simulation development
First, nine base-case simulations are carried out using
three typical materials (metal, softwood, and concrete;
see Table 1), with three different thicknesses (d1 5
0.025 m, d2 5 0.05 m, and d3 5 0.10 m) for each material.
In the simulations, Eqs. (3) and (9) are solved numerically. The nomenclature used to describe a base case is
ki_dl, meaning that the simulation was carried out for
the material i (ki 5 li/ci, i 5 1, 2, 3) with thickness l (dl, l
5 1, 2, 3). These nine simulations, when appropriately
97
1728
JOURNAL OF APPLIED METEOROLOGY AND CLIMATOLOGY
weighted by their area fraction, provide the correct results with which the averaged material simulation results
are evaluated. Second, 162 different simulations were
carried out for each proposed average; each one of these
simulations is represented by kijk_dlmn. Each of the six
subscripts can take any of the values 1, 2, and 3, but the
first three cannot be repeated, for a total of 6 3 27 5 162
different scenarios. In other words, a given simulated
average material kijk_dlmn may be ‘‘composed’’ of materials of identical depth (the idea is to consider all of the
possible combinations of thickness) but not identical
material type. The case kijk_dlmn represents the averaged material that attempts to describe an urban zone
formed by 20% of the ki material with dl thickness, 30%
of the kj material with dm thickness, and 50% of the kk
material with dn thickness (these percentages have been
chosen arbitrarily and are fixed in all of the comparisons
so as to keep the number of scenarios reasonable). To
assess the impact of including one ‘‘outlier’’ material in
terms of its thermal properties (metal), the above simulations are repeated with brick (Table 1) instead of
metal.
In numerically solving Eq. (3), two different boundary
conditions are considered here. With the first, the internal
boundary condition has been fixed (i.e., the temperature
at the deepest level inside the material is constant for the
whole simulation period); with the second, the internal
surface temperature can change freely such that ›T/›z 5
0 at the lower boundary throughout the simulation. One
might expect the former boundary condition to more
closely approximate homogeneous walls and roofs close
to the relatively constant internal building temperature,
whereas the latter is expected to better represent external
layers closer to ambient forcing and interior layers in
contact with insulation materials. At the surface, the
boundary condition is defined by solving an energy budget equation in both cases (an explanation of the symbols
used can be found in the appendix):
QG 5 Q*
Qh ,
(9)
where Q* 5 (1 2 a)KY 1 «(LY 2 sT 4sfc) and Qh 5 h(Tsfc
2 Ta). The term QG (storage heat flux density) is the net
heat flowing into the material. The sensible heat flux
from the surface is a function of the difference between
the air temperature and the surface temperature, and of
the wind speed through the convective heat transfer
coefficient h. A constant value for h is considered in the
simulations, representing a day with little wind variability, for simplicity. The air temperature Ta and the
downward shortwave radiation KY are taken as follows
(values of the parameters can be found in Table 2):
VOLUME 48
TABLE 2. Inputs and parameters used in the numerical
simulations.
Tmax (K)
Tmin (K)
h (W m22 K21)
«
a
LY (W m22)
K0 (W m22)
Time of simulation (h)
Time step (s)
Initial temperature (K)
Ta 5
T max
T min
2
2pt
sin
P
2pt
KY 5 max K0 sin
P
303
283
6
0.95
0.2
300
800
48
30
293
T
1 T min
3p
1 max
4
2
p
,0 .
2
and
(10)
Note that the radiation reaches a maximum p/4 or oneeighth of a cycle before the air temperature (i.e., 3 h for
diurnal cycles). The heat diffusion equation is solved by
an implicit finite difference approach at each time step
by inverting the corresponding tridiagonal matrix.
4. Simulation results
To compare the averaging methods, the sensible heat
obtained in each case (i.e., for each kijk_dlmn) is compared with the sensible heat representative of the urban
zone: aQhi_l 1 bQhj_m 1 gQhk_n (where Qhi_l is the
sensible heat obtained with the ki material of depth dl,
and so forth, and a 5 0.2, b 5 0.3, and g 5 1 2 a 2 b are
the area fractions fixed previously). Sensible heat flux is
used for the comparison, because equivalent results are
obtained for the surface temperature since the outdoor
temperature Ta and convective heat transfer coefficient
h do not vary between simulations. The root-meansquare error (RMSE) is computed using results at all
5760 time steps (2 days with 30-s time steps).
The results for the different averages can be seen in
Fig. 1 for the first boundary condition (internal temperature fixed), and in Fig. 2 for the second (internal
temperature free) for different area fractions and thicknesses of metal, softwood, and concrete. It is difficult to
distinguish which of the first two (approach a or approach b) averages is superior; however the approach-c
average notably improves the results—in particular, for
the case of the fixed internal temperature. This is not
unexpected given that the assumption in Eq. (8) that the
DTj are equal for all materials is better satisfied with this
boundary condition. RMSE for each combination of
the three different materials does not exhibit coherent
98
AUGUST 2009
NOTES AND CORRESPONDENCE
1729
FIG. 1. RMSE of the sensible heat Qh for fixed internal temperature obtained with the three different averages (approaches a, b, and c)
and for the combination formed by (a) 20% of the k1 material, 30% of the k2 material, and 50% of the k3 material with 27 combinations of
thicknesses dlmn (represented by the notation k123_dlmn); (b) 20% k1, 30% k3, and 50% k2 (k132_dlmn); (c) 20% k2, 30% k1, and 50% k3
(k213_dlmn); (d) 20% k2, 30% k3, and 50% k1 (k231_dlmn); (e) 20% k3, 30% k1, and 50% k2 (k312_dlmn); and (f) 20% k3, 30% k2, and 50%
k1 (k321_dlmn). In (a)–(f), at the extreme left there is d111 and at the extreme right there is d333. From left to right, the n index permutes the
fastest and the l index permutes the slowest (i.e., the sequence is d111, d112, d113, etc.).
99
1730
JOURNAL OF APPLIED METEOROLOGY AND CLIMATOLOGY
VOLUME 48
FIG. 2. As in Fig. 1, but for free internal temperature.
100
AUGUST 2009
NOTES AND CORRESPONDENCE
TABLE 3. Mean and standard deviation of the RMSE values for
each of the three thermal property averaging approaches.
Avg proposed
Mean RMSE (W m22)
Std dev (W m22)
Internal temperature fixed (metal, softwood, concrete)
Approach a
51.42
612.18
Approach b
50.42
611.34
Approach c
19.86
610.47
Internal temperature free (metal, softwood, concrete)
Approach a
38.42
68.99
Approach b
36.57
613.94
Approach c
22.57
69.73
Internal temperature free (brick, softwood, concrete)
Approach a
8.64
65.33
Approach b
8.42
64.41
Approach c
5.92
63.89
behavior dependent on the thickness of the materials
[the thinnest is represented on the left (d111) and the
thickest is represented on the right (d333) in all of the
figure panels]. Only with the free internal boundary
condition, large kj (5lj/cj) (relative to the mean kj), and
small dj does approach b clearly outperform approach a
and demonstrate RMSE equal to that of approach c (not
shown). Thus, average b only yields good results for a
relatively small subset of the total number of simulations
(about 20) when certain conditions are met. It seems
clear that the assumption of small damping depth relative to layer thickness (which underpins approach b)
does not hold for most of the combinations of depths and
materials modeled here.
To obtain a better overall picture of the differences,
the mean and standard deviation of the RMSE values
are calculated for each averaging approach. The results
show that overall the approach-b average is similar to
the approach-a average (Table 3). On the other hand,
the best results are obtained on average for the weighting
approach c. However, when brick replaces metal (simulations with the internal boundary condition ›T/›z 5 0),
RMSE magnitude decreases substantially and the three
averaging methods converge to a significant extent
(Table 3), but the relative decrease in RMSE with averaging method c still remains significant (’30% decrease vs ’40% with metal). Thus, the chosen averaging
method is less important in an absolute sense when
materials with similar thermal behavior are averaged, as
evidenced by the overall RMSE decrease. In essence,
the greater the variability in a material thermal property
is, the more poorly any average value may be expected
to represent the distribution of material thermal properties. Nevertheless, method c appears, on average, to
yield a significantly smaller RMSE in a relative sense.
Last, the downward radiation used so far in the simulations is more representative of horizontal than ver-
1731
TABLE 4. As in Table 3, but for the case in which the downward
shortwave radiation is zero before midday.
Avg proposed
Mean RMSE (W m22)
Std dev (W m22)
Internal temperature fixed (metal, softwood, concrete)
Approach a
39.87
69.27
Approach b
39.09
68.68
Approach c
15.48
67.21
Internal temperature free (metal, softwood, concrete)
Approach a
33.83
66.97
Approach b
32.59
610.48
Approach c
19.13
68.40
tical surfaces (which may be shaded for portions of the
day). To take into account the effects of vertical surface
(wall) orientation, the metal, softwood, and concrete
simulations are repeated with a different variation of
incident shortwave radiation. The KY is set to zero before midday (P/2), at which point it jumps to the value
prescribed in Eq. (10) where it remains for the duration
of the day. This rapid variation in solar forcing is typical
of a west-facing wall in the Northern Hemisphere, for
example. The results are similar to those obtained previously (see Table 4), and again the approach-c average
is superior.
5. Conclusions
This work is a first step toward the determination of
physical properties that represent the behavior (in terms
of its interaction with the atmosphere) of an ensemble of
different materials in an urban zone or neighborhood. A
new averaging method for thermal parameters has been
proposed [Eqs. (5c) and (8)], and better results have
been obtained—in particular, for an ensemble of materials with large variability in thermal behavior. To use
the new averaging scheme, information on the area occupied by each material and the thickness of each material is needed. This information can be obtained from
building-construction databases. Similar formulations
can be used for other surface descriptors necessary in
(urban) mesoscale atmospheric modeling. That is, the
equation that describes the physics of the problem for an
ideal situation is solved, and subsequently new values of
physical parameters that better represent the net effect
of the subgrid-scale heterogeneity are obtained.
Acknowledgments. The authors thank CIEMAT for
the doctoral fellowships held by Francisco Salamanca
and NSERC and UBC for the doctoral scholarships held
by Scott Krayenhoff. We also thank the reviewers for
important comments on the manuscript. This work was
funded by the Ministry of Environment of Spain.
101
1732
JOURNAL OF APPLIED METEOROLOGY AND CLIMATOLOGY
APPENDIX
List of Symbols Used in Solving Energy Budget
Equations
Q* Net radiative flux density (W m22)
Qh Sensible heat flux density (W m22)
KY Downward shortwave radiative flux density (W m22)
K0 Maximum value of the shortwave radiation (W m22)
LY Downward longwave radiative flux density (W m22)
Tsfc Surface temperature (K)
Ta Air temperature (K)
Tmax Maximum value of the air temperature (K)
Tmin Minimum value of the air temperature (K)
a
Shortwave albedo
«
Longwave emissivity
s
Stefan–Boltzmann constant (W m22 K24)
h
Convective heat transfer coefficient (W m22 K21)
VOLUME 48
REFERENCES
Clarke, J. A., P. P. Yaneske, and A. A. Pinney, 1991: The harmonisation of thermal properties of building materials. Tech.
Note 91/6, BEPAC, 87 pp.
Kanda, M., T. Kawai, M. Kanega, R. Moriwaki, K. Narita, and
A. Hagishima, 2005: A simple energy balance model for regular building arrays. Bound.-Layer Meteor., 116, 423–443.
Kusaka, H., H. Kondo, Y. Kikegawa, and F. Kimura, 2001: A simple
single-layer urban canopy model for atmospheric models:
Comparison with multi-layer and slab models. Bound.-Layer
Meteor., 101, 329–358.
Martilli, A., 2007: Current research and future challenges in urban
mesoscale modelling. Int. J. Climatol., 27, 1909–1918.
——, A. Clappier, and M. W. Rotach, 2002: An urban surface exchange parameterization for mesoscale models. Bound.-Layer
Meteor., 104, 261–304.
Masson, V., 2000: A physically-based scheme for the urban energy
budget in atmospheric models. Bound.-Layer Meteor., 94,
357–397.
Oke, T. R., 2006: Towards better scientific communication in urban
climate. Theor. Appl. Climatol., 84, 179–190.
102
Capítulo 3
103
Capítulo 3
3.4 Un estudio de la capa límite urbana usando diferentes parametrizaciones urbanas y
resoluciones de parámetros morfológicos que describen una ciudad con el modelo
atmosférico WRF.
Salamanca, F., A. Martilli, M. Tewari, and F. Chen, 2010b: A study of the urban boundary layer using
different urban parameterizations and high-resolution urban canopy parameters with WRF. Journal of
Applied Meteorology and Climatology (accepted).
En esta sección se presentan los resultados de un estudio numérico de la capa límite urbana
sobre la ciudad de Houston, Texas. En la primera parte del trabajo se ha hecho una intercomparación entre cuatro parametrizaciones urbanas disponibles en el modelo atmosférico
WRFv3.2. Aunque los resultados podrían ser dependientes del esquema turbulento usado para
la parametrización de la capa límite, la isla de calor sobre la ciudad presentó notables
diferencias dependiendo del esquema urbano utilizado. Con el esquema urbano BEP+BEM
desarrollado en esta tesis se obtuvieron buenos resultados y el calor proveniente de los
sistemas de aire acondicionado aumentó notablemente la intensidad de la isla de calor.
En la segunda parte del trabajo se repitieron las simulaciones con los esquemas urbanos BEP
y BEP+BEM pero esta vez en lugar de usar la información definida en las diferentes clases
urbanas (como se hizo en la primera parte) se utilizó información detallada de la morfología
de la ciudad con una resolución de ~ 1km 2 extraída de la base de datos NUDAPT. Los
resultados indican que el nuevo esquema BEP+BEM es más sensible a los parámetros
morfológicos que el anterior esquema BEP. Esto es debido a que el calor antropogénico
derivado de los sistemas de aire acondicionado es “proporcional” al volumen de aire que
existe en el interior de los edificios y éste depende de las dimensiones de los mismos. Cuando
se calculó el consumo energético se obtuvieron buenas estimaciones, al considerar la
información detallada existente en la base de datos NUDAPT, comparadas con valores
obtenidos con otras técnicas completamente diferentes.
104
Chapter 3. A study of the UBL using different urban parameterizations.
A study of the urban boundary layer using different urban
parameterizations and high-resolution urban canopy parameters with
WRF
Francisco Salamanca 1 · Alberto Martilli 1 · Mukul Tewari 2 · Fei Chen 2
1
CIEMAT (Research Center for Energy, Environment and Technology), Madrid, Spain
2
NCAR (National Center for Atmospheric Research), Boulder, CO, USA
ABSTRACT
In the last two decades, mesoscale models (MMs) with urban canopy parameterizations have been
important tools in the study of the Urban Boundary Layer (UBL). The correct performance of these
models requires a detailed and up-to-date land use/cover dataset of the area under study. The difficulty of
obtaining this information has been an important limitation for UBL modelling studies in the recent years.
Nowadays, high resolution urban canopy parameters (UCPs) have become available for an increasing
number of cities. One of the most important questions for the urban modelling community is to define the
needed degree of complexity of the urban canopy parameterizations and the resolution and details
necessary in the UCP datasets to obtain accurate results. In this work, and in an attempt to answer the
previous questions, four urban canopy schemes, with different degrees of complexity, have been run with
the Weather Research and Forecasting (WRF) model using a highly-detailed UCP database for the city of
Houston for two days of August 2000. Results have been intercompared and compared with
measurements. The statistical analysis shows a tendency to overestimate the air temperatures by the
simple bulk scheme and underestimate the air temperatures by the more detailed urban canopy
parameterizations. The three-dimensional analysis of the meteorological fields revealed a possible impact
of both the urban schemes and the UCP on cloud formation. Moreover, the impact of the air conditioning
(AC) systems on the air temperature and their energy consumption has been evaluated with the most
detailed urban scheme for the two simulated days. During the night, this anthropogenic heat (AH) was
responsible for an increase in the air temperature of up to 2 ºC in the most dense urban areas, and the
estimated energy consumption was of the same magnitude than energy consumption obtained with
different methodologies. Due to different meteorological conditions existing between the two selected
days, differences up to 20% in the energy consumption were obtained. Based on the results for the present
case study, we can conclude that the urban fraction, the urban material’s properties, and the AH are the
most important factors that contribute to the Urban Heat Island (UHI) phenomenon. The urban geometry
(height and dimensions of the buildings) plays a secondary role in the formation of the UHI and becomes
relevant because it affects the AH ejected by the AC systems.
______________________________________________________________________
Journal of Applied Meteorology and Climatology,
(accepted)
Corresponding author address: F. Salamanca, Research Centre for Energy,
Environment and Technology (CIEMAT), Avenida Complutense 22, 28040 Madrid,
Spain.
E-mail: [email protected]
105
Chapter 3. A study of the UBL using different urban parameterizations.
1. Introduction
At the end of the last century, the
understanding of the Urban Boundary
Layer (UBL) was greatly increased
thanks to the use of modern mesoscale
models (MMs) in combination with the
development of the urban canopy
parameterizations.
The
urban
parameterizations communicate the
mean thermal and dynamic effects of
the city to the MMs. The first urban
schemes represented the thermal effects
of the city using large values for the
heat capacity and thermal conductivity
to reproduce the large heat storage that
takes place in the urban surfaces. These
simple approaches have a disadvantage
in that they cannot represent the
heterogeneities present in the urban
areas. In the same way, large values for
roughness parameters were used to
represent the turbulence generated by
roughness elements. Subsequently, the
first single-layer urban-canopy models
appeared (e.g., Masson 2000, Kusaka et
al. 2001, Kanda et al. 2005). They
represented the urban geometry by
infinitely long street canyons, and three
different urban surfaces (walls, roofs,
and roads). With these new approaches,
different urban classes can be
considered with different thermal
properties, and the heterogeneities of
the city are better represented. Finally,
the appearance of multilayer urbancanopy models (Martilli et al. 2002,
Kondo et al. 2005) permitted a direct
interaction with the Planetary Boundary
Layer (PBL). To date, the coupling
between simple building energy models
and
multilayer
urban
canopy
parameterizations (Kikegawa et al. 2003,
Salamanca
and
Martilli,
2010)
represents the most sophisticated
approach and permits the study of the
effect of anthropogenic heat fluxes in
urban environments. This increasing
number of urban parameterizations and
the important differences existing
between them demand a study where
the positive and negative points must be
displayed to facilitate their use. In this
direction, an important effort, led by
Grimmond et al. (2009), has been
carried out by comparing energy fluxes
obtained with a wide range of urban
models run offline against site
observations.
Bulk
UCM
BEP
BEP+
BEM
How the
canopy is
resolved
No
canopy
–
roughne
ss
length
modifie
d.
Single
Layer
Multil
ayer
Multil
ayer
Anthropo
genic
heat
NO
From
fixed
temporal
profiles
NO
From
a
buildi
ng
energy
model
Accounti
ng for
fraction
of
vegetatio
n
NO
YES
YES
YES
PBL
scheme
used in
this study
Mellor
Yamada
Janjic
Mellor
Yamada
Janjic
Bouge
ault
and
Lacarr
ère
Bouge
ault
and
Lacarr
ère
Table 1. Overview of the different urban
schemes in the intercomparison.
Similarly, an intensive effort has
been carried out for the community
mesoscale WRF (Weather Research and
Forecasting) model (Chen et al. 2010)
to assess urban environmental problems
such as the urban heat island (UHI) and
urban air pollution. Following this line,
in the first part of this article, WRF has
been run with four different urban
canopy parameterizations over the city
of Houston (see Table 1), and the results
are inter-compared and compared
against measurements. The first urban
106
Chapter 3. A study of the UBL using different urban parameterizations.
parameterization (included in WRF
since 2003) is a simple bulk scheme that
represents the effects of urban surfaces
by means of a roughness length of 0.8
m to represent the turbulence generated
by roughness elements and drag due to
buildings, a surface albedo of 0.15 to
represent the radiation trapping in the
urban canyons, a volumetric heat
capacity of 3.0 Jm −3K −1 , and a thermal
conductivity of 3.24 Wm−1K −1 to
represent the large heat storage in the
urban buildings and roads. This
approach has been
successfully
employed in real time forecasts (Liu et
al.
2006).
The
second
urban
parameterization was developed by
Kusaka et al. (2001), and Kusaka and
Kimura (2004). It is a single-layer
urban-canopy model (UCM) where the
anthropogenic heat (AH) can be added
to the sensible heat flux in the urban
canopy layer. The urban geometry is
represented through infinitely long
street canyons, and three different urban
surfaces (roof, wall, and roads) are
recognized. In a street canyon,
shadowing, reflections, and trapping of
radiation are considered, and an
exponential wind profile is prescribed to
deduce the wind speed in the canyon
from the wind speed above the canyon,
where the lowest grid point of the
mesoscale model is located. The total
sensible heat flux from roof, wall, and
roads is passed to the lowest
atmospheric layer. This option has been
included in WRF (V2.2) since 2006.
The third urban parameterization was
developed by Martilli et al. (2002), and
it is a multilayer urban canopy scheme
called
BEP
(Building
Effect
Parameterization) that represents the
most sophisticated urban modelling in
WRF (included in WRF V3.1 release
since 2009) that allows a direct
interaction with the PBL. BEP
recognizes the three-dimensional nature
of urban surfaces and the fact that
buildings vertically distribute sources
and sinks of heat and momentum
through the whole urban canopy layer.
It takes into account the effects of the
vertical (walls) and horizontal (streets
and roofs) surfaces on momentum,
turbulent kinetic energy, and potential
temperature. The wall and road
radiation
considers
shadowing,
reflection, and trapping of shortwave
and longwave radiation in street
canyons.
The
last
urban
parameterization is an extension of the
BEP scheme and was developed by
Salamanca et al. (2010). It is the result
of the coupling between the BEP and a
simple building energy model (BEM)
that improves the results obtained with
the old version of BEP (Salamanca and
Martilli, 2010). BEM accounts for the 1)
diffusion of heat through the walls,
roofs, and floors; 2) radiation
exchanged through windows; 3)
longwave radiation exchanged between
indoor surfaces; 4) generation of heat
due to occupants and equipment; and 5)
air conditioning, ventilation, and
heating.
The
BEP+BEM
parameterization takes into account the
exchanges of energy between the
interior of the buildings and the outdoor
atmosphere. Consequently, the impact
of the air conditioning systems (AC)
and their energy consumption can be
estimated. The new BEP+BEM scheme
has been included in WRF V3.2 release
on April 2010.
Originally, in the WRF model, the
urban schemes looked up the input
parameters for the urban morphology
from a table with only three different
urban classes (commercial, industrial,
and high or low residential areas) that
can be derived from land use databases
(e. g., the National Land Cover Data for
the US, NLCD, developed by the U.S.
Geological Survey, USGS). In the intercomparison, three urban classes derived
from NLCD have been considered, and
the input parameters for every urban
107
Chapter 3. A study of the UBL using different urban parameterizations.
class have been extracted from the
reports of Burian et al. (2003).
To evaluate the impact of highresolution urban land cover databases in
the mesoscale weather prediction
models, a project called the National
Urban Database and Access Portal Tool
(NUDAPT) was created to provide
urban
data
and
improve
the
parameterization of UBL processes
(Ching et al. 2009). In this database,
information exists relative to urban
morphology for an important number of
cities in the USA. An important
advantage of using NUDAPT is that the
inputs of the urban parameterizations
(urban fraction, building height
histograms, building plan area fraction,
mean building height weighted by
footprint plan area, etc.) are provided
with a resolution of 1 km 2 , and not need
to be defined for every urban class.
In the second part of the work, to
evaluate the impact of high-resolution
urban land cover databases, the urban
canopy parameters (UCP) existing in
NUDAPT for the city of Houston were
introduced into the input files of the
WRF model as new variables, and the
results were obtained with the two
approaches: a) using urban parameters
from NUDAPT and b) using urban
parameters from urban look-up a table.
These results were compared. Due to
the fact that the bulk scheme does not
use urban morphological parameters,
and that the UCM is not yet ready to
directly use the urban information from
NUDAPT, only the BEP and
BEP+BEM schemes were considered
for this second comparison. For every
urban grid point, BEP and BEP+BEM
use directly the information from
NUDAPT not averaged values defined
for every urban class.
In Section 2, a description of the
simulations and the results for the four
urban models are presented. Differences
obtained using the more detailed urban
database (NUDAPT) are shown in
Section 3 for the BEP and BEP+BEM
schemes. Exploring the possibilities of
the new BEP+BEM parameterization
and the impact of the AC systems and
their energy consumption are discussed
in Section 4. Finally, conclusions and
future directions are mentioned in
Section 5.
2. WRF simulations with different
urban models
2.1 Numerical domain and set up of the
simulations
Figure 1. Configuration of the 4 two-way-nested
domains for the WRF simulations. The grid
sizes for the four domains are 27, 9, 3, and 1km,
respectively. Terrain height interval is 200 m.
Two summer days have been
analyzed: 25th and 31st of August 2000.
The 24-h simulations began at 12 UTC
(0600 LST) and a set of eight
simulations (four urban schemes for
every selected day) were performed
using the non-hydrostatic version of the
WRF V3.1 model (Skamarock et al.
2008), coupled to the Noah land surface
model (Chen and Dudhia, 2001, and Ek
et al. 2003) for the non-urban part. This
surface-hydrology model has one
canopy layer and the following
prognostic variables: soil moisture and
temperature in the soil layers, water
stored in the vegetation canopy, and
snow stored on the ground. The
horizontal domain (see Fig. 1) was
composed of four two-way nested
domains with 100 × 100 , 174 × 156 ,
219 × 186 , and 216 × 198 grid points,
and a grid spacing of 27 , 9 , 3 , and
108
Chapter 3. A study of the UBL using different urban parameterizations.
1 km , respectively. The 24-h
simulations were conducted with the
initial and boundary conditions from the
operational
National
Centre for
Environmental Prediction (NCEP) with
a grid resolution of 40km and a time
resolution of 3 h. To take full advantage
of the urban parameterizations, a
vertical resolution of 40 eta levels1 was
used (14 levels in the lowest 1.5 km )
with the lowest level 22 m above the
ground.
The
selected
radiation
parameterizations were the Dudhia
(1989) shortwave radiation scheme, and
the Rapid Radiative Transfer Model
(RRTM) longwave parameterization
(Mlawer et al. 1997). The microphysics
package is WSM3 (Hong et al. 2004);
no cumulus cloud scheme was used in
the inner domain. Two different PBL
schemes were chosen, the Mellor and
Yamada (Janjic, 1994) for the bulk and
UCM parameterizations and the
Bougeault and Lacarrère (1989) for the
BEP and BEP+BEM schemes. This
selection is motivated by the fact that
the first two parameterizations (bulk
and UCM) have been extensively tested
coupled to the MYJ scheme, while the
other two (BEP, and BEP+BEM), even
if they can also be run with the MYJ
scheme, have been mainly tested
together with the Bougeault and
Lacarrère (1989) PBL scheme. We are
aware that this choice may introduce
1
Full eta levels =1., 0.9974, 0.9940, 0.9905, 0.9850, 0.9800,
0.9700, 0.9600, 0.9450, 0.9300, 0.9100, 0.8900, 0.8650,
another source of differences, but we
considered that these are the
configurations where each scheme can
perform best. Future work will be
needed to investigate the sensitivity of
the UCP schemes to the coupling with
different PBL schemes.
In the determination of the fluxes
provided by the urban canopy
parameterization as lower boundary
conditions for coupled atmospheric
models, the fractional area occupied by
impervious surfaces (urban fraction)
plays a fundamental role. The urban
fraction (λU ) in a particular patch is
defined as the fractional area covered by
the buildings λ P (building plan area
fraction) plus the fractional area covered
by the roads. These parameters are
especially important because they are
used to obtain the dimensions of the
buildings and roads in the urban
schemes. For example, for a 2-D urban
canopy parameterization,
λP
b
=
,
(1)
λU b + w
where (b) and ( w) are the widths of the
buildings and roads, respectively. Other
morphological urban parameters used to
derive the inputs of the urban models
are the mean building height weighted
__
by building plan area ( h ) and the
__
building height-to-width ratio ( λ S ) .
These parameters are calculated using
the following equations:
N
0.8400, 0.8100, 0.7800, 0.7500, 0.7100, 0.6800, 0.6450,
0.6100, 0.5700, 0.5300, 0.4900, 0.4500, 0.4100, 0.3700,
0.3300, 0.2900, 0.2500, 0.2100, 0.1750, 0.1450, 0.1150,
0.0900, 0.0650, 0.0450, 0.0250, 0.0100, and 0.0000. The
vertical
σ
coordinate
is
surf
as
i =1
N
i =1
Ai
__
and λ S =
__
h
__
,
w
(2)
where ( Ai ) is the plan area at ground
is surface dry hydrostatic
level of building i , (hi ) is its height,
where
p is
pressure, and ptop is a constant dry hydrostatic pressure at
model top.
h=
Ai hi
the dry
( p − ptop ) /( p surf − ptop ) ,
hydrostatic pressure, p
defined
__
__
(N ) the number of buildings, and w
the mean road width. In the simulations,
the values of the above parameters used
109
Chapter 3. A study of the UBL using different urban parameterizations.
for every urban class for the city of
Houston were extracted from the reports
of Burian et al. (2003) and can be seen
in Table 2. Moreover, for the correct
performance of the BEP and BEP+BEM
schemes, a building height distribution
is necessary for every urban class, and
the considered values are in Table 2.
The thermal properties of the buildings
used in the simulations are in Table 3.
For the UCM a diurnal profile of AH
was added to the sensible heat flux
(hereafter this simulation is referred as
UCM+AH) with peak values of 90, 50,
and 20 Wm −2 for the COI (commercial
or industrial), HIR (high intensity
residential), and LIR (low intensity
residential) urban classes, respectively.
Unlike the UCM parameterization, the
BEP+BEM scheme computes the AH
released into the atmosphere to maintain
the indoor temperature of the buildings
in a range of comfort defined by the
user (see Salamanca et al. 2010) by
means of an AC model that computes
the total cooling loads for every floor in
a particular building. For these
simulations the amplitude of the range
of comfort was fixed to 1 ºC with the
target internal temperature being 25 ºC.
Other parameters of the BEP+BEM
scheme were fixed (for the three urban
classes) to COP (Coefficient of
Performance of the AC systems) = 3.5,
number of occupants
= 0.02
2
person/m of floor, sensible heat
generated by equipment = 30 Wm−2 of
floor from 0800 LST to 2000 LST and
10 Wm−2 the rest of the day. The
considered values for the sensible heat
generated by equipment are similar to
others estimations based on temporal
variations
of
electric
power
consumption for lighting (Kikegawa et
al. 2003) in business districts.
2.2 Analysis of the results.
Houston is an area subject to complex
mesoscale dynamics leading to complex
land and sea breezes. The main forcings
determining the circulation are the
contrast between the land and the sea
and the general circulations patterns.
The urban area modulates these
forcings. However, given the purpose of
this article, in the following analysis,
stress is put mainly on the impact of the
urban forcing.
LIR
HIR
COI
(Low
(High
(Commer
Intensity
cial
Intensity
Residential
Industrial)
Residenti
)
or
al)
Urban
fraction (λU
0.429
0.429
0.865
0.06
0.17
0.21
5.4
5.1
8.9
0.05
0.13
0.09
55
59
37
30
34
34
15
7
9
0
0
20
)
Building plan area
(λ P )
fraction
Building
height
weighted
by
building plan area
__
( h (m))
Building
height-
to-
width
__
ratio ( λ S )
% of buildings of
5 m of height
% of buildings of
10 m
of height
% of buildings of
15 m
of height
% of buildings of
20 m
of height
Table 2. Urban morphological parameters
considered for the three urban classes.
110
Chapter 3. A study of the UBL using different urban parameterizations.
Figure 3. Time series of 2-m air temperature for different stations for 25 August 2000 obtained with the
four urban schemes against measurements.
111
Chapter 3. A study of the UBL using different urban parameterizations.
Figure 4. As in Figure 3 but for 31 August 2000.
112
Chapter 3. A study of the UBL using different urban parameterizations.
2.2.1 Air temperature
Table 4 lists the nine monitoring
stations used in the evaluation of the
four urban models for the two selected
days. The stations were displayed over
the city of Houston, and their location
can be seen in Fig. 2b. The stations
selected are representative of the three
urban classes considered in this study,
so that the behaviour in simulating
urban areas with different morphology
can be analysed. It is important to
remember that we assume that point
measurements can be compared to
model’s outputs, which represent
spatially averaged values over a grid
cell of 1 km2. An analysis of this
assumption
requires
detailed
information on the position of the
station and the morphology of the
surrounding area and goes beyond the
scope of this work.
Surface
λ (Wm −1
K −1 )
T
(int)
m −3 K −1 )
( º C)
C (⋅10 6 J
ε
α
z0
(m )
Roof
0.695
1.32
20
0.9
0.2
0.01
Wall
0.695
1.32
20
0.9
0.2
____
Road
0.4004
1.40
20
0.95
0.15
0.01
λ is the thermal conductivity of the material, C is the
specific heat of the material, T (int) is the initial
temperature of the material and also temperature of the
deepest layer, ε is the emissivity of the surface, α is the
albedo of the surface, and z 0 is the roughness length for
momentum over the surface.
Table 3. Thermal parameters used in the urban
modules (UCM, BEP and BEP+BEM) and for
every urban class.
The inner simulation domain is plotted
in
Fig.
2a.
The observed and computed statistics
for the 2-m air temperature are shown in
Tables 5 and 6. Mean bias (MB), hit
rate (HR), and root mean square error
(RMSE) were calculated with the
criteria for the HR calculation for
model-observation agreement within 2
ºC following the same criteria of similar
studies (Miao et al. 2009). For the
statistical computations, the first
initializing value was removed to avoid
the existing gap between the observed
and computed data. It is important to
mention in terms of the comparison,
that the observed values are hourly
averaged available data whereas the
WRF
computed
values
are
instantaneous. The four urban schemes
accurately reproduce enough the surface
air temperature for the 25th of August
(see Fig. 3 and Table 5). The best
results were obtained with the
BEP+BEM and BULK schemes. The
BULK parameterization tends to
slightly overestimate the air temperature
(the MB is almost always positive)
while the other schemes tend to
underestimate it. Observing the
differences in the air temperature
between the two multilayer schemes
(BEP and BEP+BEM), it is possible to
say that the effect of the sensible heat
ejected into the atmosphere to maintain
the indoor temperature in a range of
comfort (situation simulated with the
BEP+BEM scheme) has a significant
effect from 1800-1900 LST to the dawn.
In general, the worst estimations were
obtained
with
the
single-layer
UCM+AH scheme, but they notably
improved in the COI areas where the
urban fraction (see Table 2) has the
biggest value. The 31st of August was
the warmest day, with temperatures up
to 41 ºC. The results (see Fig. 4 and
Table 6) indicate a good performance of
the urban models except for the
UCM+AH scheme, which was not able
to correctly simulate the surface air
temperature from the 1800-1900 LST
except in the COI areas. We are aware
that the UCM parameterization has been
widely tested and validated (e.g.,
Kusaka and Kimura, 2004, Lin et al.
2008; Miao et al. 2009) in different
113
Chapter 3. A study of the UBL using different urban parameterizations.
situations, so the results obtained the
31st of August were probably due to the
low urban fraction used for the HIR and
LIR urban classes, which were derived
from the available information on urban
morphology. It seems that this is not a
setback for the BEP and BEP+BEM
parameterizations. At this point, it is
important to remember that the goal of
this work is twofold: on one hand, we
want to compare the BULK, UCM, BEP,
and BEP+BEM schemes coupled to the
WRF model; on the other hand, we
want to study whether or not there is a
the necessity for using highly-accurate
urban morphology information for the
correct performance of the urban
canopy parameterizations. This is the
reason why we use the realistic urban
fraction (see Table 2) derived from the
Burian et al. (2003) reports for the
different urban classes and not
unrealistic values that could improve
the results in some cases, increasing the
sensible heat flux coming from the
urban canopy layer.
ID
Latitude
Longitude
Sampling
Urban
height
class
above
ground
(m)
C01
29.767778
-95.220556
11
COI
C55
29.733611
-95.257500
5
HIR
C81
29.735000
-95.315556
11
HIR
C146
29.695556
-95.499167
4
LIR
C167
29.734167
-95.238056
11
HIR
C169
29.706111
-95.261111
11
LIR
C404
29.806944
-95.291389
11
HIR
C409
29.623889
-95.474167
11
HIR
C603
29.765278
-95.181111
4
COI
The urban morphology characteristics for every urban class
can be seen in Table 2.
Table 4. List of monitor locations based on
information from Texas Commission on
Environmental Quality (TCEQ).
Figure 2. a) D4 inner domain and urban
fraction (from Table 2) for the city of Houston.
b) Monitoring stations used in the evaluation of
the four urban parameterizations.
25-
BULK
Aug-00
UCM+
BEP
AH
BEP+BE
ID
M
MB
0.9441
-0.1846
-0.4477
-0.1146
RMSE
1.0591
1.2252
1.3012
0.9174
HR
0.9583
0.9167
0.8750
0.9583
MB
1.0156
-0.7632
-0.6829
-0.1733
RMSE
1.4369
2.3053
1.9029
1.4190
HR
0.7917
0.6667
0.5833
0.8750
MB
0.4996
-1.0780
-1.1773
-0.7023
RMSE
0.8864
2.3070
1.7794
1.2975
HR
1.0000
0.6667
0.7083
0.8750
MB
0.9174
-0.7859
-0.7757
-0.2760
RMSE
1.1352
1.6233
1.2510
0.8454
HR
0.9583
0.8333
0.8750
1.0000
MB
0.3214
-1.3697
-1.4086
-0.8324
RMSE
1.3148
2.8136
2.4667
1.8656
HR
0.8333
0.6250
0.5833
0.6250
MB
1.4077
-0.3472
-0.2420
0.2936
RMSE
1.6013
1.7647
1.2407
1.1344
C01
C55
C81
C146
C167
C169
114
Chapter 3. A study of the UBL using different urban parameterizations.
HR
0.7917
0.7500
0.8750
0.9167
MB
0.9149
-1.9793
-0.6860
-0.1065
MB
-0.7305
-2.3791
-2.5281
-1.8296
RMSE
1.3409
3.2054
1.3515
1.1006
RMSE
1.1633
3.1474
2.9407
2.2238
HR
0.9167
0.4583
0.8750
0.9167
HR
0.9583
0.5417
0.4583
0.5000
MB
-1.4044
-4.0587
-3.0171
-2.0374
MB
-0.2036
-1.5205
-1.4305
-1.0776
RMSE
1.7845
4.8135
3.3670
2.2620
RMSE
1.1597
2.0500
1.6861
1.3535
HR
0.7500
0.2917
0.2500
0.4583
HR
0.9583
0.5417
0.6250
0.9167
MB
0.5148
-1.6033
-0.7080
-0.1324
MB
0.3717
-0.4699
-1.0604
-0.4341
RMSE
1.4530
3.1778
1.8077
1.6253
RMSE
0.7086
0.9659
1.5736
0.8885
HR
0.8750
0.6667
0.7083
0.7083
HR
1.0000
0.9583
0.7500
1.0000
MB
0.1513
-0.3021
-1.2128
-0.0001
RMSE
1.0279
1.1052
1.6225
0.9693
HR
1.0000
0.9167
0.7917
1.0000
C404
C409
C603
Table 5. Statistical comparison of the simulated
and observed 2-m air temperature ( º C ; the
criterion for hit rate calculation is 2 º C ) for the
25th Aug 2000.
31-
BULK
Aug-00
UCM+
BEP
AH
BEP+BE
ID
M
MB
0.7332
-0.6974
-0.8760
0.0400
RMSE
1.2844
1.5344
1.7177
1.2112
HR
0.8750
0.8333
0.6667
0.9583
MB
0.4255
-2.1762
-0.9776
-0.2609
RMSE
1.2351
3.6834
2.0046
1.5049
HR
0.9167
0.4167
0.6250
0.7500
MB
0.2226
-2.6085
-1.2804
-0.5661
RMSE
1.0908
3.9160
2.0344
1.3927
HR
0.9583
0.4167
0.5833
0.8750
MB
0.9850
-1.8046
-0.7235
-0.0946
RMSE
1.2923
3.3064
1.6357
1.2688
HR
1.0000
0.6667
0.7500
0.9167
MB
0.4296
-2.1180
-1.0245
-0.2328
RMSE
1.1569
3.1247
1.6067
1.1472
HR
0.9167
0.5417
0.8333
0.8750
C01
C55
C81
C146
C167
C169
C404
C409
C603
Table 6. Statistical comparison of the simulated
and observed 2-m air temperature ( º C ; the
criterion for hit rate calculation is 2 º C ) for the
31st Aug 2000.
2.2.2 Wind field
In Fig. 5, the time evolution of
the wind speed at 10 m above the
ground level (AGL) is compared with
the observations for the 25th of August.
Three monitoring stations are shown
because the rest did not present
remarkable differences. The four urban
models
were
able
to
capture
satisfactorily the rotation and magnitude
of the breeze (from the sea to the land
and vice versa) from around 1400-1500
LST to the dawn. It seems that the
BULK and UCM+AH schemes
sometimes slightly overestimate the
wind speed compared to the BEP and
BEP+BEM
parameterizations
that
underestimate the observed wind. The
same pattern was forecast for all the
parameterizations. During the first
hours of the day, from the sunrise to
1100 LST approximately, the schemes
were not able to capture the wind
direction well. Later, and for a period of
some hours, the direction of the wind
115
Chapter 3. A study of the UBL using different urban parameterizations.
was changing randomly. This behaviour
was due to a cloud modelled over the
city. The absorbed radiation does not
heat the urban surface below the cloud,
and consequently the fresh air blows
away to surrounding warmer areas. This
phenomenon (not shown) was observed
with the four urban schemes in different
places and times.
Figure 5. Time series of observed (OBS.) and
simulated (with the four urban schemes)
horizontal winds at 10 m AGL at three sites for
the 25th of August: a) C55 station, b) C409
station, and c) C603 station.
It is interesting to analyse how
the different surface schemes affect the
cloud formation. In Fig. 6 the shortwave
radiation that reaches the ground (in the
case of built-up areas, this is the
shortwave radiation that reaches the
upper limit of the urban canopy layer) is
shown at 1200 LST. This field reflects
the presence of clouds. It can be seen
that, at that time, the simulations with
the BEP and BEP+BEM schemes
produce clouds above the city, while
UCM and BULK do not. The latter is
related to strong vertical velocities (Fig.
7) at eta level Z = 9 ( ≈ 485 m AGL),
simulated by BEP and BEP+BEM. A
few hours later, similar patterns (clouds
and strong updrafts and downdrafts)
develop also for the UCM and BULK
simulations over the city. A possible
explanation of these differences is that
BEP and BEP+BEM are more sensitive
to the spatial variability of surface
fluxes due to the heterogeneity of the
urban structure. For this reason they
trigger earlier the updrafts and
downdrafts that are responsible for
cloud formation. It is difficult (and goes
beyond the scope of this article) to
decide which is the most realistic
behaviour not only because there are no
cloud
position
and
formation
measurements available during these
days over Houston, but also because
this problem involves an analysis of the
behaviour of the PBL and cloud
schemes (one of the weakest areas of
meteorological models). At this stage,
the
influence
of
the
urban
parameterization on cloud formation is
only a speculation. Further studies
should be dedicated to analyse if
model’s predictions are accurate, if
they are influenced by incorrect
parameterization of the physics of the
phenomena, or by numerical noise.
On the 31st of August, the wind
blew from the northwest during the
morning
until
the 1300
LST
approximately
over
Houston.
116
Chapter 3. A study of the UBL using different urban parameterizations.
Simulations
with
different
parameterizations were able to capture
this flow well, but the presence of
clouds some hours later made
impossible the correct prediction of the
wind field in the monitoring locations
after midday. In the evening and during
all the night, the schemes captured the
clockwise turn of the breeze well (not
shown).
Figure 6. Shortwave downward radiation
reaching the ground obtained with the four
urban models at 1200 LST 25 August 2000: a)
BULK scheme, b) UCM+AH scheme, c) BEP
scheme, and d) BEP+BEM scheme.
117
Chapter 3. A study of the UBL using different urban parameterizations.
Figure 7. Vertical velocity patterns obtained
with the four urban models at 1200 LST 25th
August 2000 at eta level Z =9 ( 485m): a)
BULK scheme, b) UCM+AH scheme, c) BEP
scheme, and d) BEP+BEM scheme.
In Fig. 8, vertical wind speed
profiles (together with the PBL height
predictions) are compared against
observations at Ellington Place for the
25th of August. The four schemes
present similar patterns for the wind
field with a shift in direction in the
lower levels compared to the observed
values. In Fig. 9, thanks to available
data for the two selected days, the PBL
heights forecast by the four schemes
against observations have been plotted
for the Ellington and Southwest Airport
places. The BULK scheme predicts the
highest values and the BEP+BEM
scheme the lowest. It must be noted,
however, that all models computed
strong spatial heterogeneities for this
field (not shown).
118
Chapter 3. A study of the UBL using different urban parameterizations.
Figure 8. Vertical wind speed profiles (together
with the Planetary Boundary Layer height
predictions) for the 25th August 2000 (UTC =
LST+6h) at Ellington Pl for: a) BULK scheme,
b) UCM+AH scheme, c) BEP scheme, d)
BEP+BEM scheme, and e) Observations.
3. WRF simulations with the
NUDAPT data
3.1 Set up of the simulations
Lo et al. (2006) verified that for
obtaining good results in the
meteorological variables in the urban
environment of Pearl River Delta (PRD)
region (China), it was not only
necessary to have an up-to-date urban
land use/cover dataset, but also it was
necessary urban schemes that were able
to distinguish the heterogeneities
present in the cities.
Continuing in this way, we have
analyzed the impact of a high-resolution
urban canopy parameters database on
the UBL over Houston. For this purpose,
the information existing in NUDAPT
for the city of Houston and surrounding
areas was analyzed. The following
gridded
urban
morphological
parameters of a region that covers
5242 km 2 were considered as inputs for
the urban parameterizations: urban
fraction, building height histograms,
building plan area fraction, building
height weighted by footprint plan area,
and building surface area to plan area
ratio (λ B ) . The λ B is defined as the sum
of building surface divided by the total
plan area of the study location. All these
parameters with a resolution of 1
km 2 were introduced as new variables
in the input files of the WRF model. In
an urban grid point inside the covering
area, then, the urban schemes use
directly the information from the
NUDAPT database and not the
averaged properties defined in Table 2.
Some modifications were necessary in
the
BEP
and
BEP+BEM
parameterizations because initially these
schemes were built to work in terms of
urban classes and not reading urban
information point to point. The
numerical domains and physical
characteristics were the same as
explained in the above section 2.1. The
same two days were considered.
Figure 9. PBL height computed against
observed in Ellington and Southwest Airport for
the 25th and 31st August respectively.
3.2 Analysis of the results
3.2.1 Air temperature
To evaluate the impact of
gridded NUDAPT data on the 2-m air
temperature, the RMSEs (NUDAPT)
were computed and compared with the
previous RMSEs (urb_class). The term
urb_class hereafter refers to the
simulations that use the inputs
parameters of Table 2 for every urban
class. To quantify the comparison, the
following relative difference (∆T 2) was
calculated,
∆T 2 = 100 ×
RMSE (urb _ class ) − RMSE ( NUDAPT )
,
RMSE (urb _ class )
(3)
for every monitoring station, and day
simulated. A positive value means an
improvement in the results and a
negative value the opposite.
119
Chapter 3. A study of the UBL using different urban parameterizations.
Figure 10. Differences between the observed and computed 2-m air temperature for 25th August
2000.
120
Chapter 3. A study of the UBL using different urban parameterizations.
Figure 11. As in Figure 10 but for 31st August 2000.
121
Chapter 3. A study of the UBL using different urban parameterizations.
The (∆T 2) values obtained are
presented in Table 7, and the
differences between the observation and
model predictions are plotted in Figs. 10
and 11. In this comparison, 13 negative
against 23 positive values for the
relative difference (∆T 2) were obtained.
It is not easy to observe a clear tendency
looking at Table 7, except that the
BEP+BEM scheme seems to present a
greater sensibility to the urban
morphology parameters than the BEP
parameterization. This fact can be
understood because the BEP+BEM
scheme computes the AH released into
the atmosphere to maintain the indoor
temperature in a range of comfort, and
this heat flux is strongly dependent on
the urban fraction and the dimensions of
the buildings. On the other hand, in the
BEP scheme the indoor surface-wall
temperature is fixed constant during all
the simulation and no anthropogenic
heat flux is directly ejected into the
atmosphere. It is important to mention
that after 1900 LST and during all the
night, the sign of the air temperature
difference between the BEP+BEM
(NUDAPT) and BEP+BEM (urb_class)
simulations is kept constant (positive or
negative) in most of the stations, which
would mean that the AH released during
all the day (that in this case has a strong
dependence on the urban geometry) is
an important component of the UHI
phenomenon. This behaviour is less
significant when the BEP (NUDAPT)
and BEP (urb_class) simulations are
compared, given that the AH is not
taken into account. Based on these
results, it can be stated that for the case
studied, the urban fraction, the urban
materials properties, and the AH are the
most important factors that contribute to
the UHI. The urban geometry is
secondary in the formation of the UHI
and becomes relevant only because it
affects the AH. In Fig. 12 the 2-m air
temperature differences (T2(NUDAPT)T2(urb_class)) for the BEP+BEM
parameterization are plotted for the two
days simulated at 0300 LST, when the
air temperature differences are the most
important. These differences are smaller
with the BEP scheme (not shown),
especially for the 25th of August (the
coldest day), which confirm that the
urban geometry plays a secondary role
in the formation of the UHI. On the 31st
of August, the surface air temperature is
slightly overestimated during the night
when the NUDAPT information is used
point-to-point with the BEP+BEM
scheme (see Fig. 11). More simulations
could help to study the impact of using
a different target temperature, 26 ºC for
example, or a wider range of comfort.
Without a larger number of monitoring
stations well-distributed over the city
and without more days simulated, it is
difficult
to
extract
definitive
conclusions from this comparison.
However, it is possible to say that the
BEP scheme is less sensitive to the
urban geometry for the surface air
temperature predictions than the
BEP+BEM parameterization because
the AH is strongly dependent of the
urban morphology of the city. This is
because, in a certain sense, the AH
resulting from air conditioning is
proportional to the number of floors.
31-Aug-00
Monitoring
25-Aug-00
stations
∆T 2(%) ∆T 2(%)
C01
C55
C81
C146
Urban
Parameteriza
tion
0.0836
2.2184
BEP
13.8690
8.5102
BEP+BEM
12.0632
9.6690
BEP
23.8582
10.5053
BEP+BEM
11.1394
13.0313
BEP
30.2381
-0.9771
BEP+BEM
-4.3597
21.8200
BEP
122
Chapter 3. A study of the UBL using different urban parameterizations.
-4.0040
C167
C169
C404
C409
C603
21.1580
BEP+BEM
6.9939
9.8589
BEP
14.8666
-8.9933
BEP+BEM
5.3486
12.9915
BEP
-6.2806
-54.9167
BEP+BEM
6.1292
1.4235
BEP
12.0630
9.5851
BEP+BEM
-3.7719
-1.3133
BEP
-4.4652
4.7144
BEP+BEM
-21.9458
-42.2202
BEP
-43.8009
-52.9864
BEP+BEM
simulated colder temperature towards
those where it simulated hotter
temperatures means that the use of
realistic NUDAPT data for the
morphology of the city of Houston,
increased the convergence of the flow
above the hotter downtown area, at least
for the days simulated.
Table 7. Relative difference errors (∆T 2)
obtained for every monitoring station and day
analyzed. The relative difference errors are
derived using the following equation:
∆T 2 = 100 ×
RMSE (urb _ class) − RMSE ( NUDAPT )
RMSE (urb _ class)
3.2.2 Wind field
In Fig. 13, the wind speed at 10
m AGL is compared with the
observations for the first day simulated
in three monitoring stations (the rest of
the stations did not show remarkable
differences). It is difficult to highlight
some conclusion from this comparison
because the wind field is very
influenced by the presence of clouds,
and when the NUDAPT information is
used the position of the clouds is
modified with respect to the urb_class
simulation. Nevertheless, it shows the
importance of using detailed urban
morphology data for the cloud
prediction. When there were no clouds
the wind field was captured reasonably
well both days. In Fig. 12, together with
the air temperature differences, the wind
speed differences have been plotted
(NUDAPT
minus
urban_class
simulation). The fact that the vectors
(differences between the winds of
NUDAPT minus the one of urb_class)
point from the region where NUDAPT
Figure 12. a) 2-m air temperature
(T2(NUDAPT)-T2(urb_class)) and wind speed
(at 10 m AGL) differences at 0300 LST (26
August) obtained with the BEP+BEM scheme.
b) as in a) at 0300 LST (01 September).
123
Chapter 3. A study of the UBL using different urban parameterizations.
Figure 13. Time series of observed (OBS.) and
simulated (with BEP and BEP+BEM schemes)
horizontal winds at 10 m AGL for the 25th
August at three sites: a) C55 station, b) C409
station, and c) C603 station.
4. Waste heat emission and energy
consumption
In the last part of this research,
the total energy consumption (EC) has
been analyzed with the BEP+BEM
scheme. The value of instantaneous
energy
consumption
ec( x, y, t )
(in Wm−2 ) due to the space
cooling/heating is computed in this
parameterization at every urban grid
point, and the total consumption can be
computed as following:
EC =
T
0
ecdxdy dt ,
urban
domain
(4)
where (T ) is the period of simulation.
In BEP+BEM, all the buildings
are considered of the same type (only
the dimensions can be different), and all
the buildings in the domain are assumed
to run the AC. Taking this into account,
the EC was calculated considering the
high
resolution
urban
canopy
parameters data set (NUDAPT) and the
urban class classification for the two
days simulated. The EC (NUDAPT) for
the 25th of August over the whole
domain was of 197347 MWh, while for
EC (urb_class) it was 13.6% higher. For
the 31st of August, the EC (NUDAPT)
over the whole domain was 251419
MWh while for EC (urb_class) it was
8.9% higher. The differences in the EC
due to the different meteorological
conditions between the two selected
days were 21.5% and 17.2% for the
NUDAPT
and
urb_class
cases,
respectively. Heiple and Sailor (2008)
estimated the daily averaged energy
consumption from all sources (space
cooling, lighting and appliances, and
water heating) for the city of Houston
for the month of August with a topdown and bottom-up approaches as
108588 and 105869 MWh, respectively.
It is difficult to compare these results
since the urban area considered by
Heiple and Sailor (2008) is only a
fraction of the urban area considered in
our simulation domain. To get a more
meaningful comparison, the energy
consumption in the grid points
classified as commercial (based on the
NLCD database) was computed and
compared to those obtained by Heiple
and Sailor (2008) for the commercial
areas (see Table 8). The following
sources of uncertainty must be kept in
mind when comparing these results:
• The values computed by
Heiple and Sailor (2008)
account for the total
energy consumption (not
only AC, but also
lighting
and
water
heating), while those
computed
in
our
simulation are only due
to AC. Nationally, 45%
124
Chapter 3. A study of the UBL using different urban parameterizations.
•
•
of the annual energy
consumed in commercial
buildings is for space
cooling (or heating). If
this proportion is valid
also for the days
simulated, it can be
concluded that the model
overestimates the energy
consumption by a factor
1.7-2.2 for NUDAPT,
and a factor 3 to 4 for
urb_class. It is likely,
however, that for the
summer days considered
the fraction of energy
used for AC was larger
than 45% and the
overestimation lower.
It is assumed that the
points
classified
as
commercial in the NLCD
data belong all to the city
of Houston and that they
overlap with the points
considered
as
commercial in the work
of Heiple and Sailor
(2008).
With
the
information
in
our
possession it is not
possible to verify this
assumption. By a simple
visual analysis of a map
of the city it can be seen
that,
although
the
majority of commercial
areas are located within
the city limits, there are
commercial
points
outside, in particular in
the southern part of the
domain (e. g., Galveston
area). This may partially
explain the larger energy
consumption obtained by
the model.
The values computed by
Heiple and Sailor (2008)
are daily averages based
on climatic data for the
month of August, while
our simulations were
done for two specific
days (25th and 31st of
August
2000).
The
st
second of them (31 ), in
particular,
was
significantly
hotter
(maximum temperature
41 ºC) than the monthly
average values (average
maximum temperature
for the month of August
in Houston was of 33 ºC
for this year). It is likely
then that the energy
consumption for AC was
higher than the monthly
average values.
Daily
energy
consumption
in MWh
25th August
31st August
NUDAPT
Urb_class
34946
66790
45058
80660
Top
down
45853
Table 8. Daily energy consumption computed by
the model and estimated by Heiple and Sailor
(2008) with the top down and bottom up
techniques.
Indeed, some refinements must
be
done
in
the
BEP+BEM
parameterization
to
correct
this
overestimation
in
the
energy
consumption;
for
example,
by
considering different typologies of
buildings that may use different types of
AC (or no AC at all). For this work,
state-of-the-art building-energy models
will be an important tool to improve the
performances of BEP+BEM by mean of
off-line simulations. However, taking
into account also the limitations of an
urban canopy parameterization (some of
them mentioned previously in this
section) with respect to the top down
and bottom up methodologies, we think
that the results are a good starting point
to create a tool that can give reasonable
125
Bottom
up
45483
Chapter 3. A study of the UBL using different urban parameterizations.
estimates of energy consumption
without the need of very detailed
information, which is often difficult to
obtain. Another interesting conclusion
that can be derived from these results is
that the total energy consumption is
very sensitive to the urban database
used. Detailed information on the urban
morphology is necessary (like those in
NUDAPT) to get a realistic estimate of
the energy consumption.
Figure 14. a) 2-m air temperature differences
(T2(AH)-T2(no AH)) at 0300 LST (26 August)
obtained with the BEP+BEM (urb_class)
simulation. The wind speed (AH) at 10 m AGL
is showed. b) As in a) with the BEP+BEM
(NUDAPT) simulation.
Finally, the impact of the AH in
the air temperature has been addressed.
For this purpose, four new simulations
with the BEP+BEM scheme were
performed, where the AH coming from
the AC systems was not ejected into the
air. In the Figs. 14-15, the 2-m air
temperature differences have been
plotted for the two different urban
configurations, NUDAPT and urb_class.
The patterns of the simulated
temperature fields present significant
differences, showing the importance of
the meteorological conditions and the
urban morphology in the quantification
of the AH in a city. It is interesting to
observe (see Figs. 14b and 15b) that the
UHI strength reflects closely the urban
fraction. Similarly as in other studies
(Ohashi et al. 2006), the waste heat
increased the air temperature by 0.5-2
ºC depending on the location inside the
city
and
the
day
considered
(meteorological conditions).
Figure 15. a) 2-m air temperature differences
(T2(AH)-T2(no AH)) at 0300 LST (01
September) obtained with the BEP+BEM
(urb_class) simulation. The wind speed (AH) at
10 m AGL is showed. b) As in a) with the
BEP+BEM (NUDAPT) simulation.
5. Conclusions
In this work, four urban canopy
schemes coupled to the WRF model
have been compared over the city of
Houston. The first scheme is a simple
BULK scheme tested successfully on
several occasions. The second scheme is
126
Chapter 3. A study of the UBL using different urban parameterizations.
a single-layer urban model (UCM)
coupled to the WRF model and
validated in different episodes. The
third parameterization is a multilayer
urban model (BEP) recently coupled to
the WRF model, and finally, the last
scheme is a multilayer urban
parameterization (BEP+BEM) with a
building energy model integrated from
WRF V3.2 model. In the first part of the
research, an up-to-date urban land cover
dataset was used defining three different
urban classes for the built-up areas in
the numerical domain. Good results for
the surface air temperature were
obtained with the four schemes for
the 25th of August. However, for the 31st
of
August,
the
UCM+AH
parameterization was not able to capture
satisfactorily the temporal evolution of
the surface air temperature during the
afternoon and the night.
In a second part of the work, a
high-resolution gridded UCP database
was utilized (instead of urban classes
derived from NLCD), and the
simulations with the BEP and
BEP+BEM schemes were repeated
again. Results show that the BEP
scheme is less sensitive to the urban
morphology parameters for the air
temperature
forecast
than
the
BEP+BEM scheme. It is important to
emphasize that the AH (computed in
BEP+BEM) is strongly dependent on
the urban geometry of the city. So if the
goal of the research is to quantify and
study the impact of the AH, and
evaluate strategies for the reduction of
the EC, the use of a high-resolution
urban land cover database for the
investigated city becomes necessary and
the BEP+BEM scheme recommended.
On the other hand, if the purpose of the
research is real time weather prediction,
the simple bulk scheme would be
sufficient, and consequently a highresolution urban land cover database
would not be necessary in mesoscale
simulations. Indeed, these are the
conclusions that can be derived for this
specific case being aware that a detailed
study on cloud prediction must be
carried out. To generalize them, it
would be necessary to also investigate
long periods, different cities and/or have
a larger number of monitoring stations
well-distributed over the domain.
Finally, in the last part of the
work, the EC over the city of Houston is
quantified with the BEP+BEM scheme,
and a reasonable value is obtained
taking into account the limitations of an
urban canopy parameterization for these
kinds of studies. Differences up to 20%
in the EC were obtained due to the
different meteorological conditions
existing between the two selected days.
The impact of the AH on the air
temperature is calculated and its
distribution examined. The urban
geometry played an important role in
the EC and in the AH spatial
distribution, and consequently the
importance of using a high-resolution
urban land cover database becomes
evident. The AH increased the air
temperature up to 2 ºC in some places
during the night. Different episodes and
cities should be simulated with the
different urban canopy models to
understand better their impact on the
complex UBL process over the cities (e.
g., cloud formation, air pollution
dispersion, etc.). Moreover, the new
BEP+BEM parameterization coupled to
the WRF mesoscale model can help to
evaluate mitigation strategies of the
UHI phenomenon and to test new
energy
consumption
reduction
programmes in the cities.
Acknowledgements
We thank CIEMAT for the doctoral
fellowships
held
by
Francisco
Salamanca. We also thank Jason Ching
for providing the Urban Canopy
Parameters for the city of Houston,
Michael Duda for his help in the
implementation of the high-resolution
urban data set in the preprocessor of
127
Chapter 3. A study of the UBL using different urban parameterizations.
WRF, and Angelines Alberto Morillas
for her help in the set up of the
computer cluster of CIEMAT where
some simulations were performed.
Finally, we thank Cody Phillips for
editing the manuscript. This work was
funded by the Ministry of Environment
of Spain.
References
Bougeault, P., and P. Lacarrère, 1989:
Parameterization of orographyinduced turbulence in a mesobetascale model. Mon. Wea. Rev., 117,
1872-1890.
Burian, Steven J., W. S. Han, S. P.
Velugubantla, and S. R. K.
Maddula, 2003: Development of
Gridded Fields of Urban Canopy
Parameters
for
Models3/CMAQ/MM5, November 25.
Burian, Steven J., and W. S. Han, 2003:
Morphological Analyses using 3D
Building Databases: Houston,
Texas. LA-UR-03-8633, Los
Alamos National Laboratory.
Chen, F., and Coauthors, 2010: The
integrated WRF/urban modelling
system: development, evaluation,
and
applications
to
urban
environmental
problems.
International
Journal
of
Climatology,
Doi:
10.1002/joc.2158.
Chen, F., and J. Dudhia, 2001: Coupling
an
advanced
landsurface/hydrology model with
the Penn State/NCAR MM5
modeling system. Part I: Model
implementation and sensitivity.
Mon. Wea. Rev., 129, 569-585.
Ching, J., and Coauthors, 2009:
National Urban Database and
Access Portal Tool. BAMS, 90,
1157-1168.
Dudhia, J., 1989: Numerical study of
convection observed during the
winter monsoon experiment using
a mesoscale two-dimensional
model. J. Atmos. Sci., 46, 30773107.
Ek, M. B., K. E. Mitchell, Y. Lin, E.
Rogers, P. Grunmann, V. Koren,
G. Gayno, and J. D. Tarpley
2003: Implementation of Noah
land surface model advances in
the National Centers for
Environmental
Prediction
operational
mesoscale
Eta
model, J. Geophys.
Res.,
108(D22),
8851,
doi:10.1029/2002JD003296.
Grimmond, C.S.B., and Coauthors,
2009: The International Urban
Energy
Balance
Models
Comparison Project: First results
from Phase 1. J. Appl. Meteorol.
Climatol.
Doi:
10.1175/2010.JAMC2354.1.
Heiple, S., and David J. Sailor, 2008:
Using building energy simulation
and
geospatial
modelling
techniques to determine high
resolution building sector energy
consumption profiles. Energy and
Buildings, 40, 1426-1436.
Hong, S.-Y., J. Dudhia, and S.-H. Chen,
2004: A Revised Approach to Ice
Microphysical Processes for the
Bulk Parameterization of Clouds
and
Precipitation.
Monthly
Weather Review, 132, 103-120.
Janjic, Z. I., 1994: The step-mountain
eta
coordinate:
Further
development of the convection,
viscous sublayer, and turbulent
closure schemes. Mon. Wea.
Rev., 122, 927–945.
Kanda, M., T. Kawai, M. Kanega, R.
Moriwaki,
K.
Narita,
A.
Hagishima, 2005: A simple energy
balance model for regular building
arrays. Boundary-Layer Meteorol.,
116, 423-443.
Kikegawa Y., Y. Genchi, H. Yoshikado,
and
H.
Kondo,
2003:
Development of a numerical
simulation
system
toward
comprehensive assessments of
128
Chapter 3. A study of the UBL using different urban parameterizations.
urban warming countermeasures
including their impacts upon the
urban buildings energy demands.
Applied Energy, 76, 449-466.
Kondo, H., Y. Genchi, Y. Kikegawa, Y.
Ohashi, H. Yoshikado, and H.
Komiyama, 2005: Development of
a multilayer urban canopy model
for the analysis of energy
consumption in a big city:
structure of the urban canopy
model and its basic performance.
Boundary-Layer Meteorol., 116,
395-421.
Kusaka, H., H. Kondo, Y. Kikegawa,
and F. Kimura, 2001: A simple
single-layer urban canopy model
for
atmospheric
models:
comparison with multi-layer and
slab models. Boundary-layer
Meteorol., 101, 329-358.
Kusaka, H., and F. Kimura, 2004:
Coupling a single-layer urban
canopy model with a simple
atmospheric model: Impact on
urban heat island simulation for an
idealized case. J. Meteor. Soc.
Japan, 82, 67-80.
Lin, C-Y, F. Chen, J.C. Huang, W-C.
Chen, Y.-A. Liou, W.-N. Chen,
and Shaw-C. Liu, 2008: Urban
Heat Island effect and its impact
on boundary layer development
and land-sea circulation over
northern Taiwan. Atmospheric
Environment, 42, 5635-5649.
Liu, Y., F. Chen, T. Warner, and J.
Basara, 2006: Verification of a
Mesoscale Data-Assimilation and
Forecasting System for the
Oklahoma City Area During the
Joint Urban 2003 Field Project. J.
Appli. Meteorol., 45, 912–929.
Lo, Jeff C. F., Alexis K. H. Lau, F.
Chen, Jimmy C. H. Fung, Kenneth
K. M. Leung, 2006: Urban
Modification in a Mesoscale
Model and the Effects on the
Local Circulation in the Pearl
River Delta Region. J. Appl.
Meteorol. Climatol., 46, 457-476.
Martilli A., A. Clappier, and M. W.
Rotach, 2002: An urban surface
exchange parameterization for
mesoscale models. BoundaryLayer Meteorol., 104, 261-304.
Masson, V., 2000: A physically-based
scheme for the urban energy
budget in atmospheric models.
Boundary-Layer Meteorol., 94,
357-397.
Miao, S., F. Chen, Margaret A. LeMone,
M. Tewari, Q. Li, and Y. Wang,
2009: An Observational and
Modelling
Study
of
Characteristics of Urban Heat
Island and Boundary Layer
Structures in Beijing. J. Appl.
Meteorol. Climatol., 48, 484-501.
Mlawer, E. J., S. J. Taubman, P. D.
Brown, M. J. Iacono, and S. A.
Clough, 1997: Radiative transfer
for inhomogeneous atmospheres:
RRTM, a validated correlated-k
model for the longwave. J.
Geophys. Res., 102 (D14), 1666316682.
Ohashi, Y., Y. Genchi, H. Kondo, Y.
Kikegawa, H. Yoshikado, and Y.
Hirano, 2006: Influence of AirConditioning Waste Heat on Air
Temperature in Tokyo during
Summer: Numerical Experiments
Using and Urban Canopy Model
Coupled with a Building Energy
Model. J. Appl. Meteorol.
Climatol, 46, 66-81.
Salamanca F., A. Krpo, A. Martilli, and
A. Clappier, 2010: A new
Building Energy Model coupled
with
an
Urban
Canopy
Parameterization for urban climate
simulations—Part I. Formulation,
verification and a sensitive
analysis of the model. Theoretical
and Applied Climatology, 99: 331344.
Salamanca, F., and A. Martilli, 2010: A
new Building Energy Model
129
Chapter 3. A study of the UBL using different urban parameterizations.
coupled with an Urban Canopy
Parameterization for urban climate
simulations—Part II. Validation
with one dimension off-line
simulations.
Theoretical
and
Applied Climatology, 99: 345-356.
Skamarock, W. C., J. B. Klemp, J.
Dudhia, D. O. Gill, D. M. Barker,
M. Duda, X.-Y. Huang, W. Wang
and
J.
G.
Powers, NCAR
Technical
Note,
2008:
A
description of the Advanced
Research WRF version 3. NCAR
Tech. Note TN-475+STR, 125 pp.
130
Capítulo 3
131
Capítulo 3
3.5 Estudio numérico de la capa límite urbana sobre la ciudad de Madrid durante la
campaña DESIREX (2008) con WRF y evaluación de diferentes estrategias de
mitigación de la isla de calor urbana.
Salamanca, F., A. Martilli, and C. Yagüe, 2010c: A numerical study of the urban boundary layer over
Madrid during the DESIREX (2008) campaign with WRF and an evaluation of simple mitigation
strategies of the UHI. Atmospheric Environment (submitted).
En esta última sección se ha simulado con la nueva parametrización BEP+BEM la ciudad de
Madrid. El período simulado escogido tuvo buenas condiciones sinópticas que favorecen la
formación de la isla de calor urbana y coincidió con la campaña meteorológica DESIREX que
tuvo lugar en el verano del 2008. La campaña tuvo como principal objetivo el estudio de la
isla de calor, haciendo uso de estaciones meteorológicas e imágenes termográficas. En una
primera parte del trabajo se validó la simulación y se estudió el impacto del calor
antropogénico proveniente de los sistemas de aire acondicionado. En una segunda parte del
trabajo se analizaron estrategias de mitigación tanto de la isla de calor como del consumo
energético. Las estrategias analizadas fueron tres: aumento del albedo, uso de materiales
aislantes en los techos y eliminación del calor antropogénico proveniente de los sistemas de
aire acondicionado. Los resultados de las tres estrategias fueron analizadas separadamente y
en conjunto.
132
Chapter 3. A numerical study of the UBL over Madrid during the DESIREX (2008) campaign.
A numerical study of the urban boundary layer over Madrid during
the DESIREX (2008) campaign with WRF and an evaluation of simple
mitigation strategies of the UHI.
Francisco Salamanca 1 · Alberto Martilli 1 · Carlos Yagüe 2
1
CIEMAT (Research Centre for Energy, Environment and Technology), Madrid, Spain
2
Dept. de Geofísica y Meteorología, Universidad Complutense de Madrid, Spain
ABSTRACT
Nowadays, mesoscale meteorological models coupled to Urban Canopy Parameterizations (UCP) can be
used to complement and interpret the information gathered from intensive meteorological campaigns on
the behaviour of the Urban Boundary Layer (UBL). Moreover, the impact of the air conditioning (AC)
systems on the air temperature, the relationships existing between the energy consumption (EC) and the
meteorological conditions, and the evaluation of strategies to mitigate the Urban Heat Island (UHI)
phenomenon can be evaluated using complex UCP. In this work, a new UCP implemented in the Weather
and Research Forecasting model (WRF, version 3.2) has been tested over the city of Madrid for the
DESIREX campaign that took place in summer of 2008 and focused on Urban Heat Island (UHI) and
Urban Thermography (UT) monitoring and assessment. Two selected days, with a high UHI intensity,
th
st
were simulated (June 30 and July 1 ), and numerical results of various physical variables were
compared against measurements showing a satisfactory performance of the model. The impact of the AC
systems on the air temperature reached up to 1.5-2 ºC in some dense urban places, and the EC was
evaluated for the simulated period. Effects of modifications in the roof albedo and building material
properties would reduce the total EC by 5 % and 3 % respectively, affecting the intensity of the UHI.
______________________________________________________________________
Atmospheric Environment,
(Submitted)
Corresponding author address: F. Salamanca, Research Centre for Energy,
Environment and Technology (CIEMAT), Avenida Complutense 22, 28040 Madrid,
Spain.
E-mail: [email protected]
133
Chapter 3. A numerical study of the UBL over Madrid during the DESIREX (2008) campaign.
1. Introduction
In the last decades, the
development of mesoscale models
(MMs) accompanied by an increasing
number of meteorological campaigns
have contributed notably to a better
understanding of the complex processes
(Oke, 2002; Arnfield, 2003) that occur
in the urban boundary layer (UBL). At
the end of the XX century (and
beginning of the XXI) the first urban
canopy parameterizations (UCPs) to
account for the mechanical and thermal
impact of cities in MMs (e.g., Masson
2000, Kusaka et al. 2001, Martilli et al.
2002, Kanda et al. 2005) appeared.
These important efforts provided a more
realistic physical description of the
urban areas, and contributed to better
understand their impact on the lower
atmosphere. Ever since, a big number of
urban areas have been investigated (e.g.,
Martilli et al. 2003, Lo et al. 2006,
Lemonsu et al. 2009, etc.) showing the
need of considering the UCPs to study
UBL processes.
A fundamental fact contributing to
increase the intensity of the urban heat
island (UHI) phenomenon in big cities
is the anthropogenic heat (AH)
generated by human activities, besides
the altered radiation and energy budget
in urban areas, mainly as a result of
replacing natural surfaces by buildings
and pavements with different thermal
inertia (Landsberg, 1981) . UHI
indicates the higher temperature
existing within cities compared to the
surrounding rural areas and its intensity
is usually defined as the difference in
temperature between two observatories,
one urban and other rural. The sources
of AH (sensible/latent) released into the
atmosphere are associated with the
energy consumption and can be
classified in three principal sectors:
industry, buildings and transportation
(see Sailor, 2010). It is known that
fundamental contributions of AH in the
commercial/residential areas are due to
energy use in buildings for heating,
ventilation, and air conditioning (AC)
systems.
In
the
first
urban
parameterizations (e.g. Masson 2000,
and Martilli et al. 2002) the AH was not
computed
explicitly and
coarse
simplifications (for example keeping the
indoor surface temperature of the
buildings constant in time) were
considered
in
the
simulations.
Alternatively, a source term of heat
estimated from energy consumption
databases was added in the atmospheric
equations (Kusaka et al. 2001). One of
the first studies where the AH (ejected
by the AC systems into the atmosphere)
was evaluated through an urban
parameterization, is the work of
Kikegawa et al. (2003), who integrated
a simple building energy model (BEM)
in an UCP. In this work, the relationship
between the outdoor temperature and
energy consumption was studied and
the AC systems were responsible of an
increase up to 1 º C (in average) of the
outdoor temperature. The coupling
between a BEM and an UCP offers a
spread of new possibilities when they
are integrated in an atmospheric model.
The effect of the AC systems in the air
temperature, the quantification of the
AH and its impact on the UHI
phenomenon, and the evaluation of
energy
consumption
mitigation
strategies are some examples of the
features that can be investigated thanks
to this linkage. However, a handicap
appears when an UCP+BEM scheme is
used, given that high resolution urban
canopy parameters data sets are needed
to obtain good estimations of the AH
fluxes.
In the last years, an important
effort is being carried out in the USA
(see Ching et al. 2009) to provide
detailed urban canopy parameters to
improve the parameterization of UBL
processes. Continuing in this line,
Salamanca et al. (2010a) developed a
new BEM accounts for: the diffusion of
134
Chapter 3. A numerical study of the UBL over Madrid during the DESIREX (2008) campaign.
heat through walls, roofs, and floors;
natural ventilation; the radiation
exchanged between indoor surfaces; the
generation of heat due to occupants and
equipments; and the consumption of
energy due to air conditioning systems.
BEM was coupled to a building canopy
parameterization (BEP, building effect
parameterization of Martilli et al. 2002),
and tested off-line over the city of Basel
(Salamanca and Martilli, 2010). It has
been integrated in the public WRF
(Weather Research and Forecasting)
model (Skamarock et al. 2005) V3.2
release. The new BEP+BEM scheme
was tested against measurements over
the city of Houston (TEXAS, USA),
and was able to produce reasonably
estimations of energy consumption (EC)
when high resolution urban canopy
parameters were used (see Salamanca et
al. 2010b).
In the summer of 2008, the
DESIREX (Dual-use European Security
IR
Experiment)
campaign
(see
DESIREX final report, Sobrino et al.
2009), run by ESA (European Space
Agency) and coordinated by Dr. J. A.
Sobrino of Valencia University, was
carried out in the city of Madrid (Spain).
An important aim of the campaign was
to study the UHI in support to
mitigation strategies and urban energy
efficiency policies. In this paper, thanks
to the availability of the measured data,
the BEP+BEM scheme, implemented in
the WRF V3.2, has been tested for
Madrid. The UHI in Madrid is known
since decades (for example, see Yagüe
et al. 1991). However, it has never been
investigated numerically with a detailed
urban canopy parameterization so far.
Meteorological modelling studies over
the region have been carried out in the
last years, but always focussing on air
pollution episodes (e.g., Palacios et al.
2002). The aim of this work is twofold:
on one hand, to evaluate the ability of
the new BEP+BEM scheme to describe
the UBL processes over the city of
Madrid in summer conditions, and on
the other hand, to evaluate some simple
energy
consumption
mitigation
strategies and its relationship with the
atmospheric variables.
In section 2, a description of the
studied
area
and
experimental
framework are displayed. In the first
part of section 3, the set up of the
simulations is presented, and in the
second part numerical results against
measurements are analysed. In section 4,
the UHI and numerical sensitive
experiments
relative
to
energy
consumption mitigation strategies are
presented, and finally in section 5,
conclusions and future work are pointed
out.
2. DESIREX campaign
2.1 Madrid area
Madrid (see fig. 1) is the capital
and largest city of Spain. It is located in
the central part of the Iberian Peninsula
(40º 25’N, 3º 41’W) in a relatively flat
area about 50 km south-east of the
Spanish Central Ridge. A small valley
crosses the city from north to south
acting frequently as channel for air
masses from the cooler northern rural
areas. The maximum city height is
located in the northern limit of the city
reaching roughly 700 m above sea level
(a.s.l.), while the lower occurs at the
Manzanares River, around 550 m. The
population of the city is roughly 3.2
million and has not had important
changes from 1970, while the estimated
population of the metropolitan area is
5.1 million, with outstanding increases
in the last decades. The urban areas
span a total of 607 km 2 . The
coordinates of the centre of the tested
area are 40º 23’ N and 3º 43’ W.
Madrid has a climate with cool winters
and temperatures that sometimes drop
below 0 º C , and dry hot summers
(mean maximum temperature 31.2 º C
in July) with temperatures that can
reach up to 40 º C in July and August
135
Chapter 3. A numerical study of the UBL over Madrid during the DESIREX (2008) campaign.
(more details can be found in Sobrino et
al. 2009).
Measu
remen
ts
ID
Radio
soundi
ng
RS1
Captiv
e
balloo
n
CB1
CUF
Nuevos
Minister
ios
40º 26' 3 º 41' 686
45.6"
27.6"
Meteo
rologi
cal
netwo
rk
MN
1
MN
2
MN
3
CUF
40º 27'
MN
4
DUF
MN
5
___
MN
6
DUF
MN
7
DUF
SN1
CUF
SN2
DUF
SN3
___
AS1
CUF
AS2
OU
A
DUF
CIEMA
T
Fuencarr
al
Junta
Mpal.
Moratal
az
Junta
Mpal.
Villaver
de
E.D.A.R
.La
China
Centro
Mpal.
De
Acústica
Junta
Mpal.
Hortalez
a
Pza. de
España
Pza. de
Castilla
Casa de
Campo
MadridParque
Retiro
Madrid
Barajas
MadridCuatro
vientos
Madrid
Getafe
MadridCiudad
Universi
taria
Madrid
Torrejón
de
Ardoz
Colmen
ar Viejo
Rugby
field
UAM
Firemen
Park
tower
Survei
llance
netwo
rk
AEM
ET
station
s
AS3
AS4
AS5
Fixed
masts
Urba
n
Clas
s
OU
A
OU
A
___
OU
A
DUF
AS6
OU
A
AS7
FM1
OU
A
DUF
FM2
DUF
Site
AEMET
-Barajas
Latitud
e (N)
Longit
ude
(W)
Altitu
de
(m)
asl
40º 27' 3º 32' 582
15"
39"
3º 43'
680
40º 30' 3º 40' 729
06"
57"
40º 21' 3º 34' 687
34"
10"
40º 20' 3º 42' 594
58"
39"
40º 19' 3º 40' 566
26"
42"
40º 26' 3º 44' 587
43"
24"
40º 27' 3º 39' 704
49"
24"
40º 25'
26.94"
40º 28'
05.77"
40º 25'
12.35"
40º 24'
40"
3º 42'
43.52"
3º 41'
19.14"
3º 44'
57.22"
3º 40'
41"
637
729
645
667
40º 27' 3º 32' 582
15"
39"
40º 22' 3º 47' 687
40"
21"
40º 18' 3º 43' 617
00"
21"
40º 27' 3º 43' 664
10"
27"
40º 29' 3º 27' 611
00"
01"
40º 41'
55"
40° 32'
50.58"
3º 45' 1004
52"
3° 41' __
53.57"
40° 23' 3° 39' __
39.98"
13.11"
FM3
CUF
New
city Hall
40° 25' 3° 41' __
8.35"
31.88"
FM4
CUF
Printing
FM5
DUF
Urbanis
m
building
40° 24'
49.96"
40° 27'
36.31"
3° 42' __
19.87"
3° 40' __
19.86"
Table 1. Measurements sites and coordinates of
the DESIREX campaign activities (source
DESIREX final report).
2.2 Experimental framework
The DESIREX campaign took
place in Summer 2008 from June
23rd to July 6th on the city of Madrid
and surroundings areas. Significant
measurements activities were carried
out involving an important number of
researchers and institutions. In this
section, only the part used in this work
relative to atmospheric measurements
are described (for more details see the
final report of the campaign, Sobrino et
al. 2009). Figure 2 shows the location of
the measurements deployed during the
campaign. In Table 1, the corresponding
measurement sites and coordinates are
indicated. The AS6, AS7 and FM1 sites
are out of the Figure 2 map. Two
balloons were released every day, at
noon ( ≈ 1400LST ) and at midnight
( ≈ 0200LST ) by the Meteorological
State Agency (AEMET), using the
Vaisala RS92-SGP system at the
AEMET-Barajas site (12km NE from
city centre). The Vaisala RS92-SGP
system was used to retrieve pressure, air
temperature, relative humidity and wind
magnitudes in each sounding.
Various soundings with tethered
balloon (up to 1000m) were taken at the
city centre (Nuevos Ministerios site)
providing the main meteorological
characteristics
and
temporal
development of the planetary boundary
layer structure. These soundings
(Atmospheric Instrumentation Research
- AIR, 1986) measured the atmospheric
pressure, dry and wet temperature, wind
velocity, and wind direction. Due to
windy conditions some of the launches
were aborted.
136
Chapter 3. A numerical study of the UBL over Madrid during the DESIREX (2008) campaign.
The city council of Madrid has
developed its own meteorological
network. The stations were located
covering an extend area. Wind
directions, wind speed, solar radiation,
relative humidity, atmospheric pressure,
air temperature and precipitations were
measured during the campaign. The city
council of Madrid and the Region of
Madrid operate also another network of
air quality stations in the area of study.
These stations (Surveillance network in
Table 1) measured wind speed, wind
direction, air temperature, pressure,
solar radiation and precipitation. Other
meteorological stations belonging to
AEMET were also included in the
campaign.
3. WRF simulations
3.1 Numerical domain and set up of the
simulations
Two summer days (high pressure period)
have been analyzed, June 30 th and July
1st 2008 (days 182 and 183
respectively). It is a typical summer
synoptic situation, showing stability at
high levels and very weak horizontal
pressure gradient, so clear skies with
slow winds are present along the two
days studied (not shown). This kind of
synoptic situation favours, during the
night, the developing of surface-based
thermal inversions, with strong stability;
in these situations, the topographic
features cause the development of
circulations that are far from being nondivergent over Madrid and convergence
katabatic flows can be developed over
the city (Terradellas and Cano, 2007).
The 60-h simulations begin at
1800 UTC (LST = UTC+2h) 29 th June
and finish 0600 UTC 2nd July. Two
different
simulations
with
the
BEP+BEM scheme (one considering the
effect of the air conditioning systems,
and the other not) were performed using
the non-hydrostatic version of WRF
(Skamarock et al. 2005) version V3.2,
coupled to the Noah land surface model
(Chen and Dudhia, 2001) for the
vegetated part. The horizontal domain
was composed of five two way nested
domains
with
100 × 100 , 174 × 156 , 219 × 186 ,
216 × 198 , and 240 × 270 grid points,
and a grid spacing of 27 , 9 , 3 , 1, and
0.333 km , respectively. The 60-h
simulations were conducted with the
Figure 1. Map of the SW of Europe with Madrid
in the middle of the Iberian Peninsula (picture
obtained from Google Earth).
initial and boundary conditions obtained
from the operational National Centre for
Environmental Prediction (NCEP) with
a grid resolution of 40km and a time
resolution of 3 h. To take full advantage
of the urban parameterizations, a
vertical resolution of 51 sigma levels 1
was used (33 levels in the lowest
1.5 km with the domain top
over ≈ 20km ). The Bougeault and
Lacarrère (1989) planetary boundary
1
Full sigma levels = 1., 0.998743415, 0.99748677,
0.996230185, 0.9949736, 0.993716955, 0.992334723,
0.990814209, 0.989141703, 0.987301886, 0.98527813,
0.983051956, 0.980603218, 0.977909565, 0.974946558,
0.971687257, 0.968101978, 0.964158237, 0.959820092,
0.955048144, 0.949799001, 0.94402492, 0.937673509,
0.930686891, 0.923001587, 0.914547801,
0.905248582, 0.895019472, 0.883767486, 0.871390283,
0.857775331, 0.842798889, 0.826324821, 0.80820334,
0.788269699, 0.7663427, 0.742223024, 0.715691328,
0.68650645, 0.65440315, 0.619089544, 0.580244482,
0.537514985, 0.49051252, 0.438809812, 0.381936818,
0.319376528, 0.250560224, 0.17486228, 0.0915945247,
and 0.0000. The vertical coordinate σ is defined
as ( p − ptop ) /( p surf − ptop ) , where p is the pressure at
each corresponding level, p
surf
is surface pressure, and
ptop is a constant pressure at model top.
137
Chapter 3. A numerical study of the UBL over Madrid during the DESIREX (2008) campaign.
layer (PBL) scheme (one-and-a-halforder closure) was used.
BEP+BEM scheme is a building
energy model linked to a multilayer
urban canopy parameterization (BEP,
Martilli et al. 2002) that takes into
account the exchange of energy
between the buildings and the outdoor
atmosphere. BEP takes into account the
effects of the building vertical and
horizontal surfaces on momentum (drag
force approach), turbulent kinetic
energy, and potential temperature. The
radiation at walls and roads considers
shadowing, reflections, and trapping of
shortwave and longwave radiation in
street canyons. In BEM, the evolution
of the indoor air temperature is
computed as a function of the energy
production and consumption in the
buildings (separately for every floor) as
well as the impact of the air
conditioning system. The AH released
into the urban atmosphere, to maintain
the indoor temperature in a comfort
range, is computed through an AC
model (more details in Salamanca et al.
2010a). To study the effect of this AH,
two simulations were carried out: in the
first (hereafter this case is referred as
BEP+BEM(AH)) the AH was ejected
directly into the atmosphere, and in the
second one it was not (case
BEP+BEM(no AH)), for example it was
placed elsewhere such as into sewage or
soil. Comparing the two simulations the
effects of the AH due to the AC systems
can be evaluated.
Figure 2. Location of the atmospheric
measurements obtained in DESIREX campaign
(picture obtained from Google Earth). See
Table 1 for icons identification. The Earth
Coordinates for the MN2 and AS4 sites are
indicated.
Three different urban classes
were defined in the inner domain (see
Figure 3) based on the CORINE land
cover (the inventory was of the year
2000) database from the European
Environment
Agency
(http://www.eea.europa.eu):
“continuous urban fabric” (CUF),
“discontinuous urban fabric” (DUF),
and “other urban areas” (OUA) that
represent a collection of points that in
the CORINE database were classified as
“Industrial or commercial units”, “Road
and rail networks and associated land”,
and “Airports”. The CUF class covered
a surface of 102.42km 2 , the DUF class
of 423.67km 2 , and finally the OUA
class of 177.89km 2 . For the non urban
part (see Figure 3), an up to date
MODIS-based land use classification
was used.
The urban parameters and
thermal properties defined for every
urban class can be seen in Table 2 and 3
respectively.
For
the
correct
performance of the BEP+BEM scheme
(see Salamanca et al. 2010a) extra
inputs are required. In the three urban
classes, the coverage area fraction of
windows in the walls was fixed to 0.2,
the number of occupants to 0.02
138
Chapter 3. A numerical study of the UBL over Madrid during the DESIREX (2008) campaign.
person/m 2 of
floor, and a thermal
efficiency of the total heat exchanger to
0.75 (this value means that the indoor
air of buildings is replaced by fresh air
six times a day that corresponds to
standard situations). For the AC model,
additional parameters are needed and
were fixed to the following values:
target internal temperature 25 º C ,
amplitude of range of comfort 1 º C , and
COP (Coefficient of Performance of the
AC systems) 3.5. The peak heat flux
generated by equipments was 36
Wm −2 of floor for OUA areas, and 20
Wm −2 for the rest of urban classes. The
considered values for the sensible heat
generated by equipments are similar to
others estimations based on temporal
variations
of
electric
power
consumption (Kikegawa et al. 2003).
Parameter
CUF
DUF
OUA
Intensity
(High
(Low
(Commercial
Residential)
Intensity
or Industrial)
Residential)
Building
plan
0.60
0.50
0.375
5
15
10
15
55
20
area
fraction
%
of
buildings of
5m
of
height
%
of
buildings of
10m
of
height
%
of
15
25
40
65
5
30
buildings of
15m
of
height
%
of
buildings of
20m
of
height
Table 2. Urban morphological parameters
considered for the three urban classes.
3.2 Analysis of the results
3.2.1 Air temperature
The
meteorological
air
temperature sensors were displayed at
different heights, often above roof,
depending of location and type of
station.
For
this
reason,
the
measurements
instead
of
being
compared with the standard modelled
2m air temperature were compared with
the first sigma level above ground
( ≈ 10 m) that was more representative.
The Root Mean Square Error (RMSE)
and Mean Bias (MB) results can be seen
in Table 4 for the two different
simulations.
Figure 3. Up to date MODIS-based land use
classification (20 + 3 urban categories) used in
the inner domain. In the figure, the 31, 32 and
33 (DUF, CUF, and OUA) values correspond to
points which the urban fraction is bigger than 0.
Continuous black lines correspond to height
a.s.l. (it can be seen the Central System
Mountains at the NW of the domain). The
location of the AS2, AS1, FM1, FM4, MN2,
MN4, MN7, and SN1 measurements sites is
indicated with green square marks.
It is known that the AH is bigger
in downtown cores (and consequently
its impact) than in residential areas
where the urban fraction is lower. This
fact can be seen observing the RMSE
for the air temperature obtained with the
two simulations BEP+BEM(AH) and
BEP+BEM(no AH). The stations MN2,
MN3, MN7, AS2, and AS7 are away of
the urban core and the two simulations
gave similar results. On the other hand,
in downtown sites as SN1, FM3, FM4,
etc., the RMSE is improved (for
example it was decreased up to 5 tenths
of degree in the FM3 site) when the AH
is ejected into the atmosphere. Other
stations, although further from the city
139
Chapter 3. A numerical study of the UBL over Madrid during the DESIREX (2008) campaign.
centre but with high population rates,
are also benefit from the AH inclusion
(MN4, MN5, SN2, SN3, AS4, FM2).
Figure 4. Temporal series of observed against
modelled air temperature at two different
stations: a) FM4, and c) FM1 sites. In b)
temporal evolution differences of the air
temperature between BEP+BEM(AH) against
BEP+BEM(no AH).
Figure 4 (a-c) shows temporal
(UTC) evolution of the air temperature
in an urban (FM4) and a rural (FM1)
site. FM4 is situated in the centre of the
city, in a very high density urban area
where the urban fraction is 100%. The
air temperature of BEP+BEM (AH) is
higher than for BEP+BEM (no AH))
(see Figures 4a-b) during all the period
simulated. The AH would be the
responsible of this increase in the air
temperature even if it does not increase
the maximum value reached during the
daytime ( ≈ 1500 UTC). A similar
behaviour was observed also for the city
of Houston, Texas (see Salamanca et al.
2010b). Once the air temperature has
reached its maximum value, the effect
of the AH is more important and can
remain until dawn. On the other hand,
this behaviour is not observed in the
FM1 station situated north of the city, in
a rural area. In fact, in the FM1 site the
vegetated part (the urban fraction covers
only a 16 %) is predominant, and
consequently
the
air-vegetation
interactions
dominates
and
the
differences between the two simulations
are small.
Figure 5. T2(AH)-T2(no AH) differences at
1800 UTC for the 30th June together with
contour lines of the urban fraction. Barajas
Airport location is indicated with a blue square
mark.
Surface
ε
λ
C
−1 −1
6
−3
α
−1
z0
(m)
(Wm K )
(⋅10 Jm K )
Roof
0.695
1.32
0.9
0.2
0.01
Wall
0.695
1.32
0.9
0.2
___
Road
0.4004
1.40
0.95
0.15
0.01
Table 3. Thermal parameters used in the urban
module BEP+BEM for the three urban classes.
λ is the thermal conductivity of the material, C is the
specific heat of the material, ε is the emissivity of the
surface, α is the albedo of the surface, and z 0 is the
roughness length for momentum over the surface.
The effect of the AH in the air
temperature is not constant during all
140
Chapter 3. A numerical study of the UBL over Madrid during the DESIREX (2008) campaign.
the day, but it is strongly linked to the
urban fraction of the corresponding area.
Figure 5 shows the 2m air temperature
differences (T2(AH)-T2(no AH)) at
1800 UTC (when this difference is
maximum) for the 30 th of June together
with contour lines of urban fraction in
the inner domain. Clearly both patterns
are overlapped, and differences up to
1.5-2 º C are reached in the urban core,
showing the importance of considering
AH in the simulation for those zones
where the urban factor is predominant.
The inclusion of the AH in the
BEP+BEM scheme, improved the
results in most of the stations (see Table
4) spread out over the city and
surroundings, although we are aware
that an exact quantification of the AH
and its effect would require more
information that was not available for
Madrid
(high
resolution
urban
morphology, correct target temperature,
percentage of buildings with air
conditioning systems, etc.)
3.2.2 Vertical profiles
open circle
BEP+BEM (AH)
closed circle
BEP+BEM (no AH)
cross
OBS.
Figure 6. Vertical profiles of the potential
temperature at Barajas Airport: a) at 0000 UTC
(30th June), b) at 1200 UTC ( 30th June), c) at
1200 UTC ( 1st July), d) at 0000 UTC ( 2nd
July), and e) 0600 UTC ( 2nd July).
Figure 6 (a-e) shows vertical
profiles of the potential temperature (θ )
at Barajas Airport (RS1), situated
( ≈ 600 m asl) at north-east of Madrid
and surrounded by urban nucleus by the
south and west. Due to its proximity to
Madrid, it is expected that air
temperature profiles can feel the
advection of the AH effects.
Measurements show a slightly stable
141
Chapter 3. A numerical study of the UBL over Madrid during the DESIREX (2008) campaign.
atmosphere, explained by the moderate
winds at surface, in the lowest 600 m at
0000 UTC ( 30 th June, Figure 6a) while
both simulations reproduce similar
behaviour but in a shallower layer (400500m above ground level). During
daytime (1200 UTC 30 th June, Figure
6b) the unstable mixed boundary layer
developed and its measured depth is
about 1 km. Model results show a
Planetary Boundary Layer (PBL)
slightly deeper (100-200m more than
measured). Interestingly, in the lowest
part of the PBL, the vertical profile
produced by BEP+BEM(AH) is closer
to measurements than the one obtained
with BEP+BEM(no AH), meaning that
the AH has a significant effect in the
vertical structure of the atmosphere that
could impact the dispersion of
pollutants, and/or cloud formation, as it
is warmer and more unstable. At the
same hour of the following day (1200
UTC, 1st July, Figure 6c), simulation
results underestimate the measurements
by 1-2 oC in the whole profile. As for
the previous day, BEP+BEM(AH)
simulates higher temperature (closer to
measurements) than BEP+BEM(no AH).
Also the PBL height is better
reproduced when AH is considered.
Finally, during the third night of the
simulation, the sounding shows a stable
atmosphere (0000 UTC
2nd July,
Figure 6d) in contrast with the situation
of the first night (Figure 6a); this is due
to the much lower wind present this
night which allows the developing of
surface-based inversions. This stability
is increased before dawn at 0600 UTC
( 2nd July, Figure 6e). At these times the
two simulations give similar results due
to the weak advection, and in general
underestimate the temperatures.
142
Chapter 3. A numerical study of the UBL over Madrid during the DESIREX (2008) campaign.
open circle
BEP+BEM (AH)
closed circle
BEP+BEM (no AH)
cross
OBS.
Figure 7. Vertical profiles of the horizontal
wind speed at Barajas Airport: a) at 0000 UTC
(30th June), b) at 1200 UTC ( 30th June), c) at
1200 UTC ( 1st July), d) at 0000 UTC ( 2nd
July), ande) 0600 UTC ( 2nd July).
Stations
ID
Meteorol
MN
ogical
1
network
MN
RMSE
MB
RMSE
MB
(AH)
(AH)
(noAH)
(noAH)
(ºC)
(ºC)
(ºC)
(ºC)
1.2143
0.5549
1.4217
1.0060
1.0897
-0.2992
1.0400
-0.0337
1.7540
-0.2354
1.7189
-0.0626
1.1519
0.2669
1.4731
0.7122
1.6381
1.2229
1.9070
1.5207
3.3073
-1.0998
3.1349
-0.7470
1.1663
-0.1058
1.1922
0.2188
2
MN
3
MN
4
MN
5
MN
6
MN
7
Surveilla
SN1
1.6905
1.3853
2.0701
1.8048
nce
SN2
2.0920
1.8420
2.4221
2.2263
network
SN3
2.5123
1.8254
2.8024
2.1442
AEMET
AS1
1.4715
0.2495
1.5695
0.7490
stations
AS2
2.2554
0.4038
2.2889
1.0291
AS3
1.8615
1.0725
2.0555
1.3692
AS4
1.4939
0.7420
1.7529
1.1321
AS5
2.0401
0.2347
2.1397
0.6571
AS6
2.1216
-0.9030
1.6560
-0.3549
AS7
1.3615
0.8291
1.3067
0.8354
Fixed
FM1
2.1332
-0.1208
2.0464
0.0335
masts
FM2
2.0561
1.6459
2.4783
2.1508
FM3
2.1716
1.9964
2.6668
2.4688
FM4
1.4634
0.9028
1.9034
1.4928
FM5
1.4838
1.0734
1.6885
1.3708
Table 4. Statistical parameters (RMSE and MB)
for the air temperature when the AH is ejected
directly into the atmosphere and when is not.
In Figure 7 (a-e) the vertical
profiles of the horizontal wind speed are
displayed. During the first night there
are no differences between the wind
speeds produced by the two simulations
that are on the same order of the
measured values. It is interesting to
observe (see Figures 7b and 7c) that
during daytime (1200 UTC), when the
differences in the air temperature
profiles produced by the two
simulations are larger, also the
differences in the wind speed are larger,
especially for July 1st at 1200 UTC.
Compared to measurements, the 30th of
June, models give acceptable results,
while the 1st of July, both simulations
overestimate the wind speed. This wind
speed overestimation is carried on to the
night of the 2nd of July. The analysis of
this overestimation would require a
detailed study of the impact of different
PBL schemes that could clarify the
reason of this setback, but it was not the
aim of this work. Finally, in Figure 8 (ae) the wind directions are showed.
There are not appreciable differences at
0000 UTC ( 30 th June, Figure 8a)
between the two simulations, and they
were able to capture the slight change in
the wind direction that took places at
1.4 km agl, but 500 m below. At noon
1200 UTC (Figures 8b and 8c) the
model captured the wind direction in the
whole UBL. The first day ( 30 th June,
143
Chapter 3. A numerical study of the UBL over Madrid during the DESIREX (2008) campaign.
Figure 8b) some differences were
observed between the two simulations
at the first 400 m agl, showing a better
fit against measurements when the AH
was considered. The second day
( 1st July) the two simulations departure
slightly the wind direction in the first
200 m agl. Finally, at night (Figures 8d
and 8e) the two simulations gave similar
results reproducing well the wind
direction except in the first 400 m.
The tendency of the model to
overestimate the wind speed the 1st of
July is confirmed by the comparison
against three vertical tethered balloon
soundings carried on at Nuevos
Ministerios (CB1 site) (not shown).
open circle
BEP+BEM (AH)
closed circle
BEP+BEM (no AH)
cross
OBS.
Figure 8. Vertical profiles of the wind direction
at Barajas Airport: a) at 0000 UTC (30th June),
b) at 1200 UTC ( 30th June), c) at 1200 UTC (
1st July), d) at 0000 UTC ( 2nd July), and e)
0600 UTC ( 2nd July).
3.2.3 Wind field
One of the most important
problems in the comparison between
wind field measurements against
mesoscale model results, in urban areas,
is the possible impact in the
measurement of microscale effects, not
captured by the model’s resolution. The
impact of these microscale local effects
on the air temperature is less important
144
Chapter 3. A numerical study of the UBL over Madrid during the DESIREX (2008) campaign.
than on the wind field. For this reason,
in urban areas, the position of a
meteorological
station
plays
a
fundamental role. The position of the
station must be sufficiently away from
walls and roofs to avoid these local
effects. Detailed information on the
exact position of the stations in the
microscale structure of the city is not
available, but it is likely that the
position can explain part of the
differences between model results and
measurements.
centre. The model captured reasonably
well the wind speed and wind direction
the first day of simulations (Figures 9a,
10a, and 11a), but not the second day
( 1st of July, Figures 9b, 10b, and 11b)
where the model has some problems to
capture correctly the wind field. There
are not important differences between
the
two
BEP+BEM(AH)
and
BEP+BEM(no AH) runs, except for the
period where the AH is maximum (from
1200 to 1800 UTC). Similar behaviour
has been observed for other stations (not
shown). Recent studies (Hu et al. 2010)
have demonstrated that the choice of the
PBL scheme in WRF can significantly
affect the wind speed. A detailed study
of the sensitivity of the results to the
PBL scheme and soil boundary
conditions could help to understand the
origin of the disagreement between
model results and measurements for the
second day of simulation.
Figure 9. Temporal evolutions of the horizontal
wind speed at 10 m agl at MN4 site: a) 30th of
Jun, and b) 1st of July.
In Figures 9-11 temporal
evolutions of the measured horizontal
wind speed against modelled (at 10m
agl) are displayed in three different sites,
MN4, AS1 and MN7 stations. The MN4
site is an urban area situated at the south
of the city, while the MN7 is an urban
site rounded by green areas situated at
north-east. The AS1 site represents an
urban park (Retiro) close to the city
Figure 10. As in Fig. 9, but at AS1 site.
145
Chapter 3. A numerical study of the UBL over Madrid during the DESIREX (2008) campaign.
Figure 11. As in Fig. 9, but at MN7 site.
4. UHI and evaluation of energy
consumption mitigation strategies
4.1 Urban Heat Island
The UHI is a known problem
very common in urban environments.
The trapping of radiation that takes
place in the urban canopy, the thermal
properties of the building materials, and
the AH due to human activities are the
principal factors that take part in the
formation of the UHI phenomenon. The
UHI during heat waves episodes can
represent a significant risk for the
population. A recent review on urban
climate,
including
turbulence,
exchanges of energy and water, and the
urban heat island can be found in
Arnfield (2003).
In this section, the urban heat
island of Madrid has been analysed. In
Figure 12a, the 2m air temperature
(BEP+BEM(AH) case) at 2200 UTC
( 30 th of June) has been plotted together
with the wind field at 10 m agl
(katabatic flows coming from SE
regions were present during the night).
Clearly, the UHI is completely
developed during the nocturnal time,
and cover the totality of the urban areas
with different degrees of intensity.
Areas with greater building plan area
fraction present a higher UHI intensity
(compare Figure 12a to Figure 3) and
areas with low urban fraction a lower
intensity. The city of Madrid is
surrounded by dry lands that not favour
the UHI during daytime. In the Figure
12b, the air temperature at 10 m agl has
been plotted at 1300 UTC ( 30 th of June)
in the inner domain, and small
differences are observed between the
city and surrounded areas. It is
interesting to observe that the increase
of temperature at night reaches even
areas not built up close to the city due to
the advection of the accumulated heat.
The UHI intensity is spatially dependent
within the urban area, and it can be
observed in the figure that the
magnitude could reach up to 6 º C ,
which is in agreement with the values
found by Yagüe et al. (1991). One of
the important consequences of the
larger nocturnal heating in the city
compared to the urban surroundings is
the weaker surface-based thermal
inversions formed (clearly seen in
Figure 13) due to the convective
motions generated by the warmer core
of the city. It is important to mention
that the inner domain presents strong
variations in height and consequently
the
UHI
intensity
could
be
misinterpreted. For this reason, a
vertical ZY section of potential
temperature ( longitude = 3.70W ) is
plotted in Figure 13. The air
temperature difference between central
areas of the city ( ≈ latitude = 40.4 N )
and southern ( ≈ latitude = 40.15 N )
areas (points with the same height asl)
reached up to five or six oC.
146
Chapter 3. A numerical study of the UBL over Madrid during the DESIREX (2008) campaign.
Due to the small differences in height
(see Table 1), the magnitude of the plots
represents the real intensity of the UHI.
Although the model is not able to
reproduce the maximum magnitude of
the UHI, it is able to capture reasonably
well the tendency of the curve.
Differences between the couple FM4
and MN2 are plotted in Figure 14b. Due
to differences in height, this picture
does not represent the real magnitude of
the UHI, but the influence of the urban
core is clear. In this case the model was
able to capture the maximum magnitude
of the air temperature differences, even
if with a time delay.
Figure 12. a) 2m air temperature
(BEP+BEM(AH) case) and wind speed at 10 m
agl at 2200 UTC (30th of June), and b) air
temperature and wind speed at 10 m agl at 1300
UTC ( 30th of June) in the inner domain.
Figure 13. Vertical section ( longitude = 3.70W
) of the air potential temperature θ (ºC) at
2200 UTC ( 30th of June) over the city of
Madrid.
Finally, the temporal evolution
of 2m air temperature differences
between couples of stations has been
analysed. In Figure 14a, results for the
SN1 (into the city centre) and AS2
(outside the city) stations are showed.
Figure 14. Temporal evolution differences of the
2m air temperature between two sites showing
the evolution of the UHI (measured against
modelled): a) SN1 and AS2 sites, and b) FM4
and MN2 sites.
Summarizing,
the
Madrid
metropolitan area experiments strong
nocturnal UHI in summer periods with
intensities that can reach up to 5 6 º C in some regions of the urban area.
4.2 Energy consumption mitigation
strategies
147
Chapter 3. A numerical study of the UBL over Madrid during the DESIREX (2008) campaign.
In the last years important
efforts have been carried out to mitigate
the UHI. High albedo surfaces, urban
forestry (planting trees in open spaces),
green roofs, and new building materials
with different thermal properties are the
principal strategies to reduce UHI
impact and energy consumption in
urban areas (see Rosenzweig, C., and
Co-authors, 2009), although some of
them could have negative effect on the
air quality. In BEP+BEM the energy
consumption due to AC systems can be
computed and different strategies can be
evaluated. In this section, three
strategies are analysed (by means of
three
new
simulations)
without
considering the possible impacts on air
quality: the first consists in a change of
the albedo of the roofs from 0.2 to 0.4
(hereafter
referred
as
BEP+BEM(AH)_ALB case), the second,
consists in a modification of the thermal
properties of the roofs (see Table 3) by
introducing an internal layer of 6cm of
insulating material (specific heat
C = 0.382MJm −3K −1 , and thermal
conductivity
= 0.09Wm −1K −1
)
(BEP+BEM(AH)_INSULATION case),
and finally the third (BEP+BEM(no
AH)_ALB_INSULATION case) groups
the mentioned modifications (change in
the albedo, introduction of the
insulating material, and no ejection of
the AH into the atmosphere) in a single
strategy. To evaluate the impact on the
UHI, in Figures 15 (a-c) the 2m air
temperature and wind speed at 10 m agl
are plotted at 2200 UTC ( 30 th of June)
for the three cases. Comparing the
BEP+BEM(AH)_ALB case against the
previous BEP+BEM(AH) simulation
(see Figs. 15a and 12a), it can be seen
that areas with greater temperatures (2930 º C ) have been reduced, and the total
energy
consumption
(EC)
was
decreased by 4.95 %. The total EC
when the AH is not ejected into the
atmosphere
(BEP+BEM(no
AH)
simulation) was reduced by 2.79 %,
showing the existing feed backs
between the AC systems and the
outdoor temperature. From the point of
view of energy saving, the albedo
strategy was better than the no AH
strategy. Although the increase of the
albedo is a known step to reduce the
UHI and EC in summer conditions, it
could be counterproductive for winter
periods (an increase in winter heating
energy consumption was found by Saiz
et al. 2006). Observing the Figures 15b
and 12a, not appreciable differences
appear
when
the
two
cases
BEP+BEM(AH)
and
BEP+BEM(AH)_INSULATION
are
compared. It seems to be that the
insertion of insulating material would
affect the total EC (the total
consumption was reduced in a 3.08 %)
but not the nocturnal UHI because the
heat released through the exterior walls
during the night it would be similar in
both simulations. An important energy
saving was obtained with the
BEP+BEM(no
AH)_ALB_INSULATION
strategy
reaching up to 10.47 %. It is interesting
to mention that the sum of the energy
saving obtained with every strategy
separately (change in the albedo of the
roofs, introduction of the insulating
material, and the not ejection of the AH
into the atmosphere) is bigger than the
energy saving obtained with the
BEP+BEM(no
AH)_ALB_INSULATION simulation.
This fact is explained because the air
temperature, the EC, and the ejected AH
are physical magnitudes interconnected
and strongly dependent. A considerable
reduction of the UHI (1-2 ºC) was
observed in some urban areas when the
two cases BEP+BEM(AH) against
BEP+BEM(no
AH)_ALB_INSULATION (see figures
15c and 12a) are compared. Finally, in
Table 5, results of the energy saving
obtained with every strategy and for
148
Chapter 3. A numerical study of the UBL over Madrid during the DESIREX (2008) campaign.
every urban classes (classification
derived of the CORINE database) are
showed. The greatest saving for the
BEP+BEM(no AH) strategy took place
in the CUF areas that curiously
represent the smallest coverage
extension. In these areas, the urban
fraction is practically 100 %, and the
building plan area fraction represents 60
% (see Table 2). All these factors
increase the AH emissions, and
consequently the energy saving was
greater in these areas. On the other hand,
observing the other studied strategies,
the greatest saving coincided with the
more extensive areas (DUF). It is not
possible to generalize the conclusions
because the results can be strongly
linked
to
urban
morphological
parameters (see Table 2) that could be
different for other cities.
STRATEGIES
%
%
%
OUA
CUF
DUF
BEP+BEM(AH)_ALB
3.64
4.57
5.73
BEP+BEM(AH)_INSULATION
2.15
2.58
3.72
BEP+BEM(no AH)
2.32
4.29
2.43
BEP+BEM(no
7.66
10.42
11.83
AH)_ALB_INSULATION
Table 5. Saving energetic produced in every
urban class over the Madrid metropolitan area.
Figure 15. 2m air temperature and wind speed
at 10 m agl at 2200 UTC (30th of June) in the
inner domain: a) BEP+BEM(AH)_ALB case, b)
BEP+BEM(AH)_INSULATION case, and c)
BEP+BEM(no AH)_ALB_INSULATION.
5. Conclusions and future work
In this paper a new UCP
(BEP+BEM) have been tested over the
Madrid metropolitan area with the WRF
model in summer conditions. The urban
scheme is integrated in the public
release WRF V3.2 from April 2010.
This UCP represents the most
sophisticated urban parameterization
coupled to the WRF model up to now,
and the heat (sensible/latent) fluxes
exchanged between the buildings and
the atmosphere are considered. Two
consecutive days have been analysed
( 30 th of June and 1st of July) and model
results have been compared against
measurements
recorded
in
the
DESIREX 2008 campaign. The urban
model was able to reproduce
satisfactorily the air temperature over
the period analysed. On the other hand,
149
Chapter 3. A numerical study of the UBL over Madrid during the DESIREX (2008) campaign.
the wind field over the city is more
difficult to validate and is strongly
dependent of the mesoscale circulations
in the surroundings areas. In any case,
the wind field was captured reasonably
well the first day of simulation, and
some difficulties appeared the second
day: WRF overestimated the wind
speed. The impact of the AH due to
space cooling loads (peaks up to
110Wm−2 were reached in the middle
of the afternoon) and their EC was
evaluated and different mitigation
strategies were addressed. At some
hours, the AH was responsible of an
increase in the air temperature up to 1.52ºC .
The UHI over Madrid reached up to 5-6
º C in some urban regions. The total EC
was reduced close to 5 % when the
albedo was increased and 3 % when the
insulating material was introduced. A
high albedo at the roofs, insulating
materials inside the walls, and AC
systems not ejecting directly into the
atmosphere are strategies that would
reduce notably the UHI (1-2ºC) and EC
(an energy saving up to 10.5 % was
obtained).
It is important to mention here
that to quantify correctly the AH
released in an urban area through an
UCP, detailed information of the urban
morphology is necessary, and high
resolution urban canopy parameters data
sets recommended. The spatial and
temporal variations of the AH, UHI and
EC over a city are physical magnitudes
difficult to quantify without a detailed
up to date urban morphology data.
Thanks to the increasing power of the
computers and detailed urban databases,
the simulation of the above three
magnitudes with a meteorological
model can be obtained. The impact of
the AH on the pollutant concentrations
and/or cloud formation can be also
investigated with these numerical tools.
Acknowledgements
We thank CIEMAT for the doctoral
fellowships
held
by
Francisco
Salamanca. We thank Dra. Rocío
Macarena Alonso and Dra. Marta
García of CIEMAT for providing an up
to date land use file of the Madrid
Community. We also thank Eugenio
Sánchez García of CIEMAT who
transformed to ASCII format the urban
information existing in CORINE
database. Finally, we also thank Dr.
Mukul Tewari of NCAR for providing
the NCEP operational data. This work
was funded by the Ministry of
Environment of Spain and partially by
the project CGL2009-12797-C03-03
(Spanish Ministry of Science and
Innovation).
References
Arnfield, A. J., 2003: Two decades of
urban climate research: A review
of turbulence, exchanges of
energy and water, and the urban
heat island. Int. J. Climatol., 23: 126.
Bougeault, P., and Lacarrère, P., 1989:
Parameterization of orographyinduced turbulence in a mesobetascale model. Mon. Wea. Rev., 117,
1872-1890.
Chen, F., and Dudhia, J., 2001:
Coupling an advanced landsurface/hydrology model with
the Penn State/NCAR MM5
modeling system. Part I: Model
implementation and sensitivity.
Mon. Wea. Rev., 129, 569-585.
Ching, J., and Coauthors, 2009:
National Urban Database and
Access Portal Tool. BAMS.
Doi:10.1175/2009BAMS2675.1.
Hu, X.-M., Nielsen-Gammon, J. W.,
and Zhang, F., 2010: Evaluation
of Three Planetary Boundary
Layer Schemes in the WRF model.
J. Appl. Meteorol. Climatol. Doi:
10.1175/2010JAMC2432.1.
Kanda, M., Kawai, T., Kanega, M.,
Moriwaki,
R.,
Narita
K.,
150
Chapter 3. A numerical study of the UBL over Madrid during the DESIREX (2008) campaign.
Hagishima, A., 2005: A simple
energy balance model for regular
building arrays. Boundary-Layer
Meteorol., 116, 423-443.
Kikegawa, Y., Genchi, Y., Yoshikado,
H., and Kondo, H., 2003:
Development of a numerical
simulation
system
toward
comprehensive assessments of
urban warming countermeasures
including their impacts upon the
urban buildings energy demands.
Applied Energy, 76, 449-466.
Kusaka, H., Kondo, H., Kikegawa, Y.,
and Kimura, F., 2001: A simple
single-layer urban canopy model
for
atmospheric
models:
comparison with multi-layer and
slab models. Boundary-layer
Meteorol., 101, 329-358.
Landsberg, H. E., 1981. The Urban
Climate. Academic Press, New
York, 277 pp.
Lemonsu, A., Belair, S., and Mailhot, J.,
2009: The New Canadian Urban
Modelling System: Evaluation for
Two Cases from the joint Urban
2003 Oklahoma City Experiment.
Boundary-Layer Meteorol., 133,
47:70.
Lo Jeff, C. F., Lau Alexis, K. H., Chen,
F., Fung Jimmy, C. H., Leung
Kenneth, K. M., 2006: Urban
Modification in a Mesoscale
Model and the Effects on the
Local Circulation in the Pearl
River Delta Region. J. Appl.
Meteorol. Climatol., 46, 457-476.
Martilli, A., Clappier, A., and Rotach,
M. W., 2002: An urban surface
exchange parameterization for
mesoscale models. BoundaryLayer Meteorol., 104, 261-304.
Martilli, A., Roulet, Y.-A., Junier, M.,
Kirchner, F., Rotach, M. W., and
Clappier, A., 2003: On the impact
of urban surface exchange
parameterizations on air quality
simulations: the Athens case.
Atmospheric Environment, 37,
4217-4231.
Masson, V., 2000: A physically-based
scheme for the urban energy
budget in atmospheric models.
Boundary-Layer Meteorol., 94,
357-397.
Oke, T. R., 1987. Boundary Layer
Climate, 2nd ed. Methuen,
London, UK, 435 pp.
Palacios, M., Kirchner, F., Martilli, A.,
Clappier,
A.,
Martín,
F.,
Rodríguez, M. E., 2002: Summer
ozone episodes in the Greater
Madrid area. Analyzing the ozone
response to abatement strategies
by
modelling.
Atmospheric
Environment, 36, 5323-5333.
Rosenzweig, C., and Co-authors, 2009:
Mitigating New York City´s Heat
Island.
BAMS.
Doi:
10.1175/2009BAMS2308.1
Sailor, D., 2010: A review of methods
for estimating anthropogenic heat
and moisture emissions in the
urban environment. International
Journal of Climatology. Doi:
10.1002/joc.2106.
Saiz, S., Kennedy, C., Bass, B., and
Pressnail, K., 2006: Comparative
Life Cycle Assessment of
Standard and Green Roofs.
Environ. Sci. Technol., 40, 43124316.
Salamanca, F., and Martilli, A., 2010: A
new Building Energy Model
coupled with an Urban Canopy
Parameterization for urban climate
simulations—Part II. Validation
with one dimension off-line
simulations.
Theoretical
and
Applied Climatology, 99, 345-356.
Salamanca, F., Krpo, A., Martilli, A.,
Clappier, A., 2010a: A new
Building Energy Model coupled
with
an
Urban
Canopy
Parameterization for urban climate
simulations—Part I. Formulation,
verification and a sensitive
analysis of the model. Theoretical
151
Chapter 3. A numerical study of the UBL over Madrid during the DESIREX (2008) campaign.
and Applied Climatology, 99, 331344.
Salamanca, F., Martilli, A., Tewari, M.,
and Chen, F., 2010b: A study of
the urban boundary layer using
different urban parameterizations
and high-resolution urban canopy
parameters with WRF (The case
of Houston). J. Appl. Meteorol.
Climatol.(accepted).
Skamarock, W. C., and Coauthors,
2005: A description of the
Advanced Research WRF version
2. NCAR Tech. Note TN468+STR, 88 pp. [Available from
NCAR, P. O. Box 3000, Boulder,
CO 80307.]
Sobrino, J. A., and Coauthors, 2009:
DESIREX 2008 Final Report
Contract
nº:
21717/08/I-LG
(http://www.uv.es/desirex).
Terradellas, E., and Cano, D., 2007:
Implementation of a singlecolumn model for fog and low
cloud forecasting at CentralSpanish airports. Pure and
Applied Geophysics, 164, 13271345.
Yagüe, C., Zurita, E., and Martínez, A.,
1991: Statistical analysis of the
Madrid urban heat island.
Atmospheric Environment, 25B,
327-332.
152
Capítulo 4
153
Capítulo 4
CAPÍTULO 4
DISCUSIÓN INTEGRADORA
154
Capítulo 4
4.1 Discusión integradora
Los trabajos que se han presentado a lo largo de esta tesis doctoral han pretendido
avanzar en el conocimiento de la formación de la isla de calor urbana. La isla de calor es un
viejo conocido problema de los estudiosos del clima urbano que puede afectar negativamente
a los habitantes de las grandes ciudades. Aunque gran parte del conocimiento del fenómeno
de la isla de calor se adquirió hace algunas décadas, su modelización sólo ha sido posible en
los últimos años. Hoy en día su estudio está muy extendido gracias en buena parte al aumento
en la capacidad de cálculo de los ordenadores y al desarrollo teórico de las parametrizaciones
urbanas. Éstos proporcionan las condiciones de contorno a los modelos atmosféricos
resolviendo los efectos térmicos y dinámicos producidos por los edificios en la atmósfera. Las
ciudades con extensiones de varios kilómetros pueden modificar de un modo importante el
clima local.
Un componente fundamental originado por las distintas actividades humanas y que
afecta directamente a la isla de calor urbana es el calor antropogénico. Las principales fuentes
de este calor son debidas al consumo energético y están originadas por el tráfico, las
actividades industriales y la regulación térmica interna que tiene lugar en los edificios. Estas
fuentes de calor no se resuelven explícitamente en la mayoría de los esquemas urbanos. Sin
embargo, en algunos esquemas se tienen en cuenta añadiendo al flujo de calor total valores
horarios que se basan en datos de consumo energético mensuales y perfiles temporales.
Desafortunadamente, la obtención de estos datos es, para la mayoría de las ciudades, muy
difícil. Además estos perfiles diarios dependen fuertemente de las condiciones meteorológicas
y la extrapolación a partir de datos medios mensuales es una importante fuente de error. Los
aires acondicionados son una fuente de calor para la atmósfera y contribuyen a aumentar el
efecto de la isla de calor generando un fenómeno de retroalimentación positiva que no es
155
Capítulo 4
considerado con esta técnica. Por estas razones, en esta tesis se ha optado por resolver de
forma explícita los intercambios de calor (sensible/latente) que tienen lugar entre los edificios
y la atmósfera. Nuestro modelo considera los intercambios de calor a través de la ventilación
natural y a través del uso de los sistemas que regulan la temperatura interna. Si bien el
consumo energético medio anual puede ser muy diferente dependiendo del país o de la ciudad
considerada, se estima que en torno al 50 % del consumo energético anual de los hogares de
las ciudades de los EEUU tiene su origen en la regulación térmica interna de los edificios
(EIA, 2005). Este gran porcentaje muestra la importancia de esta componente en el cómputo
total del calor antropogénico producido en las ciudades y consecuentemente de su efecto en la
isla de calor. Por consiguiente, es de esperar que el resolver térmicamente los edificios, tenga
un efecto notable en la modelización atmosférica sobre las ciudades. La regulación de la
temperatura interna de los edificios se lleva a cabo en verano por los sistemas de aire
acondicionado y en invierno por los sistemas de calefacción. Si bien en esta tesis hacemos
referencia constantemente a los sistemas de aire acondicionado, el modelo desarrollado regula
la temperatura interna tanto en períodos estivales (enfriando) como en invernales
(calentando).
Una vez que se desarrolló el modelo energético BEM (Salamanca et al., 2010a), se
realizó una inter-comparación frente a otros programas utilizados en el análisis térmico de
edificios utilizados en estudios de ingeniería (Salamanca & Martilli, 2010). Los resultados
muestran que BEM es capaz de describir los mecanismos más importantes que gobiernan la
generación de calor dentro de los edificios y los intercambios con el exterior. La ventaja de
BEM frente a los otros programas es que es fácilmente integrable en una parametrización
urbana y se pueden llevar a cabo simulaciones a mesoescala del clima urbano (objetivo
principal de esta tesis) teniendo en cuenta los efectos de retroalimentación, mencionados más
156
Capítulo 4
arriba, entre el uso de los aires acondicionados y la isla de calor. En BEM se calculan los
flujos de calor (sensible/latente) intercambiados con el exterior debido al uso de los sistemas
de aire acondicionado cuando regulamos la temperatura interna de los edificios. La
posibilidad de estimar con un modelo atmosférico el consumo energético debido a factores
climáticos (aire acondicionado y calefacción) a nivel de una ciudad entera, abre la posibilidad
de utilizar esta herramienta para evaluar estrategias tanto de mitigación de la isla de calor
urbano, como de reducción del consumo energético. Es importante recordar que las ciudades
son los lugares donde más se consume energía a nivel mundial, y poder reducir este consumo
es una clave para poder controlar el cambio climático.
Varias simulaciones en una dimensión vertical (off-line) se llevaron a cabo con el
nuevo modelo energético una vez que estuvo unido a la parametrización urbana BEP para una
pequeña zona de la ciudad suiza de Basel (Salamanca & Martilli, 2010), donde se había
realizado una campaña de medidas. Estas simulaciones muestran que la inclusión de los flujos
de calor provenientes de los aires acondicionados acercan los resultados a las medidas de flujo
de calor sensible tomadas por encima de los edificios de la calle. Estos primeros resultados
indican, además, que los sistemas de aire acondicionado pueden aumentar, para la ciudad
considerada, la temperatura exterior hasta 2-3 ºC en períodos estivales, aunque el impacto
real podría ser algo menor ya que la advección no estuvo considerada en estas primeras
simulaciones. Flujos diarios medios de calor de 50 a 160 W/m 2 fueron emitidos por los
sistemas de aire acondicionado con picos que alcanzaron hasta 200 W/m 2 durante algunas
horas del día (valores típicos de flujos de calor sensibles totales en ciudades en latitudes
similares a la de Basel alcanzan entre 300-400 W/m 2 en verano). Estos resultados ponen de
manifiesto la necesidad de tener en cuenta estos flujos en las simulaciones con modelos
mesosescalares y de su posible impacto no solo sobre el clima urbano, sino que también en la
157
Capítulo 4
dispersión de contaminantes y en la predicción de la formación de nubes porque pueden
afectar de manera significativa la estructura entera de la capa limite atmosférica. Se han
realizado diferentes estudios de sensibilidad del consumo energético y los resultados indican
que un aumento de la temperatura exterior de 1 ºC incrementa entre 3-5 % el consumo
energético. Por otro lado se ha visto que el uso de materiales aislantes reduce el consumo en
torno a un 10 % (Salamanca & Martilli, 2010). Para evaluar la importancia de estos flujos en
una simulación real en 3D, en la figura 4.1 se muestran los flujos totales de calor sensible
(HFX) intercambiados con la atmósfera cuando se consideran los flujos provenientes de los
sistemas de aire acondicionado y cuando no se tienen en cuenta. La figura corresponde a una
zona céntrica (latitud= 40.41387778º, longitud= -3.705519444º) de la ciudad de Madrid y las
simulaciones fueron realizadas con el modelo atmosférico WRF utilizando el esquema urbano
BEP+BEM, los resultados mostrados corresponden al día 30 de Junio del año 2008
(Salamanca et al., 2010c).
158
Capítulo 4
Figura 4.1. Flujos totales de calor sensible: la línea negra representa el caso en el que el calor proveniente de
los sistemas de aire acondicionado es considerado y la línea roja cuando no se tienen en cuenta.
En la figura 4.1 se observa que el flujo total de calor sensible no solo se ve afectado
notablemente durante las horas diurnas, sino que el efecto de los sistemas de aire
acondicionado se prolonga durante toda la noche. Los valores positivos del flujo de calor
durante la noche indican que las superficies urbanas siguen calentando el aire y por lo tanto
acentúan la formación de la isla de calor.
La primera ciudad que se ha simulado en esta tesis con el modelo atmosférico WRF ha
sido la ciudad de Houston (Texas, EEUU) (Salamanca et al., 2010b). En la primera parte de
este estudio se han comparado las cuatro parametrizaciones urbanas disponibles en el modelo
con el fin de mostrar las posibles diferencias existentes entre ellas. Con el nuevo esquema
BEP+BEM se obtuvieron buenas estimaciones de la temperatura del aire y los sistemas de
159
Capítulo 4
aire acondicionado fueron los responsables de un aumento de la temperatura de hasta 2 ºC en
algunos lugares de la ciudad. En la segunda parte del estudio se hizo uso de información
morfológica detallada de la ciudad (existente en la base de datos NUDAPT) con una
resolución de 1 km 2 y se repitieron las simulaciones con los esquemas BEP y BEP+BEM.
Los resultados ponen de manifiesto que el nuevo esquema BEP+BEM es más sensible que el
anterior esquema BEP a los parámetros geométricos que describen la ciudad, ya que con el
nuevo modelo se calcula el calor antropogénico proveniente de los sistemas de aire
acondicionado. El flujo de calor proveniente de estos sistemas depende de las dimensiones de
los edificios y éstas de los parámetros morfológicos. Se ha calculado el consumo energético
de la ciudad con la nueva parametrización BEP+BEM y se obtuvieron buenos resultados al
compararlos con valores obtenidos por otras metodologías totalmente diferentes (top-down y
botton-up) cuando se usó la información morfológica detallada. Es importante comentar aquí
que diferentes simulaciones de períodos más largos (se han simulado sólo dos días) junto con
una mayor red de observaciones (tanto en superficie como en altura) deberían llevarse a cabo
para extraer conclusiones más definitivas. En la figura 4.2 mostramos algunos perfiles
verticales de la temperatura potencial obtenidos con los cuatro esquemas urbanos en dos
zonas diferentes de la ciudad de Houston durante la noche. Comparando BEP con BEP+BEM
vemos claramente el efecto de los sistemas de aire acondicionado en los primeros 100-150 m
de la capa límite atmosférica. Cuando se tuvieron en cuenta estos flujos, se logró reproducir
mejor la temperatura del aire sobre la ciudad. Claramente comparando las figuras 4a y 4b
vemos que el efecto del calor antropogénico no es homogéneo y depende fuertemente de la
morfológia urbana de la zona considerada.
160
Capítulo 4
161
Capítulo 4
Figura 4.2. Perfiles verticales de la temperatura potencial obtenidos con los cuatro esquemas urbanos
(BULK, UCM, BEP y BEP+BEM) a las 0300 LST (1 Septiembre 2000) en dos zonas diferentes de la ciudad de
Houston, Texas: a) zona comercial y b) zona residencial.
Finalmente también se ha simulado la ciudad de Madrid (España) con el modelo
atmosférico WRF (Salamanca et al., 2010c). El período analizado tuvo buenas condiciones
sinópticas favoreciéndose la formación de la isla de calor y coincidió con la campaña
meteorológica DESIREX que tuvo lugar en el verano del 2008. El único esquema urbano
utilizado para este estudio ha sido la nueva parametrización BEP+BEM. En la primera parte
del trabajo hemos estudiado el impacto de los sistemas de aire acondicionado y en la segunda
se ha hecho una evaluación de diferentes estrategias de reducción del consumo energético y
162
Capítulo 4
mitigación de la isla de calor. Los resultados mostraron picos de calor provenientes de los
sistemas de aire acondicionado de 110 W/m 2 y fueron los responsables de un aumento de la
temperatura del aire de hasta 1.5-2 ºC en algunos lugares de la ciudad durante la noche. La
isla de calor sobre Madrid alcanzó 5-6 ºC cuando estuvo totalmente desarrollada de noche. En
la figura 4.3 mostramos la diferencia de temperatura obtenida cuando el calor antropogénico
procedente de los sistemas de aire acondicionado es liberado a la atmósfera y cuando no es
liberado para los dos días simulados. Comparando ambas figuras se observa claramente que el
impacto del calor emitido es mayor el primer día simulado. El segundo día fue menos
caluroso y por consiguiente el calor antropogénico liberado menor. Estos resultados muestran
la relación directa que existe entre el calor antropogénico y las condiciones meteorológicas.
163
Capítulo 4
Figura 4.3. Diferencias de la 2-m temperatura (ºC) del aire T2(AH)-T2(noAH) a las 2000 LST sobre la
ciudad de Madrid para los dos días simulados: a) 30 de Junio y b) 01 de Julio del 2008.
Las siguientes tres estrategias de reducción del consumo energético fueron analizadas:
aumento del albedo de los tejados, uso de materiales aislantes en los muros y finalmente la
eliminación del calor proveniente de los sistemas de aire acondicionado. El aumento del
albedo y el uso de materiales aislantes redujeron el consumo energético en un 5 % y un 3 %
respectivamente. Cuando se consideraron las tres estrategias de un modo global se obtuvo un
ahorro en el consumo total de hasta un 10 % y la isla de calor disminuyó en 1-2 ºC.
En esta tesis también se ha investigado el papel que juegan las propiedades térmicas de
los diferentes materiales presentes en una zona urbana a la hora de calcular el flujo total de
164
Capítulo 4
calor sensible intercambiado con la atmósfera (Salamanca et al., 2009). Cuando se calcula
este flujo, se necesita conocer las propiedades térmicas del material representativo de la zona
bajo estudio y la forma estándar de proceder es “promediando” las distintas propiedades
térmicas teniendo en cuenta el porcentaje presente de cada material en la zona. En esta tesis
proponemos calcular los valores térmicos del material representativo de una forma totalmente
diferente teniendo como objetivo el reducir la diferencia entre la suma de los flujos
intercambiados por cada una de las superficies de los materiales presentes con la atmósfera, y
el flujo de calor que se obtendría con el nuevo material representativo. Se ha considerado una
situación sencilla donde se conocía la solución analítica y se ha igualado la suma de los flujos
de calor obtenidos con los diferentes materiales con la que se obtendría con el nuevo material.
De esta forma se han derivado las propiedades térmicas del material representativo de la zona
y varias simulaciones han demostrado que el error en el cálculo del flujo de calor se ha
reducido en torno a un 50 % en situaciones reales.
El trabajo presentado en esta tesis contribuye al desarrollo de una herramienta
numérica capaz de describir de forma detallada el impacto de las ciudades en la atmósfera.
Con este tipo de herramientas se obtienen distribuciones espaciales tanto del calor
antropogénico como del consumo energético. Gracias al trabajo realizado en esta tesis
podemos contribuir al desarrollo de ciudades más sostenibles desde un punto de vista
energético.
165
Capítulo 4
166
Capítulo 4
CHAPTER 4
DISCUSSION
167
Capítulo 4
4.1 Discussion
The works that have been presented in this doctoral thesis have the aim to improve the
knowledge of the formation of the urban heat island. The urban heat island is a well known
feature of the urban climate that affects inhabitants of the big cities. Although a significant
part of the knowledge of the urban heat island has several decades, its modelling has been
possible only in the recent years. Nowadays, its study is very extended in part thanks to the
increase in the computer power and the theoretical development of the urban canopy
parameterizations. The urban schemes provide the boundary conditions to the atmospheric
models solving the thermal and dynamical effects produced by the buildings in the
atmosphere. The cities with extensions of several kilometres can modify significantly the
local climate.
An important component originated by the different human activities and that it affects
directly the urban heat island is the anthropogenic heat. The principal sources of this heat are
due to the energy consumption and are originated by the traffic, the industrial activities, and
the internal regulation of the air temperature inside the buildings. These heat sources are not
solved explicitly in most of the urban schemes. However, in some schemes they are taken into
account by adding to the total sensible heat flux daily profiles of heat based monthly data of
energy consumption. For many cities, however, it is often impossible to get this information.
Moreover, these values depend strongly on the meteorological conditions and the use of
monthly values can introduce significant errors. In addition, this technique does not take into
account the positive feedback between the use of the air conditioning systems (which
contributes to the Urban Heat Island, since they eject heat to the atmosphere) and the UHI
itself. For these reasons in this thesis we have chosen to solve explicitly the heat
(sensible/latent) exchanges that take place between the buildings and the atmosphere. Our
168
Capítulo 4
model considers the direct exchanges through the natural ventilation and through the use of
systems that regulate the indoor temperature. Although the energy annual consumption can be
very different depending on the country and the city considered, it is estimated that over 50 %
of the annual energy consumption of the houses in cities of the USA has its origin in the
indoor temperature regulation of the buildings (EIA, 2005). This percentage shows the
importance of this component in the total calculation of the anthropogenic heat produced in
the cities and consequently of its effect in the formation of the urban heat island. It is
expected, then, that resolving the thermal behaviour of the buildings may have a significant
effect in the urban mesoscale modelling. The regulation of the indoor air temperature is
carried out in summer periods by the air conditioning systems and in winter periods by
heating systems. Although in this thesis we refer constantly to the air conditioning systems,
the model developed regulates the indoor temperature in summer (cooling) and in winter
(heating) periods.
As soon as the new energy model (BEM) was developed (Salamanca et al., 2010a), it
was compared against other well-known programs used in thermal analysis of buildings.
Results show that BEM is able to describe the most important mechanisms that govern the
generation of heat inside the buildings and their exchange with the exterior. The advantage of
BEM against other programs is that it can be easily implemented in an urban scheme for
simulations of urban climate (principal aim of this thesis) taking into account the feedback
effects, mentioned above, between the use of the air conditioning systems and the urban heat
island. In BEM the heat fluxes (sensible/latent) exchanged with the exterior due to the use of
the air conditioning systems are computed when we regulate the indoor air temperature. The
possibility of estimating with an atmospheric model the energy consumption due to climatic
factors (air conditioning and heating) to city scale opens the possibility of using this tool to
169
Capítulo 4
evaluate strategies of mitigation both urban heat island and energy consumption. It is
important to remember that the cities are the places where more energy is consumed to world
scale, and to be able to reduce this consumption is an important steep to be able to control the
climate change.
Several simulations have been carried out in a vertical column (off-line) with the new
energy model integrated in the urban scheme BEP in an urban area of the Swiss city of Basel
(Salamanca & Martilli, 2010), where a field campaign took place. These first results indicate
that the new scheme improved the estimation of the sensible heat fluxes when the air
conditioning are taken into account compared to measurements above the urban canyon.
Results show also that air conditioning systems could increase the outdoor air temperature up
to 2-3 ºC, for this city in summer, although the real impact might be slightly less since
advection is not considered in these first simulations. Daily average heat fluxes from 50 to
160 W/m 2 were computed by the air conditioning systems with peak values that reached up to
200 W/m 2 during some hours of the day (note that maximum total sensible heat fluxes for this
city in summer were of the order of 300-400 W/m 2 ). These results show the need to consider
these fluxes in mesoscale models and study their possible impact on pollutant dispersion and
cloud formation since they may affect the whole structure of the PBL. Different sensitivity
studies have been carried out to evaluate the impact in the energy consumption. The results
indicate that an increase of 1 ºC in the outdoor temperature increases the energy consumption
by 3 % to 5 %. On the other hand, the use of insulating materials reduces the consumption by
about 10 % (Salamanca & Martilli, 2010). To evaluate the importance of these fluxes in a real
3D simulation, the Fig. 4.1 shows the total sensible heat (HFX) flux exchanged with the
atmosphere when the heat fluxes coming from the air conditioning systems are taken into
account and when are not. The figure shows the results for the 30 June 2008 (Salamanca et
170
Capítulo 4
al., 2010c) and represents a central area (latitude = 40.41387778º, longitude = -3.705519444º)
of the city of Madrid. The simulations were done with the atmospheric WRF model using the
urban scheme BEP+BEM.
Figure 4.1. Total sensible heat fluxes: the black line represents the case in which the heat fluxes coming from the
air conditioning systems are taken into account and the red line when are not.
In the Fig. 4.1 it is observed that the total sensible heat flux is affected not only during
the diurnal hours but the effect is extended during the whole night. The positive values
obtained during the night means that the urban surfaces continue heating the air and are the
responsible of a better representation of the formation of the urban heat island.
The first city that has been simulated in this thesis has been the city of Houston
(Texas, US) with the atmospheric WRF model (Salamanca et al., 2010b). In the first part of
171
Capítulo 4
this study, the four urban schemes available in the atmospheric model have been compared
against observations, and important differences have been observed between them. With the
new urban BEP+BEM scheme good estimations of the air temperature were obtained and the
air conditioning systems were responsible of an increase in the temperature up to 2ºC in some
places of the city. In the second part of the study detailed urban morphological information
(existing in the database NUDAPT) with a grid resolution of 1 km 2 was used, and the
simulations with the BEP and BEP+BEM schemes were repeated. Results reveal that the
BEP+BEM scheme is more sensitive than the previous BEP scheme to the geometric
parameters of the city. The reason is that in the new urban scheme the anthropogenic heat
coming from the air conditioning systems is computed, and this heat flux depends on the
morphological parameters that describe the city (e. g., the dimensions of the buildings). The
energy consumption of the city has been calculated and good estimations were obtained
against the values obtained with other methodologies totally different (bottom-up and topdown) when morphologic detailed information was used. It is important to comment here that
analyse longer periods (more than two days) together with a larger network of observations
(both surface stations and vertical profiles) could help to extract more definitive conclusions.
The Fig. 4.2 shows some vertical profiles of the potential temperature obtained with the four
urban schemes in two different areas of the city of Houston during night. Comparison
between BEP and BEP+BEM clearly shows the effect of the air conditioning systems in the
first 100-150 m of the urban boundary layer. When these fluxes were taken into account the
observed air temperature over the city was better represented by the model. Clearly,
comparing both Figs. 4a and 4b we see that the effect of the air conditioning systems is not
homogeneous and it depends strongly of the urban morphology of the zone considered.
172
Capítulo 4
173
Capítulo 4
Figure 4.2. Vertical profiles of the potential temperature obtained with the four urban schemes (BULK, UCM,
BEP and BEP+BEM) at 0300 LST (1 September 2000) in two different areas of the city of Houston, Texas: a)
commercial area and b) high residential area.
Finally, the city of Madrid (Spain) has been simulated with the WRF model
(Salamanca et al., 2010c). The period analyzed had good synoptic conditions and coincided
with the meteorological campaign DESIREX that took place in the summer of 2008. For this
study, only the urban scheme BEP+BEM has been used. In the first part of the work we have
studied the impact of the air conditioning systems on the atmosphere. In the second, an
evaluation of different energy consumption reduction programs and urban heat island
mitigation strategies has been done. The results showed peaks of heat coming from the air
174
Capítulo 4
conditioning systems of 110 W/m 2 and they were the responsible of an increase in the air
temperature up to 1.5-2 ºC in some places of the city during night. The heat island over
Madrid reached 5-6 ºC. In the Fig. 4.3 the 2-m air temperature differences are showed when
the air conditioning systems are considered and when are not for the two days simulated.
Comparing both figures the relation that exists between the heat ejected and the
meteorological conditions is clearly deduced. The second day was less warm and
consequently the anthropogenic heat originated lower.
175
Capítulo 4
Figure 4.3. 2-m air temperature (ºC) differences T2(AH)-T2(noAH) at 2000 LST over the city of Madrid for the
two days simulated: a) 30 June and b) 01 July 2008.
Three strategies of energy saving were analyzed: increase of the albedo, use of
insulating materials in the walls, and finally elimination of the heat created by the air
conditioning systems. The increase of the albedo and the use of the insulating material
reduced the energy consumption by 5 % and 3 % respectively. When the mentioned strategies
were considered globally the energy saving reached 10 % and the urban heat island was
reduced by 1-2 ºC.
In this thesis the role that the thermal properties of different materials play in the
calculation of the total sensible heat flux exchanged with the atmosphere in a heterogeneous
176
Capítulo 4
urban area has been investigated (Salamanca et al., 2009). When we compute this heat flux,
the thermal properties of the material most representative have to be known. The standard
way to derive them is averaging the different thermal properties of the different materials
presents taking into account the percentage of area that each of them covers. In this thesis we
propose to compute the thermal properties in a different way with the objective to reduce the
difference between the sum of fluxes exchanged between the surfaces and the atmosphere and
the heat flux obtained with the representative material. A simple situation has been considered
where the analytical solution was known and the sum of the heat fluxes obtained with the
different materials has been equal to the one that would be obtained with the new material. In
this way we have derived the thermal properties of the representative material and several
simulations have demonstrated that the error in the calculation of the sensible heat flux is
reduced by about 50 % in real situations.
The work presented in this thesis contributes to the development of a numerical tool is
able to describe in detail the impact of the cities in the atmosphere. With this kind of tools
spatial distributions of the anthropogenic heat and energy consumption can be derived.
Thanks to the work done in this thesis we can contribute to the development of cities more
energetically sustainable.
177
Capítulo 4
178
Capítulo 5
CAPÍTULO 5
CONCLUSIONES Y FUTURAS LINEAS DE INVESTIGACIÓN
179
Capítulo 5
5.1 Conclusiones
El estudio del clima urbano es una de las áreas recientes con mayor interés por parte
de la comunidad científica ya que la mayoría de la población vive en las ciudades. Hace
algunas décadas sólo era posible estudiar el impacto de las ciudades en el clima regional a
través de las observaciones. Sin embargo, gracias al aumento de la capacidad de cálculo de los
ordenadores y al desarrollo de los modelos numéricos, hoy en día podemos modelizar el
impacto de las ciudades en la atmósfera. Las primeras parametrizaciones urbanas aparecieron
en la década de los noventa. Estos primeros esquemas eran simples y no ofrecían una
descripción detallada de la ciudad. A pesar de esto se obtienen buenos resultados con ellas y
se siguen utilizando cuando el objetivo no es el estudio del clima urbano. A medida que las
interacciones de los edificios con la atmósfera se fueron incorporando en las
parametrizaciones urbanas, éstas aumentaron en complejidad. A principios del siglo XXI
prácticamente la totalidad de los esquemas urbanos consideraba el atrapamiento radiativo que
tiene lugar en el canyon urbano y resolvían los diferentes tipos de superficies (verticales y
horizontales) presentes en cualquier ciudad. Aparecen asimismo las parametrizaciones más
avanzadas multicapa que permitían una interacción directa con la capa límite atmosférica. El
efecto de las superficies en la temperatura, el viento y la energía cinética turbulenta eran
consideradas en este tipo de esquemas. Con estos esquemas modernos se había logrado dar un
paso hacia delante ya que eran capaces de distinguir las heterogeneidades presentes en una
ciudad y reproducían satisfactoriamente la isla de calor urbana. Sin embargo, el calor
(sensible/latente) generado por las distintas actividades humanas (tráfico, industria, sistemas
de aire acondicionado, etc.) no estaba considerado. A lo sumo se añadía un perfil diario de
calor al flujo sensible total. En esta tesis se resuelve teóricamente una de las fuentes más
importantes de calor antropogénico (el calor originado por los sistemas de aire acondicionado
180
Capítulo 5
en las grandes ciudades) y se estudia su papel en el clima urbano.
En estudios de ingeniería de la edificación se hace uso de software avanzado
(EnergyPlus Energy Simulation Program, UIUC LBNL., 2005) para el estudio y el diseño de
los edificios de nuestras ciudades. Estos programas permiten un análisis detallado del
consumo energético que tiene lugar en los edificios dependiendo de las condiciones
meteorológicas externas, pero sin tener en cuenta la retroalimentación existente entre los
flujos de calor (sensible/latente) y la atmósfera. Además al ser programas muy detallados,
necesitan de muchos parámetros para poder resolver un edificio particular. Sin embargo, en el
estudio del clima urbano nos movemos en escalas espaciales mucho mayores que el tamaño
de un edificio (entorno a ~ 1 km ) y estamos interesados en resolver una ciudad completa para
poder estudiar su impacto en la atmósfera a escala regional. Por consiguiente, se ha optado en
esta tesis por desarrollar un modelo más simple (Building Energy Model, BEM) que resuelva
energéticamente los edificios y que sea fácilmente integrable en una parametrización urbana
para simulaciones del clima urbano. Los principales fenómenos de transferencia de calor que
se resuelven en el modelo BEM y para cada planta de un edificio son:
•
la difusión del calor a través de las paredes, suelos y tejados,
•
la ventilación natural, así como la reflexión y emisión radiativa que tiene lugar
entre las superficies interiores de los muros,
•
el calor generado por los equipos domésticos y las personas,
•
el flujo de calor intercambiado por los sistemas de aire acondicionado y el
exterior.
De esta forma el calor generado por los sistemas de aire acondicionado (al fijar una
temperatura de confort interior) puede ser calculado para cada planta de un mismo edificio
tipo.
181
Capítulo 5
La comparación de BEM con otros programas desarrollados expresamente para el
análisis térmico de edificios ha dado resultados satisfactorios, demostrando que el modelo
energético es capaz de resolver los principales fenómenos de transferencia de calor e
intercambio con el exterior. Posteriormente el esquema BEM ha sido integrado en la
parametrización urbana BEP (Building Effect Parameterization) y se ha podido estudiar su
impacto en la temperatura del aire. Estos primeros resultados muestran que el resolver
energéticamente los edificios es importante y debería ser considerado cuando se estudia el
clima urbano. Además, diferentes estrategias de ahorro energético como el cambio del albedo
de los tejados, el uso de materiales aislantes y la eliminación del calor proveniente de los
sistemas de aire acondicionado han sido evaluadas cuantitativamente.
Posteriormente se ha procedido al acoplamiento del esquema urbano BEP+BEM en el
modelo atmosférico WRF y se han simulado las ciudades de Houston (Texas, US) y Madrid
(España). Con la nueva parametrización se ha cuantificado el impacto del calor antropogénico
y se han evaluado diferentes estrategias de mitigación del consumo energético y de la isla de
calor. Los resultados han sido satisfactorios y se ha observado que el calor proveniente de los
sistemas de aire acondicionado (en días especialmente calurosos) puede aumentar la
temperatura del aire en un par de grados centígrados y modificar claramente la estructura
vertical de la capa límite atmosférica. Las diferentes estrategias de ahorro energético
(incremento del albedo de los tejados, uso de materiales aislantes en los muros y eliminación
del calor originado por los sistemas de aire acondicionado) cuando fueron consideradas
conjuntamente en una sola redujeron la intensidad de la isla de calor en toda la ciudad y en
algunas zonas en un par de grados centígrados. Estos resultados ponen de manifiesto que el
calor antropogénico debería considerarse a la hora de estimar la concentración de
contaminantes, ya que este calor favorece la inestabilidad atmosférica y por consiguiente la
182
Capítulo 5
mezcla turbulenta.
En este trabajo vemos que herramientas numéricas como WRF junto con
parametrizaciones urbanas detalladas del tipo BEP+BEM permiten evaluar diferentes
estrategias de ahorro del consumo energético y cuantificar el calor antropogénico tanto
espacial como temporalmente, algo impracticable hace algunos años. Este tipo de
herramientas permitirán a los planificadores urbanos evaluar diferentes escenarios futuros a la
hora de diseñar el crecimiento de nuestras ciudades.
5.2 Futuras líneas de investigación
Las futuras líneas de investigación son numerosas y trataremos de presentar las más
relevantes que se pueden llevar a cabo en esta última sección de la tesis.
•
Explotar en su totalidad la información muy detallada sobre morfología
urbana que está disponible para un número creciente de ciudades (tipo la
existente en NUDAPT). En las simulaciones sobre Houston hemos visto que
para el cálculo del consumo energético la información morfológica detallada
de la ciudad mejoró notablemente los resultados, pero un estudio exhaustivo
del impacto en las variables meteorológicas en diferentes condiciones y
períodos más largos no se ha llevado a cabo y requiere de un mayor análisis.
•
Investigar el impacto de la calefacción en invierno. BEP+BEM tiene las
potencialidades para calcular el consumo energético y los flujos de calor
antropogénico en períodos invernales pero este tema todavía no ha sido
explorado. En la literatura científica no es frecuente encontrar estudios del
clima urbano en períodos no estivales y estimaciones de la isla de calor y del
consumo energético podrían llevarse a cabo en estos períodos de más bajas
temperaturas.
183
Capítulo 5
•
Relaciones con la calidad del aire. Es sabido que la calidad del aire se ve
afectada por los cambios de las propiedades del suelo urbano (aumento zonas
verdes, cambio del albedo, cambio de las propiedades térmicas de los
materiales, etc.) y por el impacto del calor antropogénico porque afectan los
flujos de calor en la superficie, y por consiguiente la estructura de la capa de
mezcla. Estudiar estos efectos podría ayudar en la predicción de la
contaminación atmosférica a escala regional.
• Profundizar el estudio del impacto de los edificios y el calor antropogénico
sobre la estructura de la capa limite atmosférica. Para ello se hace necesario
acoplar nuevos esquemas turbulentos de cierre de la capa límite atmosférica a
los esquemas urbanos BEP y BEP+BEM. De este modo estudios de
sensibilidad de las diferentes variables meteorológicas a estos esquemas se
podrían llevar a cabo.
184
Capítulo 5
185
Capítulo 5
CHAPTER 5
CONCLUSIONS AND FUTURE RESEARCH LINES
186
Capítulo 5
5.1 Conclusions
The study of the urban climate has attracted an increasing interest from the scientific
community since the majority of the population lives in the cities. Some decades ago, it was
possible to study the impact of the cities on the urban climate only through observations.
Thanks to the increase in the computing power and the development of numerical models,
nowadays we can model the impact of the cities on the atmosphere. The first urban schemes
appeared in the 90´s. These first parameterizations were simple and did not present a detailed
description of the cities. In spite of this, good results for the different meteorological variables
were obtained and they continue to be used when the goal of the study is not the urban
climate. When the interactions of the buildings with the atmosphere were taking into account,
the urban schemes increased in complexity. At the beginning of the 21´s century, the totality
of the urban schemes considered the radiative trapping that takes place in the urban canyon
and accounted for the different urban surfaces (vertical and horizontal) present in the city. The
most advanced urban parameterizations were multilayer and the direct interaction with the
planetary boundary layer was allowed. The effect of the urban surfaces in the potential
temperature, wind speed and turbulent kinetic energy were considered in these schemes. With
these moderns schemes it was possible to distinguish the heterogeneities in the cities and
reproduce satisfactorily the urban heat island phenomenon. Nevertheless, the heat
(sensible/latent) generated by the different human activities (traffic, industry, air conditioning
facilities, etc.) was not considered in these models. At most a daily profile of heat was added
to the total sensible heat flux. In this thesis a methodology to estimate one of the most
important sources of anthropogenic heat is proposed and its impact in the urban climate is
studied.
In building engineering the use of advanced software (EnegyPlus Energy Simulation
187
Capítulo 5
Program, UIUC LBNL., 2005) for building design is common. These programs allow a
detailed analysis of the energy consumption that takes places in the buildings depending on
the exterior meteorological conditions, but without considering the existing feedbacks
between the heat fluxes (sensible/latent) and the atmosphere. Moreover, these programs need
a lot of parameters to be able to solve a particular building. In the study of the urban climate
we work at scales greater than a building, and we are interested in solving a city to study its
impact in the atmosphere at regional scale. Consequently, it has chosen to develop a simpler
model (Building Energy Model, BEM) that solves energetically the buildings and that it can
be easily implemented in an urban parameterization.
The principal heat transfer phenomenon’s that are solved in BEM and for every floor
of a building are;
•
the heat diffusion through the walls, roofs and floors,
•
the natural ventilation, as well as the reflexion of radiation and radiative
emission that takes place between the indoor surfaces,
•
the heat generated by the equipments and occupants,
•
the heat fluxes exchanged by the air conditioning systems and the exterior.
In this way the heat generated by the air conditioning systems (having fixed an indoor target
temperature) can be computed for every floor of the same buildings type.
The comparison of BEM with other programs developed for the thermal analysis of
buildings it has given satisfactorily results, showing that the energy model is able to solve the
principal heat transfer phenomena and exchanges with the exterior. The BEM scheme has
been integrated in the urban parameterization BEP (Building Effect Parameterization) and it
has been possible to study its impact in the air temperature. These first results show that to
solve energetically the buildings is important and it should be considered in studies of the
188
Capítulo 5
urban climate. In addition, different strategies of energy saving as increase of the albedo of
the roofs, use of insulating materials, and elimination of the heat coming from the air
conditioning systems have been evaluated quantitatively.
Later, we proceeded to the coupling of the urban scheme BEP+BEM in the
atmospheric WRFv3.2 model and the cities of Houston (Texas, US) and Madrid (Spain) have
been simulated. With the new urban scheme the impact of the anthropogenic heat has been
quantified, and different strategies to reduce the energy consumption and to mitigate the urban
heat island have been evaluated. The results are satisfactory and it has been observed that the
heat coming from the air conditioning systems (specially in warm days) can increase the air
temperature by a couple of degrees Celsius and clearly modify the vertical structure of the
planetary boundary layer. The different strategies of energy saving (increase of the albedo of
the roofs, use of insulating materials in the walls, and elimination of the heat coming from the
air conditioning systems) when they were considered together in only one strategy, reduced
the urban heat island by a couple of degrees Celsius in some areas of the city. These results
indicate that the anthropogenic heat should be considered to estimate the concentration of
pollutants since this heat favours the turbulence mixing.
In this work, it has been shown that numerical tools like WRF together with detailed
urban schemes like the one developed in this thesis (BEP+BEM), allow to evaluate different
strategies of energy saving and quantify the anthropogenic heat spatially and temporally,
something impracticable some years ago. This type of tools will allow the urban planners to
evaluate different future scenarios at the moment to design cities development.
5.2 Future research lines
The future research lines are numerous and we will try to present the most relevant
that can be carried out in this last section of the thesis.
189
Capítulo 5
•
Take full advantage of the urban morphological data (like NUDAPT) that are
becoming available for an increasing number of cities. In the simulations over
Houston, we have seen that for the calculation of the energy consumption the detailed
morphological information improved notably the results, but an exhaustive study of
the impact on the meteorological variables for different meteorological conditions has
not been carried out and it needs a deeper analysis.
•
Study the impact of heating in winter. BEP+BEM has the potencialities to estimate the
impact of heating in winter periods but it has not been explored yet. In the scientific
literature is not frequent to find studies of the urban climate in winter periods, and
estimations of the energy consumption and urban heat island in these periods of lower
temperatures could be carried out.
•
Interactions with air quality. It is known that the urban air quality is affected by the
modifications of the surface properties (increase of green zones, change of the albedo,
change in the thermal properties of the materials, etc.) and for the impact of the
anthropogenic heat because of the changes in surface heat fluxes, and consequently,
on the vertical structure of the PBL. Studying these effects might help in the
prediction of pollutant concentrations at regional scale.
•
Deepen the study of the impact of buildings and anthropogenic heat fluxes on the
structure of the PBL. To do this it becomes necessary to link others PBL schemes to
the urban parameterizations BEP and BEP+BEM. In this way, studies of sensitivity to
the different meteorological variables could be carried out.
190
Referencias
191
Referencias
Referencias
BLAST-3.0-1981. The Building Loads Analysis and System Thermodynamics Program,
Users Manual, U. S. Army Construction Engineering Research Laboratory,
Champaign, Illinois, March.
Bornstein, R. D., 1968: Observations of the Urban Heat Island Effect in New York City.
Journal of Applied Meteorology, 7, 575-582.
Bougeault, P., and P. Lacarrère, 1989: Parameterization of orography-induced turbulence in a
mesobeta-scale model. Mon. Wea. Rev., 117, 1872-1890.
Burian, Steven J., and W. S. Han, 2003: Morphological Analysis using 3D Building
Databases: Houston, Texas. LA-UR-03-8633, Los Alamos National Laboratory.
Ching, J., and Co-authors, 2009: National Urban Database and Access Portal Tool. BAMS, 90,
1157-1168.
EIA, Annual Energy Review 2005, Energy Information Administration, U. S. Department of
Energy, Washington, DC, DOE/EIA-0384.
Georgakis, C., and M. Santamouris, 2009: The Vertical Stratification of Air Temperature in
the Center of Athens. Journal of Applied Meteorology and Climatology, 49: 12191231.
Godowitch, J. M., Ching, J. K. S., Clarke, J. F., 1985: Evolution of the Nocturnal Inversion
Layer at an Urban and Nonurban Location. Journal of Applied Meteorology, 24, 791805.
Grell, G. A., Steven E. Peckham, R. Schmitz, S. A. McKeen, G. Frost, W. C. Skamarock, and
B. Eder, 2005: Fully coupled “online” chemistry within the WRF model. Atmospheric
Environment, 39, 6957-6975.
Grimmond, C. S. B., and Co-authors, 2010a: The international Urban Energy Balance Model
192
Referencias
Comparison Project. First results from Phase 1. J. Appl. Meteorol. Climatol., 49, 12681292.
Haeger-Eugensson, M., and B. Holmer, 1999: Advection caused by the urban heat island
circulation as a regulating factor on the nocturnal urban heat island. Int. J. Climatol.,
19: 975-988.
Heiple, S., and David J. Sailor, 2008: Using building energy simulation and geospatial
modelling techniques to determine high resolution building sector energy consumption
profiles. Energy and Buildings, 40, 1426-1436.
Jacobson, M. Z., 1999: Fundamentals of Atmospheric Modeling. Published by Cambridge
University Press.
Janjic, Z. I., 1994: The step-mountain eta coordinate: Further development of the convection,
viscous sublayer, and turbulent closure schemes. Mon. Wea. Rev., 122, 927-945.
Kikegawa, Y., Y. Genchi, H. Yoshikado, and H. Kondo, 2003: Development of a numerical
simulation
system
toward
comprehensive
assessments
of
urban
warming
countermeasures including their impacts upon the urban buildings energy demands.
Applied Energy, 76, 449-466.
Kondo, H., and Y. Kikegawa, 2003: Temperature Variation in the Urban Canopy with
Anthropogenic Energy Use. Pure Appl. Geophys., 160, 317-324.
Kusaka, H., H. Kondo, Y. Kikegawa, and F. Kimura, 2001: A simple single-layer urban
canopy model for atmospheric models: comparison with multi-layer and slab models.
Boundary-Layer Meteorol., 101, 329-358.
Kusaka, H., and F. Kimura, 2004: Coupling a single-layer urban canopy model with a simple
atmospheric model: Impact on urban heat island simulation for an idealized case. J.
Meteor. Soc. Japan, 82, 67-80.
193
Referencias
Landsberg, H. E., 1981: The Urban Climate. Academic Press, 275 pp. ISBN-13: 9780124359604.
Lemone, M. A., F. Chen, M. Tewari, J. Dudhia, B. Geerts, Q. Miao, R. L. Coulter, and R. L.
Grossman, 2010: Simulating the IHOP_2002 Fair-Weather CBL with the WRF-ARWNoah Modeling System. Part-II: Structures from a Few Kilometers to 100 km across.
Mon. Wea. Rev., 138, 745-764.
Liu, Y., F. Chen, T. Warner, and J. Basara, 2006: Verification of a Mesoscale DataAssimilation and Forecasting System for the Oklahoma city Area During the Joint
Urban 2003 Field Project. Journal of Applied Meteorology and Climatology, 45, 912–
929.
Martilli, A., A. Clappier, and M. W. Rotach, 2002: An urban surface exchange
parameterization for mesoscale models. Boundary-Layer Meteorol., 104, 261-304.
Masson, V., 2000: A physically-based scheme for the urban energy budget in atmospheric
models. Boundary-Layer Meteorol., 94, 357-397.
Ohashi, Y., Y. Genchi, H. Kondo, Y. Kikegawa, H. Yoshikado, Y. Hirano, 2007: Influence of
Air-Conditioning Waste Heat on Air Temperature in Tokyo during Summer:
Numerical Experiments Using an Urban Canopy Model Coupled with a Building
Energy Model. Journal of Applied Meteorology and Climatology, 46, 66-81.
Oke, T. R., 1981: Canyon geometry and the nocturnal urban heat island: Comparison of scale
model and field observations. Int. J. Climatol., 1, 237-254.
Oke, T. R., 1988: The urban energy balance. Prog. Phys. Geogr., 12, 471-508.
Oke, T. R., 1995: The heat island of the urban boundary layer: Characteristics, causes and
effects. Wind Climate in Cities, J. E. Cermak et al., Kluwer Academic Publishers, 81107.
194
Referencias
Oleson, K. W., G. B. Bonan, J. Feddema, and T. Jackson, 2010: An examination of urban heat
island characteristics in a global climate model. Int. J. Climatol. Doi:
10.1002/joc.2201.
Rotach, M. W., 1993: Turbulence close to a rough urban surface part I: Reynolds stress.
Boundary-Layer Meteorol, 65, 1-28.
Rotach, M. W., and Co-authors, 2005: BUBBLE-an urban boundary layer meteorology
project. Theor. Appl. Climatol., 81, 231-261.
Roth, M., and T. R. Oke, 1993: Turbulent transfer relationships over an Urban Surface I:
Spectral characteristics. Quart. J. Roy. Meteorol. Soc., 119, 1071-1104.
Salamanca, F., E. S. Krayenhoff, and A. Martilli, 2009: On the Derivation of Material
Thermal Properties Representative of Heterogeneous Urban Neighbourhoods. Journal
of Applied Meteorology and Climatology, 48, 1725-1732.
Salamanca, F., A. Krpo, A. Martilli, and A. Clappier, 2010a: A New Building Energy Model
coupled with an urban canopy parameterization for urban climate simulations-part I.
formulation, verification, and sensitivity analysis of the model. Theor. Appl. Climatol.,
99, 331-344.
Salamanca, F., and A. Martilli, 2010: A New Building Energy Model coupled with an urban
canopy parameterization for urban climate simulations-part II. Validation with one
dimension off-line simulations. Theor. Appl. Climatol., 99, 345-356.
Salamanca, F., A. Martilli, M. Tewari, and F. Chen, 2010b: A study of the urban boundary
layer using different urban parameterizations and high-resolution urban canopy
parameters with WRF. Journal of Applied Meteorology and Climatology. (accepted).
Salamanca, F., A. Martilli, and C. Yagüe, 2010c: A numerical study of the urban boundary
layer over Madrid during the DESIREX (2008) campaign with WRF and an evaluation
195
Referencias
of simple mitigation strategies of the UHI. Submitted to Atmospheric Environment.
Sobrino, J. A., and Co-authors, 2009: DESIREX 2008 Final Report Contract nº: 21717/08/ILG (http://www.uv.es/desirex).
Stull, R. B., 1988: An Introduction to Boundary Layer Meteorology. Dordrecht, Boston,
London: Kluwer Academic Publishers, 666 pp.
UIUC, LBNL., 2005: EnergyPlus Engineering Reference: The Reference to EnergyPlus
Calculations, U.S. Department of Energy.
Uno, I., S. Wakamatsu, H. Ueda, and A. Nakamura, 1988: An observational study of the
structure of the nocturnal urban boundary layer. Boundary-Layer Meteorol., 45, 59-82.
Walton, G. N. 1983. Thermal Analysis Research Program (TARP) Reference Manual, U. S.
Department of Commerce, National Bureau of Standards, National Engineering
Laboratory, Washington, DC, March.
196
Appendices
197
Appendices
APPENDIX A
This appendix contains several abstracts of publications co-authored by the author of this
thesis.
198
Boundary-Layer Meteorol (2010) 136:105–127
DOI 10.1007/s10546-010-9491-2
ARTICLE
On the Impact of Anthropogenic Heat Fluxes
on the Urban Boundary Layer: A Two-Dimensional
Numerical Study
Andrea Krpo · Francisco Salamanca ·
Alberto Martilli · Alain Clappier
Received: 3 March 2009 / Accepted: 23 March 2010 / Published online: 21 April 2010
© Springer Science+Business Media B.V. 2010
Abstract The heat generated in buildings and the manner in which this heat is exchanged
with the ambient environment can play an important role in urban climate. Recent studies
have shown that anthropogenic heat from air-conditioning facilities can increase the exterior
ambient temperature and should be taken into account for a more complete urban heat island
(UHI) mitigation study. For this purpose, the first part of the present work is focused on
the coupling of a new building energy model (BEM) and an urban canopy parameterisation
(UCP). The new scheme is implemented in a finite volume mesoscale model (MM) and
tested in a two-dimensional (2D) configuration of a city over flat terrain. A sensitivity study
is performed with respect to different parameters in order to test the simulation system and
enhance the understanding of the possible impacts of the BEM on the exterior microclimate.
Keywords Anthropogenic heat · Building energy model · Urban canopy parameterisation ·
Urban heat island
1 Introduction
Urban regions are perhaps the most complex of all microclimates. Nowadays more than 50%
of the global population can be classified as urban, and this proportion is forecast to increase
to 75% during the next twenty five years. So it is not surprising that these areas have been
A. Krpo (B)
HBI Haerter Ltd., Thunstrasse 32, 3005 Bern, Switzerland
e-mail: [email protected]
F. Salamanca · A. Martilli
CIEMAT, Avenida Complutense 22, 28040 Madrid, Spain
A. Clappier
Université de Strasbourg, Laboratoire Image Ville Environnement 3,
rue de l’Argonne, 67000 Strasbourg, France
123
199
INTERNATIONAL JOURNAL OF CLIMATOLOGY
Int. J. Climatol. (2010)
Published online in Wiley InterScience
(www.interscience.wiley.com) DOI: 10.1002/joc.2158
The integrated WRF/urban modelling system: development,
evaluation, and applications to urban environmental
problems
Fei Chen,a * Hiroyuki Kusaka,b Robert Bornstein,c Jason Ching,d† C. S. B. Grimmond,e
Susanne Grossman-Clarke,f Thomas Loridan,e Kevin W. Manning,a Alberto Martilli,g
Shiguang Miao,h David Sailor,i Francisco P. Salamanca,g Haider Taha,j Mukul Tewari,a
Xuemei Wang,k Andrzej A. Wyszogrodzkia and Chaolin Zhangh,l
a National Center for Atmospheric Research, Boulder, CO, USA
Center for Computational Sciences, University of Tsukuba, Tsukuba, Japan
c Department of Meteorology, San Jose State University, San Jose, CA, USA
d National Exposure Research Laboratory, ORD, USEPA, Research Triangle Park, NC, USA
Environmental Monitoring and Modelling, Department of Geography, King’s College London, London, UK
f Global Institute of Sustainability, Arizona State University, Tempe, AZ, USA
g Center for Research on Energy, Environment and Technology, Madrid, Spain
h Institute of Urban Meteorology, China Meteorological Administration, Beijing, China
i Mechanical and Materials Engineering Department, Portland State University, Portland, OR, USA
j Altostratus Inc., Martinez, CA, USA
k Department of Environmental Science, Sun Yat-Sen University, Guangzhou, China
l Department of Earth Sciences, National Natural Science Foundation of China, Beijing, China
b
e
ABSTRACT: To bridge the gaps between traditional mesoscale modelling and microscale modelling, the National Center
for Atmospheric Research, in collaboration with other agencies and research groups, has developed an integrated urban
modelling system coupled to the weather research and forecasting (WRF) model as a community tool to address urban
environmental issues. The core of this WRF/urban modelling system consists of the following: (1) three methods with
different degrees of freedom to parameterize urban surface processes, ranging from a simple bulk parameterization to a
sophisticated multi-layer urban canopy model with an indoor–outdoor exchange sub-model that directly interacts with
the atmospheric boundary layer, (2) coupling to fine-scale computational fluid dynamic Reynolds-averaged Navier–Stokes
and Large-Eddy simulation models for transport and dispersion (T&D) applications, (3) procedures to incorporate highresolution urban land use, building morphology, and anthropogenic heating data using the National Urban Database and
Access Portal Tool (NUDAPT), and (4) an urbanized high-resolution land data assimilation system. This paper provides
an overview of this modelling system; addresses the daunting challenges of initializing the coupled WRF/urban model and
of specifying the potentially vast number of parameters required to execute the WRF/urban model; explores the model
sensitivity to these urban parameters; and evaluates the ability of WRF/urban to capture urban heat islands, complex
boundary-layer structures aloft, and urban plume T&D for several major metropolitan regions. Recent applications of this
modelling system illustrate its promising utility, as a regional climate-modelling tool, to investigate impacts of future
urbanization on regional meteorological conditions and on air quality under future climate change scenarios. Copyright 
2010 Royal Meteorological Society
KEY WORDS
urban modelling; mesoscale modelling; urban environmental issues; WRF urban model
Received 14 October 2009; Revised 22 March 2010; Accepted 28 March 2010
1.
Introduction
We describe an international collaborative research and
development effort between the National Center for
* Correspondence to: Fei Chen, National Center for Atmospheric
Research/RAL, PO Box 3000, Boulder, CO 80307-3000, USA.
E-mail: [email protected]
† The United States Environmental Protection Agency through its
Office of Research and Development collaborated in the research
described here. It has been subjected to Agency review and approved
for publication.
Atmospheric Research (NCAR) and partners with regard
to a coupled land-surface and urban modelling system for
the community weather research and forecasting (WRF)
model in this paper. The goal of this collaboration is
to develop a cross-scale modelling capability that can
be used to address a number of emerging environmental
issues in urban areas.
Today’s changing climate poses two formidable challenges. On the one hand, the projected climate change
by Intergovernmental Panel on Climate Change (IPCC)
Copyright  2010 Royal Meteorological Society
200
1268
JOURNAL OF APPLIED METEOROLOGY AND CLIMATOLOGY
VOLUME 49
The International Urban Energy Balance Models Comparison Project:
First Results from Phase 1
C. S. B. GRIMMOND,a M. BLACKETT,a M. J. BEST,b J. BARLOW,c J.-J. BAIK,d S. E. BELCHER,c
S. I. BOHNENSTENGEL,c I. CALMET,e F. CHEN,f A. DANDOU,g K. FORTUNIAK,h
M. L. GOUVEA,a R. HAMDI,i M. HENDRY,b T. KAWAI,j Y. KAWAMOTO,k
H. KONDO,l E. S. KRAYENHOFF,m S.-H. LEE,d T. LORIDAN,a A. MARTILLI,n
V. MASSON,o S. MIAO,p K. OLESON,f G. PIGEON,o A. PORSON,b,c Y.-H. RYU,d
F. SALAMANCA,n L. SHASHUA-BAR,q G.-J. STEENEVELD,r M. TOMBROU,g
J. VOOGT,s D. YOUNG,a AND N. ZHANGt
a
King’s College London, London, United Kingdom
b
Met Office, Exeter, United Kingdom
c
University of Reading, Reading, United Kingdom
d
Seoul National University, Seoul, South Korea
e
Laboratoire de Mécanique des Fluides, CNRS-Ecole Centrale de Nantes, Nantes, France
f
National Center for Atmospheric Research, Boulder, Colorado
g
National and Kapodistrian University of Athens, Athens, Greece
h
University of Łódź, Łódź, Poland
i
Royal Meteorological Institute, Uccle, Belgium
j
Ehime University, Matsuyama, Japan
k
The University of Tokyo, Tokyo, Japan
l
National Institute of Advanced Industrial Science and Technology, Tsukuba, Japan
m
University of British Columbia, Vancouver, British Columbia, Canada
n
CIEMAT, Madrid, Spain
o
CNRM-GAME Meteo France-CNRS, Toulouse, France
p
IUM, CMA, Beijing, China
q
Ben Gurion University of the Negev, Beer-Sheva, Israel
r
Wageningen University, Wageningen, Netherlands
s
University of Western Ontario, London, Ontario, Canada
t
Nanjing University, Nanjing, China
(Manuscript received 27 July 2009, in final form 4 February 2010)
ABSTRACT
A large number of urban surface energy balance models now exist with different assumptions about the
important features of the surface and exchange processes that need to be incorporated. To date, no comparison of these models has been conducted; in contrast, models for natural surfaces have been compared
extensively as part of the Project for Intercomparison of Land-surface Parameterization Schemes. Here, the
methods and first results from an extensive international comparison of 33 models are presented. The aim of
the comparison overall is to understand the complexity required to model energy and water exchanges in
urban areas. The degree of complexity included in the models is outlined and impacts on model performance
are discussed. During the comparison there have been significant developments in the models with resulting
improvements in performance (root-mean-square error falling by up to two-thirds). Evaluation is based on a
dataset containing net all-wave radiation, sensible heat, and latent heat flux observations for an industrial area in
Vancouver, British Columbia, Canada. The aim of the comparison is twofold: to identify those modeling approaches that minimize the errors in the simulated fluxes of the urban energy balance and to determine the
degree of model complexity required for accurate simulations. There is evidence that some classes of models
perform better for individual fluxes but no model performs best or worst for all fluxes. In general, the simpler
models perform as well as the more complex models based on all statistical measures. Generally the schemes
have best overall capability to model net all-wave radiation and least capability to model latent heat flux.
Corresponding author address: Sue Grimmond, Environmental Monitoring and Modelling Group, Department of Geography, King’s
College London, London, WC2R 2LS, United Kingdom.
E-mail: [email protected]
DOI: 10.1175/2010JAMC2354.1
Ó 2010 American Meteorological Society
201
INTERNATIONAL JOURNAL OF CLIMATOLOGY
Int. J. Climatol. (2010)
Published online in Wiley Online Library
(wileyonlinelibrary.com) DOI: 10.1002/joc.2227
Initial results from Phase 2 of the international urban energy
balance model comparison
C. S. B. Grimmond,a * M. Blackett,a M. J. Best,b J.-J. Baik,c S. E. Belcher,d J. Beringer,e
S. I. Bohnenstengel,d I. Calmet,f F. Chen,g A. Coutts,e A. Dandou,i K. Fortuniak,j
M. L. Gouvea,a R. Hamdi,k M. Hendry,b M. Kanda,l T. Kawai,m Y. Kawamoto,n H. Kondo,o
E. S. Krayenhoff,p S.-H. Lee,c T. Loridan,a A. Martilli,q V. Masson,r S. Miao,s K. Oleson,h
R. Ooka,n G. Pigeon,r A. Porson,b,d Y.-H. Ryu,c F. Salamanca,q G.J. Steeneveld,t M. Tombrou,i
J. A. Voogt,u D. T. Younga and N. Zhangv
a
Department of Geography, King’s College London, London WC2R 2LS, UK
b Met Office, FitzRoy Road, Exeter, EX1 3PB, UK
c School of Earth and Environmental Sciences, Seoul National University, Seoul 151-742, Republic of Korea
d Department of Meteorology, University of Reading, Reading, RG6 6BB, UK
e School of Geography and Environmental Science, Monash University, Melbourne, Vic, 3800, Australia
f Equipe Dynamique de l’Atmosphère Habitée Laboratoire de Mécanique des Fluides (UMR CNRS 6598) Ecole Centrale de Nantes, B.P. 92101,
F-44321 NANTES Cedex 3, France
g Research Applications Laboratory, National Center for Atmospheric Research, Boulder, Colorado, 80307, USA
h Earth System Laboratory, National Center for Atmospheric Research, Boulder, Colorado, 80307, USA
i National and Kapodistrian University of Athens, Faculty of Physics, Department of Environmental Physics and Meteorology, Laboratory of
Meteorology, Building Physics V, University Campus, 157 84 Athens, Greece
j Department of Meteorology and Climatology University of Lodz Narutowicza 88 Lodz Poland 90139
k Royal Meteorological Institute, Department II, section 3 Avenue Circulaire, 3, B-1180 Brussels, Belgium
l Department of International Development Engineering. Tokyo Institute of Technology,2-12-1-14-9, O-okayama, Megro-KU, Tokyo, Japan
m Research Center for Environmental Risk, National Institute for Environmental Studies, 16-2 Onogawa, Tsukuba-City, Ibaraki, 305-8506 Japan
n School of Engineering, The University of Tokyo, 7-3-1 Hongo, Hongo, Bunkyo-ku, Tokyo, 113-8656, Japan
o Research Institute for Environmental Management Technology, National Institute of Advanced Industrial Sciecne and Technology,Tsukuba,
Ibaraki, 305-8569, JAPAN
p Department of Geography, University of British Columbia, Vancouver, British Columbia, V6T 1Z2, Canada
q Department of Environment, CIEMAT, Madrid, 28040, Spain
r CNRM-GAME, Météo France/CNRS, Toulouse, 31057 Cedex 1, France
s Institute of Urban Meteorology, China Meteorological Administration, Beijing, 100089, China
t Meteorology and Air Quality Section, Wageningen University, P.O. Box 47, 6700 AA Wageningen, The Netherlands
u Department of Geography, University of Western Ontario, London ON N6A 5C2 Canada
v School of Atmospheric Sciences, Nanjing University 22 Hankou Road, Nanjing, 210093, China
ABSTRACT: Urban land surface schemes have been developed to model the distinct features of the urban surface and
the associated energy exchange processes. These models have been developed for a range of purposes and make different
assumptions related to the inclusion and representation of the relevant processes. Here, the first results of Phase 2 from
an international comparison project to evaluate 32 urban land surface schemes are presented. This is the first large-scale
systematic evaluation of these models. In four stages, participants were given increasingly detailed information about an
urban site for which urban fluxes were directly observed. At each stage, each group returned their models’ calculated
surface energy balance fluxes. Wide variations are evident in the performance of the models for individual fluxes. No
individual model performs best for all fluxes. Providing additional information about the surface generally results in better
performance. However, there is clear evidence that poor choice of parameter values can cause a large drop in performance
for models that otherwise perform well. As many models do not perform well across all fluxes, there is need for caution in
their application, and users should be aware of the implications for applications and decision making. Copyright  2010
Royal Meteorological Society
KEY WORDS urban climate; energy balance; surface atmosphere exchanges; land surface modelling; sustainable cities;
radiation; turbulent heat fluxes; evaporation
Received 29 March 2010; Revised 18 August 2010; Accepted 21 August 2010
1.
Introduction
Land surface models (LSMs) parameterize energy exchanges between the surface and the atmosphere for a
* Correspondence to: C. S. B. Grimmond, Department of Geography,
King’s College London, London WC2R 2LS, UK.
E-mail: [email protected]
wide range of different land surface types (e.g. deciduous
trees, coniferous trees, grasses, bare soil, and urban).
They provide the lower boundary conditions (fluxes)
to meso- and global-scale atmospheric models and are
forced with meteorology from the overlying atmospheric
model. A wide variety of approaches are taken to model
the influence of the underlying land surface type. To
Copyright  2010 Royal Meteorological Society
202
Appendices
APPENDIX B
Numerical treatment to solve the heat diffusion equation using an energetic
balance as boundary conditions.
The diffusion equation in one dimension can be written as,
∂T
∂
λ ∂T
∂
∂T
=
=
K
,
∂t ∂z ρC P ∂z
∂z
∂z
(B1)
where T (K ) is the temperature, λ (W / mK ) is the conductivity, ρ (kg / m 3 ) is the density, and
C P ( J / kgK ) is the heat capacity of the material. Grouping the different physical
parameters K =
λ
ρC P
, and supposing that the material is divided in n layers (see Fig. B.1), we
can discretize the diffusion equation in the following way for a particular indoor layer
1< k < n:
Tkm +1 − Tkm
T m +1 − Tkm +1
T m+1 − Tkm−1+1
1
=
K k +1 k +1
− Kk k
∆z k +1 + ∆z k
∆z k + ∆z k −1
∆t
∆z k
2
2
.
(B2)
Grouping the previous equation we can write the eq. (B2) in the form;
Tkm = a (0, k )Tkm +1 + a (1, k )Tkm+1+1 + a (−1, k )Tkm−1+1 ,
(B3)
where
a (0, k ) = 1 +
a (1, k ) = −
2∆tK k +1
2∆tK k
+
∆z k (∆z k +1 + ∆z k ) ∆z k (∆z k + ∆z k −1 )
2∆tK k +1
∆z k (∆z k +1 + ∆z k )
a (−1, k ) = −
.
(B4)
2∆tK k
∆z k (∆z k + ∆z k −1 )
203
Appendices
Figure B.1. Schematic picture showing the different layers that compose an arbitrary material.
The treatment at the boundary conditions is different and requires the knowledge of
the net heat fluxes φ1 and φ 2 (W / m 2 ) at the external boundary surfaces. The criterion followed
here is that a positive value means a gain for the surface. For the first layer and using the
previous eq. (B1) can write:
T1m +1 − T1m
T m +1 − T1m +1
φm
1
.
=
K2 2
+ 1
∆z1 + ∆z 2
∆t
∆z1
ρ1C P1
2
(B5)
Again, after some algebraic operations we transform the eq. (B5) in
∆tφ1m
T +
= a (1,1)T2m +1 + a (0,1)T1m +1 ,
∆z1 ρ1C P1
m
1
(B6)
204
Appendices
where
a (0,1) = 1 +
2 K 2 ∆t
∆z1 (∆z 2 + ∆z1 )
2∆tK 2
a (1,1) = −
∆z1 (∆z 2 + ∆z1 )
.
(B7)
For the other external layer the treatment is similar, and the use of the eq. (B1) leads to
Tnm +1 − Tnm
φ 2m
Tnm+1 − Tnm−1+1
1
=
− Kn
∆z n + ∆z n −1
∆t
∆z n ρ n C Pn
2
,
(B8)
and finally we derive that
∆tφ 2m
T +
= a (0, n)Tnm +1 + a (−1, n)Tnm−1+1
∆z n ρ n C Pn
m
n
(B9)
where
a (0, n) = 1 +
2∆tK n
∆z n (∆z n + ∆z n −1 )
2∆tK n
a (−1, n) = −
∆z n (∆z n + ∆z n−1 )
.
(B10)
Proceeding in this way, the problem is solved inverting the three-diagonal matrix
system AX = B , where
T1m +
∆tφ1m
∆z1 ρ1C P1
T2m
B=
.
,X =
.
T1m+1
a (0,1)
.
a (−1,2) a (0,2)
.
.
Tnm−1
.
∆tφ 2m
T +
∆z n ρ n C Pn
Tnm+1
m
n
,A= 0
a (1,1)
0
0.......................0
a (1,2) 0.......................0
a (−1,3) a (0,3) a (1,3) 0............0
(B11)
...............................................................
0 .....................................0 a (−1, n) a (0, n)
205
Appendices
By inverting the matrix system X = A −1 B it is possible to obtain the temperature for every
layer of the material at time m + 1 knowing the value at time m .
206
CURRICULUM VITAE
Family Names: Salamanca Palou
Name: Francisco
Place of Birth: Mallorca (Balearic Islands, Spain)
Professional position
Center:
Research
Centre
for
Energy,
Environment
and
Technology
(CIEMAT,
http://www.ciemat.es/).
Department: Environmental Department, Atmospheric Pollution Modelling Division.
Address: Avenida Complutense 22, Madrid 28040, Spain.
Telephone: +34-913466299
e-mail: [email protected]
Position: recipient of a grant since January 2007.
Research Lines
Atmospheric sciences, mesoscale modelling, urban climate, urban boundary layer, and open
to any creative physical-mathematical challenge.
Academic Degrees
‘Licenciado’ in Physical Sciences at University of the Balearic Islands (UIB), February 1999.
PhD in Physics “Development of numerical models to investigate the Urban Heat Island in
cities, and sensitivity study of different urban parameters”, University Complutense of
Madrid (UCM). Defense planned in December 2010.
207
Thesis Advisors: Dr. Alberto Martilli and Dr. Carlos Yagüe Anguís.
Participation in Research Projects
1.-
Urban
Surface
Energy
Balance:
Land
Surface
Scheme
Comparison
(www.kcl.ac.uk/ip/suegrimmond/model_comparison.htm)
2.- Mesoscale simulations of urban climate, and development of an evaluation technique of
Urban Heat Island mitigation strategies (funded by the Ministry of Environment of Spain, file
200800050084408).
Publications in international journals
1.- Salamanca, F., A. Krpo, A. Martilli, and A. Clappier, 2010. A New Building Energy
Model coupled with an Urban Canopy Parameterization for urban climate
simulations–Part I. Formulation, verification and a sensitive analysis of the model.
Theoretical and Applied Climatology, 99, 331-344.
2.- Salamanca, F., and A. Martilli, 2010. A New Building Energy Model coupled with an
Urban Canopy Parameterization for urban climate simulations–Part II. Validation with
one dimension off-line simulations. Theoretical and Applied Climatology, 99, 345356.
3.- Salamanca, F., E. S. Krayenhoff, and A. Martilli, 2009. On the derivation of material
thermal properties representative of heterogeneous urban neighbourhoods. Journal of
Applied Meteorology and Climatology, 48, 1725-1732.
4.- C.S.B. Grimmond, M. Blackett, M.J. Best, J. Barlow, J.-J. Baik, S.E. Belcher, S.I.
Bohnenstengel, I. Calmet, F. Chen, A. Dandou, K. Fortuniak, M.L. Gouvea, R.
Hamdi, M. Hendry, T. Kawai, Y. Kawamoto, H. Kondo, E.S. Krayenhoff, S.-H.
Lee, T.
Loridan, A.
Martilli, V.
Masson, S.
Miao, K.
Oleson, G.
Pigeon, A.
Porson, Y.-H. Ryu, F. Salamanca, L. Shashua-Bar, G.-J. Steeneveld, M. Tombrou, J.
208
Voogt, D. Young, N. Zhang. The International Urban Energy Balance Models
Comparison Project: First results from Phase 1 (2010). Journal of Applied
Meteorology and Climatology, 49, 1268-1292.
5.- Krpo, A., F. Salamanca, A. Martilli, and A. Clappier, 2010. On the impact of
anthropogenic heat fluxes on the urban boundary layer: a two-dimensional numerical
study. Boundary Layer Meteorology, 136, 105-127.
6.- Chen, F., Hiroyuki Kusaka, Robert Bornstein, Jason Ching, C.S.B. Grimmond, Susanne
Grossman-Clarke, Thomas Loridan, Kevin W. Manning, Alberto Martilli, Shiguang
Miao, David Sailor, Francisco P. Salamanca, Haider Taha, Mukul Tewari, Xuemei
Wang, Andrzej A. Wyszogrodzki, Chaolin Zhang, 2010. The integrated WRF/urban
modelling system: development, evaluation, and applications to urban environmental
problems. International Journal of Climatology. Doi: 10.1002/joc.2158.
7.- Salamanca, F., A. Martilli, M. Tewari, and F. Chen, 2010. A study of the urban boundary
layer using different urban parameterizations and high-resolution urban canopy
parameters with WRF. Journal of Applied Meteorology and Climatology. (In press).
8.- CSB Grimmond, M Blackett, MJ Best, J-J Baik, SE Belcher, J Beringer, SI
Bohnenstengel, I Calmet, F Chen, A Coutts, A Dandou, K Fortuniak, ML Gouvea, R
Hamdi, M Hendry, M Kanda, T Kawai, Y Kawamoto, H Kondo, ES Krayenhoff, S-H
Lee, T Loridan, A Martilli, V Masson, S Miao, K Oleson, R Ooka, G Pigeon, A
Porson, Y-H Ryu, F Salamanca, G-J Steeneveld, M Tombrou, JA Voogt, D Young, N
Zhang. Initial Results from Phase 2 of the International Urban Energy Balance
Comparison Project. International Journal of Climatology. Doi: 10.1002/joc.2227.
9.- Salamanca, F., A. Martilli, and C. Yagüe, 2010. A numerical study of the urban boundary
layer over Madrid during the DESIREX (2008) campaign with WRF and an evaluation
209
of simple mitigation strategies of the UHI. Atmospheric Environment. (submitted).
Oral and posters communications in Conferences
1.- F. Salamanca, A. Krpo, A. Martilli, A. Clappier. Implementation of a Building Energy
Model in an Urban Canopy Parameterization (poster). 7th Symposium on the Urban
Environment, San Diego, USA. September 2007.
2.- Sue Grimmond, M. Blackett, M. Best with Baik, J., Bohnenstengel, S., Calmet, I., Chemel,
C., Chen, F., Dandou, A., Fortuniak, K., Gouvea, M., Hamdi, R., Kondo, H.,
Krayenhoff, S., Lee, S., Loridan, T., Martilli, A., Masson, V., Miao, S., Oleson, K.,
Pigeon, G. Porson, A., Salamanca, F., Shashua-Bar, L., Steeveveld, G., Sugar, L.,
Trombou, M., Voogt, J., Zhang N. An international urban surface energy balance
model comparison study: first results (oral presentation). 8th Symposium on the
Urban Environment, Phoenix, Arizona USA. January 2009.
3.- F. Salamanca and A. Martilli. A detailed study of the different turbulent fluxes in an urban
environment considering a Building Energy Model coupled with an Urban Canopy
Parameterization (one dimension off-line simulations) (oral presentation). 7th
International Conference on Urban Climate (ICUC-7), Yokohama, Japan. Jun 2009
Grimmond CSB, Blackett M and Best M with Baik J, Bohnenstengel S, Calmet I, Chen F,
Danndou A, Fortuniak K, Gouvea M, Hamdi R, Hendry M, Kondo H, Krayenhoff S,
Lee S, Loridan T, Martilli A, Masson V, Miao S, Oleson K, Pigeon G, Porson A,
Salamanca F, Shashua-Bar L, Steeveveld G, Trombou M, Voogt J, Zhang N. Results
from the international urban surface energy balance model comparison study (oral
presentation). 7th International Conference on Urban Climate (ICUC-7),
Yokohama, Japan. Jun 2009.
5.- F. Salamanca, A. Martilli, M. Tewari, F. Chen, C. Yagüe. A study of the Urban Boundary
210
Layer considering different Urban Canopy Parameterizations and high resolution
urban databases with WRF (the case of Houston) (poster). European Geosciences
Union General Assembly 2010, Vienna, Austria. May 2010.
6.- F. Salamanca, A. Martilli, M. Tewari, F. Chen. A study of the Urban Boundary Layer
considering different Urban Canopy Parameterizations and high resolution urban
canopy parameters with WRF (the case of Houston) (oral presentation). Ninth
Symposium on the Urban Environment, August 2-6, 2010, Keystone, Colorado, USA.
August 2010.
7.- F. Salamanca, A. Martilli, C. Yagüe. A numerical study of the Urban Boundary Layer
over Madrid during the DESIREX (2008) campaign with WRF (oral presentation).
Ninth Symposium on the Urban Environment, August 2-6, 2010, Keystone,
Colorado, USA. August 2010.
Other
I have experience with graphical packages like FERRET, GRADS, and programming
language like FORTRAN. I am familiar with WRF (WPS, WRF-ARW, and ARW-post),
where I implemented the urban parameterization that I developed during my thesis, and that I
used to simulate Houston and Madrid. I am also familiar with the linux environment and
supercomputers (clusters).
Stages
(longer than four weeks)
1.- From May to July 2009. Implementation of a building energy model in an urban
parameterization in the atmospheric WRF model. National Center for Atmospheric Research
(NCAR), Boulder, CO, USA. The new urban scheme was included in the public release of
WRF (V3.2) from April 2, 2010.
211
2.- From July to August 2010. Development and test of a new version of the atmospheric
WRF model that is able to use urban morphology information point to point in the numerical
grid domain. National Center for Atmospheric Research (NCAR), Boulder, CO, USA.
Short stages (less than four weeks)
1.- From 9 to 15 December 2007. Ecole Polytechnique Federale de Lausanne (Switzerland).
2.- From 23 to 28 Jun 2008. Ecole Polytechnique Federale de Lausanne (Switzerland).
Referee for Building and Environment, and International Journal of Climatology.
212