Simulation of a Flash Flooding Storm at the Steep Edge

Transcription

Simulation of a Flash Flooding Storm at the Steep Edge
212
JOURNAL OF HYDROMETEOROLOGY
VOLUME 15
Simulation of a Flash Flooding Storm at the Steep Edge of the Himalayas*
ANIL KUMAR
NASA Goddard Space Flight Center, Greenbelt, and Earth System Science Interdisciplinary Center, University of Maryland,
College Park, College Park, Maryland
ROBERT A. HOUZE JR. AND KRISTEN L. RASMUSSEN
Department of Atmospheric Sciences, University of Washington, Seattle, Washington
CHRISTA PETERS-LIDARD
NASA Goddard Space Flight Center, Greenbelt, Maryland
(Manuscript received 25 October 2012, in final form 27 June 2013)
ABSTRACT
A flash flood and landslide in the Leh region of the Indus Valley in the Indian state of Jammu and
Kashmir on 5–6 August 2010 resulted in hundreds of deaths and great property damage. Observations
have led to the hypothesis that the storm, which formed over the Tibetan Plateau, was steered over the
steep edge of the plateau by 500-hPa winds and then energized by the ingestion of lower-level moist air,
which was approaching from the Arabian Sea and Bay of Bengal and rose up the Himalayan barrier. A
coupled land surface and atmospheric model simulation validates this hypothesized storm scenario, with
the model storm taking the form of a traveling mesoscale squall line with a leading convective line, trailing
stratiform region, and midlevel inflow jet. In this region, the development of a mesoscale storm over high
terrain is highly unusual, especially one in the form of a propagating squall line system. This unusual
storm occurrence and behavior could serve as a warning sign in flash flood prediction. The coupled atmosphere and land surface model showed that the excessive runoff leading to the flood and landslide were
favored by the occurrence of this unusual meteorological event coinciding temporally and spatially with
favorable hydrologic conditions. Additionally, the model simulations showed that previous rainstorms
had moistened the soil during the entire season and especially over the few days leading up to the Leh
flood, so the normally arid mountainsides were likely not able to rapidly absorb the additional rainfall of
the sudden 5 August squall line.
1. Introduction
Convective storms occurring at the steep edge of
broad high topography, such as the Rocky Mountains
and Himalayas, are notorious for producing surprising
and lethal flash floods. Such a storm occurred over the
* Supplemental information related to this paper is available at the Journals Online website: http://dx.doi.org/10.1175/
JHM-D-12-0155.s1.
Corresponding author address: Dr. Anil Kumar, Hydrological
Science Branch (Code 617), NASA Goddard Space Flight Center,
Greenbelt, MD 20770.
E-mail: [email protected]
DOI: 10.1175/JHM-D-12-0155.1
Ó 2014 American Meteorological Society
Himalayan range on 5 August 2010. It produced a devastating flash flood at the town of Leh in the high arid
Indus Valley in northwest India (Fig. 1). Rasmussen and
Houze (2012, hereafter RH) have recently described the
meteorological setting and sequence of events for this
storm. They arrived at a conceptual model of the storm
in which cells generated by diurnal heating over the high
terrain of the Tibetan Plateau gathered into a mesoscale
convective system (MCS), which the 500-hPa winds then
steered over the bank of the Plateau such that the storm
could then draw upon low-level moisture arriving from
the Arabian Sea and Bay of Bengal. They hypothesized
that the ingestion of this moisture then energized the
MCS just as it was passing over Leh and thus produced
heavy convective and/or stratiform rain over Leh and
FEBRUARY 2014
KUMAR ET AL.
213
Research and Forecasting Model (ARW-WRF, hereafter just WRF; Skamarock et al. 2008) coupled with
NASA’s Land Information System (LIS) model. We will
show that the conceptual model is borne out by numerical modeling, show the sources of moist air ingested
into the storm as it moves off the Plateau, and elucidate
the physical processes producing the rainfall and land
surface conditions that led to the flood.
2. Model description and experimental design
FIG. 1. (a) Topography near the city of Leh, India. The Indus
River flows to the northwest and is just south of the city of Leh.
Source: Google Maps. (b) The rivers of India and Pakistan. The
Indus River originates in China and flows through India and
Pakistan before reaching the Arabian Sea to the southwest.
the surrounding mountainsides. The flash flood was a
consequence of this heavy rain and runoff.1
Although RH’s conceptual model based on observations is consistent with the available data for this storm,
physical insight into the storm’s dynamics and precipitation-producing processes can best be derived from
a numerical model given the remote nature of the region
and limited observations of the flash flood. The purpose
of this paper is, therefore, to provide such insight via
a simulation with the Advanced Research Weather
1
In India, sudden heavy rain events such as the one described in
this paper are often referred to as ‘‘cloudbursts,’’ and many discussions online and elsewhere have referred to this event as the Leh
cloudburst. For example, see the official description by the Indian
Meteorological Department at http://www.imd.gov.in/doc/cloudburst-over-leh.pdf.
The simulations in this paper were carried out with the
National Aeronautics and Space Administration (NASA)
Unified Weather Research and Forecasting Model (NUWRF) in an effort to unify WRF with NASA’s existing
weather models and assimilation systems, such as Goddard Earth Observing System Model, version 5 (GEOS-5)
and LIS. Physical processes like cloud–aerosol–land
surface processes have been implemented in NU-WRF
in order to 1) better use WRF in scientific research and
2) allow for better use of satellite data for research
(short-term climate and weather). Detailed information
about the NU-WRF can found at https://modelingguru.
nasa.gov/community/atmospheric/nuwrf. The most important feature of NU-WRF for this study is the capability to couple with version 6.1 of LIS to provide a more
realistic land surface state.
We ran the LIS model on its own (uncoupled) to
generate input land surface values for the NU-WRF–
LIS coupled system to investigate the Leh flood event.
As described by Kumar et al. (2006), Peters-Lidard et al.
(2008), and Kumar et al. (2008), LIS integrates satelliteand ground-based observational data products with advanced land surface modeling to produce optimal fields
of land surface states and fluxes. Recent studies have used
this modeling framework to successfully investigate the
effects of soil moisture, albedo effects, and land cover in
various environments (Zaitchik et al. 2013; Santanello
et al. 2011; Tao et al. 2013), demonstrating the utility of
the coupled modeling system despite uncertainties in the
estimates. The LIS has evolved from two earlier efforts,
the North American Land Data Assimilation System
(NLDAS; Mitchell et al. 2004) and the Global Land Data
Assimilation System (GLDAS; Rodell et al. 2004), which
focused primarily on improving numerical weather prediction skill through the characterization of land surface
conditions. LIS not only consolidates the capabilities of
these two systems, but also enables a much larger variety
of configurations with respect to horizontal and spatial
resolution, input datasets, and choice of land surface
models. The coupling of LIS to WRF is both one way and
two way to allow hydrometeorological modeling for evaluating the impact of land surface processes on hydrologic
214
JOURNAL OF HYDROMETEOROLOGY
VOLUME 15
FIG. 2. Flowchart detailing the WRF-LIS coupling framework.
prediction. Figure 2 contains a flowchart of the LIS and
WRF-LIS systems. The uncoupled LIS simulation was
implemented with Tropical Rainfall Measuring Mission
(TRMM) precipitation forcing from January 2008 through
August 2010 to allow the model adequate spinup time
for more realistic soil conditions to develop over a longer time. The TRMM precipitation forcing data are the
3B42 version 6 products from the TRMM Multisatellite
Precipitation Analysis (TMPA) archive. The land surface fields generated were used as input to the coupled
WRF-LIS to simulate the Leh flash flood event. We have
conducted two simulations using the WRF-LIS coupled
model: one with two-nested domains (27 and 9 km) at a
coarser resolution that was conducted from 2 to 6 August
2010 and the other as a high-resolution simulation with
four-nested domains that was conducted from 5 to 6 August
2010. The purpose of conducting two simulations is
mainly because 1) the coarser simulation provides important information on larger-scale synoptic flows, moisture transport, and regional-scale precipitation patterns
that occurred from 2 to 5 August 2010 and 2) the very
high-resolution simulation provides important information about the storm structure that passed over Leh.
The boundary and initial conditions for the large-scale
atmospheric fields, soil parameters (moisture and temperature), and sea surface temperature (SST) are given
by the 18 3 18 6-hourly National Centers for Environmental Prediction (NCEP) Final (FNL) Operational
Global Analysis data (http://rda.ucar.edu/datasets/ds083.2/).
These forcing data were interpolated to the respective
NU-WRF grid using the WRF Preprocessing System
(http://www.mmm.ucar.edu/wrf/users/wpsv3/wps.html).
The two-way nested model domain is shown in Fig. 3:
the outer domain has 27-km horizontal grid spacing with
190 3 190 grid points along the x and y axes, the second
domain has 9-km horizontal grid spacing with 409 3 385
grid points, the third domain has 3-km horizontal grid
spacing with 688 3 487 grid points, and the fourth domain has 1-km horizontal grid spacing and is zoomed in
over Leh with 259 3 256 grid points. The model’s vertical layers are in terrain-following coordinates with 35
vertical layers from Earth’s surface to 10 hPa.
FIG. 3. Model nested domains (at 27-, 9-, 3-, and 1-km grid
spacing) with topography in the background. The city of Leh is
indicated with a white circle.
The high-resolution, four-nested domain simulation
was initialized at 0000 UTC 5 August 2010 and was run
until 0600 UTC 6 August 2010. The coupling of LIS to
WRF enables integrated land–atmosphere modeling
through both one- and two-way coupling, leading to
a hydrometeorological modeling capability that can be
used to evaluate the impact of land surface processes on
hydrologic prediction. The modeled convective storms
required 6–12 h of model spinup time for better prediction, so we initialized the model 15 h before the rain
event in Leh. The coupled WRF-LIS run was done in two
steps. First, the Noah land surface model initial conditions were generated by running a spinup integration
using the offline LIS model with uniform initial conditions from 1 January 2008 to 30 August 2010. Second, the
land conditions generated by LIS were used to initialize
the WRF-LIS model. The LIS and WRF-LIS model and
documentation are available on the LIS website (https://
modelingguru.nasa.gov). For the backward trajectory
analysis, which is used to track the air parcels to examine
the synoptic-scale flow, we have used a separate set of
model output using the same domain configuration and
two-nested domains (27 and 9 km), and this experiment
was run for 96 h (0000 UTC 2 August to 0000 UTC
6 August).
The model simulations were conducted using the
following physics options and tools:
d
cloud microphysics, using Milbrandt two-moment,
seven-class microphysics, with separate graupel and
hail species (Milbrandt and Yau 2005);
FEBRUARY 2014
d
d
d
d
d
d
KUMAR ET AL.
215
cumulus parameterization, using Kain–Fritsch cumulus
scheme for 27- and 9-km domains (Kain and Fritsch
1993; Kain 2004);
boundary layer, using the Yonsei University planetary
boundary layer (PBL) scheme of Hong and Pan (1996);
land surface, using the Noah land surface model (Chen
and Dudhia 2001);
radiation, using the Rapid Radiative Transfer Model
(RRTM; Mlawer et al. 1997) for longwave radiation
calculations and basing shortwave radiation calculations on Dudhia (1989);
surface layer, using the Monin–Obukhov scheme; and
radar simulator, using the National Center for Atmospheric Research (NCAR) Read/Interpolate/Plot, version 4 (RIP4; http://www.mmm.ucar.edu/wrf/users/docs/
ripug.htm) program.
3. Meteorological setting
A meteorological observatory located in the valley
near Leh reported that a total rain amount of 12.8 mm
fell during the early morning hours of 5 August 2010
(http://blogs.wsj.com/indiarealtime/2010/08/06/lehs-flashfloods-how-much-did-it-rain). This amount is abnormally large for this desert valley, where 15.4 mm is the
monthly average rain accumulation for August. An additional important factor is that this rainfall was not
simply local to the town, but rather, fell over the surrounding area on the valley’s bare mountainsides (see
photo on the title page of RH), where it ran off into
the town and flooded the Indus River. Figure 4a shows
the soil moisture gradient during 2–5 August in the LIS
simulation. The model indicates that a significant amount
of soil moisture was present prior to the landslide and
flash flood. Preexisting soil moisture saturation is one of
the key ingredients that, when combined with heavy rain,
can lead to landslides (Kirschbaum et al. 2012). Multiple
heavy rain episodes occurred during 3–5 August near
Leh, likely exacerbating the already wet soil conditions
and making the land surface especially conducive to
landslides when the flash flood occurred (Fig. 5). As
mentioned in RH, the Ladakh region is known for having
low-nutrient and poor soil conditions that are unsuitable
for agricultural purposes (Goodall 2004), which could
have contributed to the severity of the Leh flood and
related landslides. The results presented in RH and this
study based on satellite and model data show approximately 90–100-mm accumulation from 2 to 6 August. On
a longer time scale, the slow seasonal buildup of surface
water vapor mixing ratio in this normally arid region likely
preconditioned the area for a disaster when the period of
heavy rains occurred (Fig. 6).
FIG. 4. (a) Soil moisture and (b) precipitation (mm) from 2 to
5 Aug 2010. The soil moisture and precipitation is obtained from
the uncoupled LIS simulation.
Since the observed synoptic and satellite settings of
the Leh flood are described in RH, we summarize here
only a few key points. During the three days leading
up to the Leh flood, the 500-hPa winds were generally
easterly to northeasterly over Leh, coming across the
Himalayas after flowing around the south side of the
high-pressure system centered over the Tibetan Plateau
(Fig. 7). This anomalous wind formed a jet that helped to
organize MCSs and steer them over Leh. For three
successive days, well-organized MCSs traveled in the
downstream direction of the Indus River and passed
directly over Leh, which likely contributed to increasing
soil moisture in the entire region in addition to raising
the levels of the Indus River.2 At the same time, relatively stationary low-pressure systems over north-central
India were situated so that moist lower-level flow was
directed into the Leh region from the south-southeast.
The vortex over northwest India (seen most strongly at
700 hPa; see RH) was a midlevel monsoon cyclone of the
type described by Krishnamurti and Hawkins (1970),
2
See the supplementary material for a Meteosat-7 loop showing
the three successive MCSs passing over Leh, India.
216
JOURNAL OF HYDROMETEOROLOGY
VOLUME 15
FIG. 5. The LIS resolved (a),(c),(e),(g) precipitation and (b),(d),(f),(h) soil moisture rate per day is shown from 2 to
5 Aug 2010. All panels are derived from the uncoupled LIS simulation.
FEBRUARY 2014
KUMAR ET AL.
FIG. 6. The surface water vapor mixing ratio over Leh from
1 May to 30 Aug 2010. The surface moisture time series is constructed from NCEP reanalysis data.
Mak (1975), Carr (1977), Goswami et al. (1980), and Das
et al. (2007). The lower row of panels of Fig. 7 show the
progression of this low-pressure system weakening over
the three-day period leading up to the Leh flood. However, it seems that the simulation conducted from 2 to
5 August 2010 did not adequately capture the strength,
size, and location of the midlevel monsoon cyclone
217
compared to the reanalysis data shown in RH. The midlevel monsoon cyclone is known as a stationary pattern as
it typically grows and decays in place (Krishnamurti and
Hawkins 1970; Mak 1975; Carr 1977; Goswami et al.
1980). However, during the numerical simulation, the
cyclone propagates to the west and is located over
Pakistan on 5 August 2010 (Fig. 7f), which likely altered
the moisture flux into the lowlands of India and Pakistan
by this feature. Additionally, the Bay of Bengal depression is deeper than the midlevel monsoon cyclone
(Figs. 7d–f), which likely affected the dominant source
of the lowland moisture in the simulation compared to
the observations. Regardless of the shortcomings in the
numerical simulation, the midlevel low from the Arabian
Sea was instrumental in persistently driving low- to
midlevel moisture toward the Leh region from the south.
A Bay of Bengal depression was strengthening during
this period, and flow around its north side paralleled
the Himalayas and imported low-level moisture into the
Leh region from the southeast. As mentioned above, the
relative strength of the depression in the simulation is
greater than the reanalysis data, but it is nonetheless an
FIG. 7. (a)–(c) Model simulated geopotential height (m) contours and geostrophic wind vectors (m s21) at 500 hPa and (d)–(f) geopotential height contours at 700 hPa and terrain-following surface winds for 3–5 Aug 2010. Leh is indicated on each panel with a red circle.
The model analysis presented is from the 9-km domain from the coarser resolution (27 and 9 km) nested simulation conducted from 2 to
5 Aug 2010.
218
JOURNAL OF HYDROMETEOROLOGY
VOLUME 15
FIG. 8. Conceptual model demonstrating key meteorological elements that led to the
anomalous flash flooding case in Leh. Convective cells on the Tibetan Plateau organize upscale
and propagate to the west. The MCS on the edge of the Himalayas taps into the upslope lowlevel moisture (from Rasmussen and Houze 2012).
important source of moisture. RH hypothesized that the
moisture coming from the south and southeast, below
the level of the Himalayan Plateau, rose up the steeply
sloping terrain and strengthened convective systems
past the edge of the Plateau. Although convection was
able to form over the diurnally heated plateau, the
moisture over the high terrain was in too short of a supply
to maintain large mesoscale systems. The newly formed
and growing mesoscale convective systems, however,
could rapidly strengthen upon tapping the lower-level
moisture, and mesoscale systems could then deposit
a large amount of precipitation over the Leh valley. The
conceptual model of RH reproduced in Fig. 8 illustrates
this hypothesized sequence of events.
valley region as indicated by the time series of variables
along the model trajectories (Fig. 10b). The altitude and
wind speed of the parcels along both trajectories are also
shown in Figs. 10c and 10d. As suggested by RH, both
the Bay of Bengal and the Arabian Sea were important
sources of moisture that were necessary for the MCSs
propagating along the Tibetan Plateau. Additionally,
Medina et al. (2010) showed that surface fluxes of sensible
heat can increase buoyancy and previous precipitation
features can contribute to moisture fluxes from the
Arabian Sea. Thus, in addition to the oceanic moisture
sources, local moisture sources and surface fluxes likely
4. Model trajectory analysis
Figure 9 shows trajectories from the WRF model
simulation of the storm producing the Leh flood on
5 August 2010. Two families of trajectories occurred.
One family, arriving from the south-southwest, was associated with the flow around the monsoon vortex over
northwestern India. The other family, arriving from the
southeast, was associated with flow around the north
side of the Bay of Bengal depression. These trajectories
in the WRF simulation conform to the conceptual model
in Fig. 8. Trajectory 1 (marked on Fig. 9) originates over
the Bay of Bengal and brings plentiful moisture into the
lowlands of India and Pakistan. Trajectory 2 begins over
the Arabian Sea, picks up moisture from the Arabian
Sea, and arrives near the India–Pakistan border, helping
to increase the low-level moisture near the foothills of
the Himalayas. Trajectory 1 also shows that the moist air
present in the vicinity of Leh was brought up the steep
slope of the Himalayas, importing moisture into the Leh
FIG. 9. Model simulation backward trajectories of parcels from
the final state to initial state to track the source of the moisture. The
final state of all parcels is 1800 UTC 5 Aug and the initial state of
the parcel is 0000 UTC 2 Aug 2010. The model analysis presented is
from the 9-km domain from the coarser resolution (27 and 9 km)
nested simulation conducted from 2 to 5 Aug 2010.
FEBRUARY 2014
KUMAR ET AL.
219
slope of the Himalayas. The trajectories of warm, moist
air flowing into the Leh region contributed to building
an unstable stratification. Model-derived soundings over
Leh are shown in Figs. 11a and 11b at 0000 and 1200 UTC
5 August 2010. The 0000 UTC sounding shows that
the air was mostly humid up to 400 hPa, the precipitable
water was about 2 cm, and the CAPE was 655 J kg21.
After 12 h on the same day (1200 UTC) at the same place,
the sounding showed that the CAPE had increased to
3819 J kg21, with precipitable water of about 3 cm.
5. Mesoscale storm organization
FIG. 10. Values of variables along trajectories 1 and 2 in Fig. 9:
(a) temperature (K), (b) water vapor mixing ratio (g kg21), (c) altitude (m), and (d) wind speed (m s21) along trajectories 1 (originating in Bay of Bengal) and 2 (originating over the Arabian Sea).
The model analysis presented is from the 9-km domain from the
coarser resolution (27 and 9 km) nested simulation conducted from
2 to 5 Aug 2010.
contributed to the enhanced moisture content in trajectory 2 while over the Indian continent.
Trajectory analysis demonstrated that moist air in the
lowlands of India and Pakistan traveled up the steep
The initiation of convection over the Himalayan region is not well documented with surface-based observations. However, satellite data have established certain
climatological relationships between cloud patterns,
landforms, and orography in the Himalayan region
(Houze et al. 2007; Barros et al. 2004; Romatschke et al.
2010; Romatschke and Houze 2011; Hirose and Nakamura
2002). Medina et al. (2010) used the WRF model to understand the synoptic events, mesoscale dynamics, and
surface fluxes leading to the extreme forms of convection
occurring in this region. In view of these previous studies,
the heavy rain over Leh studied here was unusual; that is,
it did not conform to the climatology and convective
forms seen in these previous studies. The anomalous
nature of the Leh event was twofold. First, the previous
studies using satellite data show that convection over the
Tibetan Plateau hardly ever forms mesoscale systems; the
high-altitude arid conditions support primarily isolated
FIG. 11. Model-based soundings over Leh city at (a) 0000 and (b) 1200 UTC of 5 Aug. The modeled sounding is
constructed using the 1-km inner domain output.
220
JOURNAL OF HYDROMETEOROLOGY
convective-scale showers, which may be locally intense
but practically never grow upscale to form MCSs. Second, squall line systems with trailing stratiform regions
are very rare in this region, probably because of the
absence of a midlevel jet, which helps to organize convection into a squall line (Smull and Houze 1987). Houze
et al. (2007) described one rare exception when a 500-hPa
westerly jet was present during the monsoon and a squall
line propagated along the Himalayan barrier.
In the Leh flood case, both of these general rules were
broken. Convection grew upscale to form a mesoscale
system over the Plateau, and the 500-hPa flow formed
a jet that organized a propagating squall line system that
moved over Leh. Analysis of the 30-min resolution
Meteosat-7 infrared satellite data indicated that the
MCS was in the vicinity of Leh for approximately 7 h
(1700 UTC 5 August to 0000 UTC 6 August; shown in
the supplementary materials). Figure 12 shows the simulated reflectivity development over a 6 h period leading
up to the Leh flood. Convection over the Plateau was
cellular and merged into a larger combined system that is
consistent with the geosynchronous infrared satellite data
presented in RH. By 2000 UTC 5 August, at about the
time of the Leh flood, some of the convection had devolved into a stratiform region with a bright band while
newer reflectivity cells were deep and intense (Fig. 13).
The portion of the system shown in this figure was over
the high terrain of the Tibetan Plateau, but other portions of the system were moving over the Leh Valley at
this time. Figure 13 shows that the basic mesoscale structure of a squall line system had already taken shape over
the Plateau. This structure was likely tapping lower-level
moisture on the upslope side of the Himalayan barrier
(Fig. 12), allowing the convection to become even more
intense and capable of producing heavy rain. It should be
mentioned that Barros and Lang (2003) described similar
types of storms over Nepal, and their findings are consistent with the model obtained reflectivity.
A system of this type moving over a valley like that of
Leh is a prescription for flash flooding, especially after
the model-derived soil had likely become saturated as
described in section 3. The Leh flood was unlike the
slow-rising Pakistan flood of the lower Indus basin
a week earlier (Houze et al. 2011). That flood occurred
when the broad stratiform regions of mesoscale systems
became locked into place over the Himalayan slopes
surrounding the Indus River well downstream of Leh.
The widespread and steady nature of the broad stratiform precipitation regions fed by moist airflow from the
Bay of Bengal built up the rainwater runoff upstream,
which slowly traveled downriver, where it affected a
broad region of Pakistan. In contrast, the storm studied
here contained both intense convective cells and a trailing
VOLUME 15
stratiform region that was transitory over Leh. As will be
shown below, the rain at Leh fell over a short period.
However, the intensity of the downpour and the steepness of the slopes led to a sudden flood, somewhat similar
to those that occur occasionally in the Front Range of
the Rocky Mountains, such as the Big Thompson flood of
1976 (Caracena et al. 1979; Maddox et al. 1978). Although both the Leh and Big Thompson storms drew
upon the upslope flow of moist, unstable, tropical air
masses, they differed in that the Big Thompson storm was
a quasi-stationary convective system, while the Leh flood
was caused by a traveling squall line system, which had
formed over the higher terrain but became energized
when it moved over the escarpment and tapped the
moisture at lower levels (Fig. 9). Another factor contributing to the Leh flood was undoubtedly that on two
previous days, similar but smaller convective systems
passed over Leh (see RH for the full description and the
Meteosat-7 loop in the supplementary materials). The
model results suggest that these prior storms likely
contributed to the increasing saturation of the soil (Fig. 5),
which exacerbated the rapid runoff of the heavy precipitation from the 5 August MCS.
6. Internal storm structure
Figure 14 shows horizontal and vertical cross sections
through the storm when it was near the edge of the Plateau and passing over the Leh valley region. The storm
was propagating as a squall line with leading convective
and trailing stratiform precipitation, steered by the 500-hPa
wind (Fig. 7c). At the time of the cross section, the leading
convective line had reflectivity values exceeding 40 dBZ
reaching up to 12-km altitude, with the echo top exceeding 16 km (Fig. 14c). Convective cells with these
characteristics of maximum reflectivity and echo top
height are among the most extreme seen in the South
Asian monsoon region (Houze et al. 2007; Romatschke
et al. 2010). The air motions in the modeled storm show
the strong convective updraft over the Leh valley and a
strong midlevel inflow into the stratiform region, which
then sinks downslope toward the back edge of the leading
convective line. The midlevel jet in the stratiform region
is an extension of the environmental midlevel flow (about
500 hPa). It is well known that the midlevel rear-to-front
flow in the stratiform region often is a continuation and
an enhancement of the environmental midlevel flow
(Smull and Houze 1987; Kingsmill and Houze 1999;
Houze 2004). The air motions of this storm thus exhibit
the typical properties of a squall line MCS with trailing
stratiform precipitation and robust rear inflow. The
1-km resolution model simulation accumulated rainfall
between 1500 and 2300 UTC on 5 August 2010 is shown
FEBRUARY 2014
KUMAR ET AL.
FIG. 12. Sequence of model reflectivity showing the initiation and evolution of convection on 5 Aug 2010. The red
lines show the location of the vertical cross sections (see right), taken from west to east at (a),(b) 1400, (c),(d) 1600,
(e),(f) 1800, and (g),(h) 2000 UTC. The wind barbs are surface winds that follow the terrain. The analysis presented is
based on the 3-km resolution model output.
221
222
JOURNAL OF HYDROMETEOROLOGY
VOLUME 15
2011), Yuan et al. (2011), and Powell et al. (2012). The
upper portions of the anvils, both leading and trailing,
are dominated by cloud ice. The only precipitating anvil
is in the trailing region, immediately following the deep
convective region. At the time of the cross section, the
stratiform rain and part of the convective rain is falling
on the slopes of the mountainside above Leh, where it
would quickly run off and contribute to the flooding at
the base of the valley.
7. Precipitation and land surface moisture analysis
Spaceborne precipitation estimates from passive microwave sensors and the TRMM satellite have revolutionized the study of global precipitation and its distribution
(Kucera et al. 2013). Despite known uncertainties in
the satellite-derived precipitation estimates (e.g., Adler
et al. 2012; Kidd and Levizzani 2011; Gao and Liu 2013;
Rasmussen et al. 2013), the ability to examine precipitation in remote regions of the world has enabled an
unprecedented view of the climatological occurrence,
seasonal and diurnal variability, and individual extreme
precipitation events that was not possible before this era.
A recent study comparing precipitation gauge measurements over the complex terrain of the Tibetan Plateau to four merged precipitation datasets showed that
the TMPA product had the lowest overall errors (Gao
and Liu 2013). However, satellite datasets tended to
underestimate the higher precipitation rates on average
(Gao and Liu 2013; Rasmussen et al. 2013), so rain rates
derived from merged satellite data for heavy precipFIG. 13. (a) Horizontal reflectivity of the model storm at
2000 UTC 5 Aug 2010. The red line shows the location of the (b) itation events are likely a lower bound on the actual
vertical cross section, which was taken from west to east. The precipitation rates. An advantage of the satellite methods
analysis presented is based on the 3-km resolution model output. for studying flood cases is that they provide area-wide
estimates so that the rain over a whole hydrologic basin
can be seen. The areal rainfall is more useful than a point
in Fig. 14b, and it seems that the model slightly missed measurement, since the latter may not be representative
the heaviest precipitation over Leh with upward of of the region draining into the basin that floods. Re50 mm of rain accumulations just south of Leh.
gardless of the inherent and well-studied errors associThe simulated microphysical structure of the MCS ated with satellite data, it is the best source of information
producing the Leh flood is shown in Fig. 14d. For ref- for heavy precipitation events in remote regions with
erence, note that the 08C level is found at altitudes of limited ground-based rain-measuring instruments and
6–6.5 km above sea level. In the convective cell over the networks, and it provides both a spatial and temporal
Leh region, graupel dominates the active region above perspective that is crucial for studying events like the
the 08C level, while high concentrations of rainwater Leh flash flood.
dominate at lower levels. As suggested by Houze et al.
The rainfall evolution for the Leh flood is well cap(2007) for intense convective cells in this region, the tured by the TRMM 3B42 product (which includes
graupel extends to very high levels. In this simulation, information from TRMM and other satellites). Gridded
the 1.0 g kg21 graupel contour reaches 16-km height in rainfall estimates are on a 3-h temporal resolution and
the convective region, whereas in the stratiform region, a 0.258 3 0.258 spatial resolution in a global belt extending
little graupel occurs. Snow mixing ratio contours domi- from 508S to 508N. At 1200 UTC 5 August, rainfall patnate the lower portions of the anvils, consistent with the terns (Fig. 15a) are scattered in small patches over the
cloud radar data analyzed by Cetrone and Houze (2009, Himalayan belt, with rain accumulations of ;5–15 mm.
FEBRUARY 2014
KUMAR ET AL.
223
FIG. 14. Model storm internal structure shown by the 1-km horizontal domain zoomed in over Leh at 1815 UTC on
5 Aug 2010: (a) horizontal map of maximum reflectivity and horizontal wind at the 850-mb level, (b) model simulated
accumulated rainfall (1500–2300 UTC) on 5 Aug 2010, (c) vertical cross section along the red line of reflectivity and
air motion vectors in the plane of the cross section, and (d) vertical cross section of hydrometeor mixing ratios. In (c)
cloud water mixing ratios are color-filled contours according to the color bar on the right, and contours show mixing
ratios of rainwater in red, graupel in blue, snow in purple, and cloud ice in green. The red line in (a) shows the location
of the cross sections and the black dot indicates the location of Leh.
At 1500 UTC, rain is more concentrated and located
over the Tibetan Plateau in an elongated region paralleling the Himalayas, with rain accumulations up to
;50 mm in the northwestern region close to Leh. By
1800 UTC, heavy rain is concentrated near Leh in a zone
of 10–20 km in diameter with maximum accumulation
of about 75 mm (Fig. 15c). To facilitate a reasonable
comparison between the model and observational satellite precipitation data, the model data were resampled
to the same resolution as the TRMM 3B42 data (0.258 or
;26 km), and the rain accumulation over Leh is presented in Fig. 16. Comparisons to the TRMM 3B42 data
from around the same time (Fig. 15c) indicate that the
model simulation produced a net accumulation of ;39 mm
(Fig. 16), which gives greater confidence in the model
simulated precipitation. A nearby Indian Air Force observatory recorded ;13 mm of precipitation (http://www.
imd.gov.in/doc/cloud-burst-over-leh.pdf). These precipitation amounts do not, at first glance, seem excessive. So why
did the flash flood and mudslide occur? First of all, station measurements of precipitation are not directly
comparable to larger, gridded satellite estimates because of the difference in temporal and spatial scales.
However, aside from this consideration, several factors
must have combined. The storm maximized in intensity
as it moved off the Plateau and tapped the moisture
from the lowlands. Nearly all the rain fell over steeply
sloping land. Unused to rain, the land was barren. Because
of the buildup of soil moisture over the season (Fig. 6) and
especially in the preceding several days (Fig. 5), the soil
224
JOURNAL OF HYDROMETEOROLOGY
VOLUME 15
FIG. 16. Model simulated accumulated rainfall (mm) over Leh
on 5 Aug 2010. The model resolved precipitation is based on the
1-km resolution inner domain, but was resampled to 26-km resolution to match the TRMM 3B42 pixel size.
could have saturated. These factors, in addition to heavy
precipitation from the final MCS, came together to produce the flood at Leh at about 1900–2000 UTC (0100–
0200 local time).
Based on available observations over complex terrain,
the atmospheric and hydrological processes must be
accurately simulated (Warner 2011) in order to identify
and develop a physical understanding of the Leh flash
flood. We have tried to determine how well the model
reproduced the actual rainfall leading to the Leh flood
by comparing with standard sources of observational
rainfall information. Total accumulations from the 24 h
period that encompasses the Leh flood are presented in
Fig. 17, which includes the TRMM 3B42 product as well
as the Climate Prediction Center’s morphing technique
(CMORPH) product that was bias corrected with precipitation gauge data (Joyce et al. 2004). The CMORPH
and TRMM products show many of the same features
but also have some inconsistencies, both qualitatively
and quantitatively. The CMORPH product (Fig. 17a)
shows overall less rain than the TRMM estimated rainfall
(Fig. 17b), which likely results from the lack of rain gauge
observations in the region. Despite the various inconsistencies, the overall rainfall patterns from TRMM
and CMORPH are generally similar over Nepal, Tibet,
and northwest India. The latter shows values of 40–50 mm
of rainfall over the India–Pakistan border on 5 August,
whereas TRMM estimated amounts similar to those observed over Leh and is consistent with the simulated rain
accumulation from Fig. 16. The model-derived rainfall
shows a basic qualitative agreement with both observational products, but is in somewhat closer agreement with
the TRMM estimated rainfall given the spatial distriFIG. 15. TRMM 3-hourly accumulated precipitation (3B42) at bution and accumulated rain amounts. The model also
(a) 1200, (b) 1500, (c) 1800, and (d) 2100 UTC on 5 Aug 2010. Leh is
captured the localized rainfall spots over Nepal and
indicated on each panel with a black circle.
adjoining areas seen in both the TRMM and CMORPH
products.
The atmospheric processes associated with flash
floods that cause landslides over complex terrain have
FEBRUARY 2014
KUMAR ET AL.
225
FIG. 18. Hydrological parameters from May to August 2010 over
Leh showing (a) top soil layer (at 5-cm soil depth) soil moisture
(m3 m23), (b) hourly rain rate (mm h21), and (c) surface runoff
(mm h21). All panels are derived from the uncoupled LIS simulation.
FIG. 17. Total precipitation occurring between 0600 UTC 5 Aug
to 0600 UTC 6 Aug 2010 from (a) the bias-corrected CMORPH
merged with gauge analysis on a 0.258 latitude–longitude grid, (b)
TRMM product 3B42, and (c) the model-simulated precipitation
from the 3-km domain. Leh is indicated in (b) and (c) with a black
circle.
been shown to have anomalous hydrometeorological
characteristics, such as soil moisture, surface runoff,
rainfall amount, and surface storage water (Kirschbaum
et al. 2012). Land surface characteristics were not directly
observed during the Leh flood event but have been hypothesized to play a role in the severity of the flood and
media coverage–reported landslides in the region. The
uncoupled LIS system provides improved estimates of
land surface conditions such as soil moisture, evaporation, snowpack, and runoff, which were used as input
conditions to the coupled LIS-WRF runs. In addition,
this approach allows for further understanding of the
hydrological aspects of features at much longer time
scales, such as 3–6 months prior to the flash flood event.
Using the uncoupled LIS system to investigate the Leh
flash flooding event, analysis from May to September
2010 (Fig. 18) indicates that the soil moisture in the top
layer was dry before July, but it increased in steps after
multiple rain events. A sharp increase in soil moisture
with a peak of approximately 0.42 m3 m23 volumetric
soil moisture occurred on 5 August 2010 during the Leh
flash flood (Figs. 18a,b) and then gradually decreased
through early September. Compared to the climatological volumetric soil moisture value of approximately
0.1 m3 m23 over Leh derived from 62 years (1948–2010)
of GLDAS, version 2 (GLDAS-2), 18-resolution data
from the Noah model (not shown; Rodell et al. 2004),
the enhanced moisture preconditioning likely played
a large role in the land surface impact on the flash flood
226
JOURNAL OF HYDROMETEOROLOGY
event. Anomalous surface runoff is also evident during
the highly unusual flash flood over Leh, shown in Fig. 18c.
High values of surface runoff are critical and particularly
hazardous during such a short period of time because of
the potential to trigger landslides. While the current LIS
system lacks the potential influences and feedbacks between vegetation and geomorphic processes, such as
landslides, debris flows, and channel evolution, using an
integrated approach can lead to greater insight into
hazardous conditions and natural disasters in remote
regions, such as the Leh flash flood. As demonstrated in
this study, both the land surface processes and the meteorological conditions contributed to the disaster in
Leh, illuminating the potential benefit in implementing
integrated land–atmosphere forecast models in vulnerable regions of the world (Kirschbaum et al. 2012).
8. Conclusions
The Leh flood was not only a tragedy worth understanding in its own right, but it is also representative of
flash floods of the type that are prone to occur at the
steep edge of a high plateau. The Himalayas are the
mountainous edge of the Tibetan Plateau, somewhat
like how the Front Range of the Rocky Mountains constitutes the steep boundary of the vast plateau of the
western United States. The Big Thompson flood of 1976
is an example of such a flash flood over the Front Range.
Like that flood, the Leh flood and its associated landslide killed several hundred people and did terrible
damage. Both the Big Thompson and the Leh flood involved upslope flow of moist unstable tropical air masses;
however, they were the result of very different types of
convective storms. The Big Thompson storm was stationary over the Front Range, while the Leh Flood was
due to a moving squall line with trailing stratiform rain
forming over the Tibetan Plateau. It drew upon the warm
moist air rising up the steep Himalayan barrier as it
moved over the steep edge of the Plateau and was exacerbated by prior storms moistening the soil on the sides of
the mountains surrounding Leh. As presented in Doswell
et al. (1996), identifying similar ingredients that are
present during a variety of flash flood–producing storms
provides lessons for understanding and predicting flash
floods and leads to insights into flash flood–producing
scenarios in various regions of the world. The analysis in
this study has investigated the physical mechanisms,
storm structure, land surface characteristics, and precipitation of the MCS to provide a detailed analysis of
the ingredients that combined to produce the flash flood
in Leh.
RH examined synoptic and satellite data for the Leh
case and constructed the conceptual model in Fig. 8 to
VOLUME 15
explain the atmospheric aspects of the event. A key
feature of the conceptual model was its suggestion that
when the propagating MCS that formed over the Tibetan Plateau moved over Leh it was able to feed upon
warm, moist air from both the Arabian Sea and Bay of
Bengal. That moist air was thought to be flowing up the
Himalayan slope in the Leh region as a consequence
of low-pressure systems over the northwestern subcontinent and the Bay of Bengal. Trajectories in the model
simulation presented here confirm this suggestion. The
simulated storm approximately reproduced the rainfall
pattern of the storm and total storm accumulations are
consistent between the model and the TRMM satellite
estimates (Figs. 15, 16). The model simulation further
shows that the Leh storm’s mesoscale organization, internal air motions, and hydrometeor distributions were
consistent with those of a typical traveling squall line
with trailing stratiform precipitation. The convective
region maximized over the Leh valley, and the stratiform
region contributed rain there as well, as the system moved
through the Leh region. The mesoscale proportions of the
systems assured that the surrounding mountainsides, as
well as Leh itself, received the heavy but short-lived
downpour.
The hydrological aspects of the Leh flooding event
were investigated with the coupled WRF-LIS modeling
tool. In the vicinity of Leh, anomalously high values of
soil and surface moisture were present just prior to the
flash flooding event that likely preconditioned the soil
for maximum immediate runoff and destabilized the
ground, making it susceptible to landslides. The saturation of the soil in the model simulations was intensified
by heavy rainfall from the succession of two MCSs
propagating over the city of Leh on the two days before
the flood. On 5 August 2010, when the third and most
intense MCS propagated over Leh, all of the moisture
parameters in the simulation were maximized both in
the atmosphere and on the ground, indicating the hazardous nature of this flash flood event. Landslides and
flash flooding occurred and were reported by local media. These events were indicated in the WRF-LIS simulation as high values of surface runoff during the event,
as the barren land surface in this region could not absorb
all of the water deposited in the short time of the 5 August
storm. A coupled model like the WRF-LIS system or
a similar coupled model is capable of providing hydrologic information and potentially dangerous scenarios
that could be very useful in high landslide- and floodprone regions.
Flash floods on the edges of major plateau regions are
infrequent and devilishly hard to predict. Besides a model
such as the WRF-LIS system, basic climatological
awareness can be useful in forecasting events such as the
FEBRUARY 2014
KUMAR ET AL.
Leh flood, specifically awareness of what is usual in order to be able to recognize an outlier event. From recent
satellite climatologies (Houze et al. 2007; Romatschke
et al. 2010), we know that the Leh storm was highly
unusual in two respects: 1) convection over the Tibetan
Plateau seldom grows upscale into a mesoscale convective system and 2) traveling squall line systems are
rare in the Himalayan region. RH demonstrated how the
large-scale flow pattern and ordinary geosynchronous
satellite imagery indicated that this highly unusual condition was occurring for several days. This fact suggests
that, along with coupled atmospheric–land surface models,
better use of real-time observations against the background of recent satellite climatologies of the region
might also contribute to recognizing the warning signs
for a flash flood like the Leh event.
Acknowledgments. This research was sponsored by
NSF Grants ATM-0820586 and AGS-1144105 and NASA
Grants NNX10AH70G and NNX11AL65H (Dr. Ramesh
Kakar). This research was also supported by National
Aeronautics and Space Administration grants from the
NASA Modeling, Analysis and Prediction (MAP) Program (Dr. David Considine, MAP program manager).
The NASA Center for Climate Simulation (NCCS) computing system provided resources for the model simulations. Thara Prabhakaran participated in an early phase of
this study. Graphics art and manuscript editing were
provided by Beth Tully. The authors thank Dr. Russ
Schumacher and two anonymous reviewers for their
comments and suggestions, which have greatly improved
this manuscript.
REFERENCES
Adler, R. F., G. Gu, and G. J. Huffman, 2012: Estimating climatological bias errors for the Global Precipitation Climatology
Project (GPCP). J. Appl. Meteor. Climatol, 51, 84–99,
doi:10.1175/JAMC-D-11-052.1.
Barros, A. P., and T. J. Lang, 2003: Monitoring the monsoon in the Himalayas: Observations in central Nepal,
June 2001. Mon. Wea. Rev., 131, 1408–1427, doi:10.1175/
1520-0493(2003)131,1408:MTMITH.2.0.CO;2.
——, G. Kim, E. Williams, and S. W. Nesbitt, 2004: Probing orographic controls in the Himalayas during the monsoon using
satellite imagery. Nat. Hazards Earth Syst. Sci., 4, 29–51,
doi:10.5194/nhess-4-29-2004.
Caracena, F., R. A. Maddox, L. R. Hoxit, and C. F. Chappell, 1979:
Mesoanalysis of the Big Thompson storm. Mon. Wea. Rev., 107, 1–
17, doi:10.1175/1520-0493(1979)107,0001:MOTBTS.2.0.CO;2.
Carr, F. H., 1977: Mid-tropospheric cyclones of the summer monsoon. Pure Appl. Geophys., 115, 1383–1412, doi:10.1007/
BF00874415.
Cetrone, J., and R. A. Houze Jr., 2009: Anvil clouds of tropical
mesoscale convective systems in monsoon regions. Quart.
J. Roy. Meteor. Soc., 135, 305–317, doi:10.1002/qj.389.
227
——, and ——, 2011: Leading and trailing anvil clouds of West
African squall lines. J. Atmos. Sci., 68, 1114–1123, doi:10.1175/
2011JAS3580.1.
Chen, F., and J. Dudhia, 2001: Coupling an advanced land
surface–hydrology model with the Penn State–NCAR
MM5 modeling system. Part I: Model implementation and
sensitivity. Mon. Wea. Rev., 129, 569–585, doi:10.1175/
1520-0493(2001)129,0569:CAALSH.2.0.CO;2.
Das, S., R. Ashrit, M. W. Moncrieff, M. D. Gupta, J. Dudhia,
C. Liu, and S. R. Kalsi, 2007: Simulation of intense organized
convective precipitation observed during the Arabian Sea
Monsoon Experiment (ARMEX). J. Geophys. Res., 112,
D20117, doi:10.1029/2006JD007627.
Doswell, C. A., III, H. E. Brooks, and R. A. Maddox, 1996:
Flash flood forecasting: An ingredients-based methodology. Wea. Forecasting, 11, 560–581, doi:10.1175/
1520-0434(1996)011,0560:FFFAIB.2.0.CO;2.
Dudhia, J., 1989: Numerical study of convection observed during
the winter monsoon experiment using a mesoscale twodimensional model. J. Atmos. Sci., 46, 3077–3107, doi:10.1175/
1520-0469(1989)046,3077:NSOCOD.2.0.CO;2.
Gao, Y. C., and M. F. Liu, 2013: Evaluation of high-resolution
satellite precipitation products using rain gauge observations
over the Tibetan Plateau. Hydrol. Earth Syst. Sci., 17, 837–849,
doi:10.5194/hess-17-837-2013.
Goodall, S. K., 2004: Rural-to-urban migration and urbanization in Leh, Ladakh: A case study of three nomadic pastoral communities. Mt. Res. Dev., 24, 220–227, doi:10.1659/
0276-4741(2004)024[0220:RMAUIL]2.0.CO;2.
Goswami, B. N., R. N. Keshavamurthy, and V. Satyan, 1980: Role
of barotropic, baroclinic, and combined barotropic-baroclinic
instability for the growth of monsoon depressions and midtropospheric cyclones. Proc. Indian Acad. Sci., Earth Planet.
Sci., 89, 79–97.
Hirose, M., and K. Nakamura, 2002: Spatial and seasonal variation of rain profiles over Asia observed by spaceborne
precipitation radar. J. Climate, 15, 3443–3458, doi:10.1175/
1520-0442(2002)015,3443:SASVOR.2.0.CO;2.
Hong, S.-Y., and H.-L. Pan, 1996: Nonlocal boundary layer
vertical diffusion in a medium-range forecast model.
Mon. Wea. Rev., 124, 2322–2339, doi:10.1175/
1520-0493(1996)124,2322:NBLVDI.2.0.CO;2.
Houze, R. A., Jr., 2004: Mesoscale convective systems. Rev. Geophys., 42, RG4003, doi:10.1029/2004RG000150.
——, D. C. Wilton, and B. F. Smull, 2007: Monsoon convection in the
Himalayan region as seen by the TRMM Precipitation Radar.
Quart. J. Roy. Meteor. Soc., 133, 1389–1411, doi:10.1002/qj.106.
——, K. L. Rasmussen, S. Medina, S. R. Brodzik, and U. Romatschke,
2011: Anomalous atmospheric events leading to the summer
2010 floods in Pakistan. Bull. Amer. Meteor. Soc., 92, 291–298,
doi:10.1175/2010BAMS3173.1.
Joyce, R. J., J. E. Janowiak, P. A. Arkin, and P. Xie, 2004:
CMORPH: A method that produces global precipitation estimates from passive microwave and infrared data at high
spatial and temporal resolution. J. Hydrometeor., 5, 487–503,
doi:10.1175/1525-7541(2004)005,0487:CAMTPG.2.0.CO;2.
Kain, J. S., 2004: The Kain–Fritsch convective parameterization: An update. J. Appl. Meteor., 43, 170–181, doi:10.1175/
1520-0450(2004)043,0170:TKCPAU.2.0.CO;2.
——, and J. M. Fritsch, 1993: Convective parameterization for
mesoscale models: The Kain–Fritsch scheme. The Representation of Cumulus Convection in Numerical Models, Meteor.
Monogr., No. 46. Amer. Meteor. Soc., 165–170.
228
JOURNAL OF HYDROMETEOROLOGY
Kidd, C., and V. Levizzani, 2011: Status of satellite precipitation
retrievals. Hydrol. Earth Syst. Sci., 15, 1109–1116, doi:10.5194/
hess-15-1109-2011.
Kingsmill, D. E., and R. A. Houze Jr., 1999: Thermodynamic
characteristics of air flowing into and out of precipitating
convection over the west Pacific warm pool. Quart. J. Roy.
Meteor. Soc., 125, 1209–1229, doi:10.1002/qj.1999.49712555606.
Kirschbaum, D. B., R. Adler, Y. Hong, S. Kumar, C. Peters-Lidard,
and A. Lerner-Lam, 2012: Advances in landslide hazard
nowcasting: Evaluation of a global and regional modeling
approach. Environ. Earth Sci., 66, 1683–1696, doi:10.1007/
s12665-011-0990-3.
Krishnamurti, T. N., and R. S. Hawkins, 1970: Mid-tropospheric
cyclones of the southwest monsoon. J. Appl. Meteor., 9, 442–
458, doi:10.1175/1520-0450(1970)009,0442:MTCOTS.2.0.CO;2.
Kucera, P. A., E. E. Ebert, F. J. Turk, V. Levizzani, D. Kirschbaum,
F. J. Tapiador, A. Loew, and M. Borsche, 2013: Precipitation
from space: Advancing Earth system science. Bull. Amer.
Meteor. Soc., 94, 365–375, doi:10.1175/BAMS-D-11-00171.1.
Kumar, S. V., and Coauthors, 2006: Land information system:
An interoperable framework for high resolution land surface
modeling. Environ. Model. Software, 21, 1402–1415, doi:10.1016/
j.envsoft.2005.07.004.
——, C. Peters-Lidard, J. L. Eastman, and W.-K. Tao, 2008: An
integrated high-resolution hydrometeorological modeling
testbed using LIS and WRF. Environ. Model. Software, 23,
169–181, doi:10.1016/j.envsoft.2007.05.012.
Maddox, R. A., L. R. Hoxit, C. F. Chappell, and F. Caracena, 1978:
Comparison of meteorological aspects of the Big Thompson
flood and Rapid City flash floods. Mon. Wea. Rev., 106, 375–389,
doi:10.1175/1520-0493(1978)106,0375:COMAOT.2.0.CO;2.
Mak, M.-K., 1975: The monsoonal mid-tropospheric cyclogenesis. J. Atmos. Sci., 32, 2246–2253, doi:10.1175/
1520-0469(1975)032,2246:TMMTC.2.0.CO;2.
Medina, S., R. A. Houze Jr., A. Kumar, and D. Niyogi, 2010:
Summer monsoon convection in the Himalayan region: Terrain
and land cover effects. Quart. J. Roy. Meteor. Soc., 136, 593–616,
doi:10.1002/qj.601.
Milbrandt, J. A., and M. K. Yau, 2005: A multimoment bulk microphysics parameterization. Part II: A proposed threemoment closure and scheme description. J. Atmos. Sci., 62,
3065–3081, doi:10.1175/JAS3535.1.
Mitchell, K. E., and Coauthors, 2004: The multi-institution North
American Land Data Assimilation System (NLDAS): Utilizing multiple GCIP products and partners in a continental
distributed hydrological modeling system. J. Geophys. Res.,
109, D07S90, doi:10.1029/2003JD003823.
Mlawer, E. J., S. J. Taubman, P. D. Brown, M. J. Iacono, and S. A.
Clough, 1997: Radiative transfer for inhomogeneous atmosphere: RRTM, a validated correlated-k model for the long
wave. J. Geophys. Res., 102 (D14), 16 663–16 682, doi:10.1029/
97JD00237.
VOLUME 15
Peters-Lidard, C., D. Mocko, M. Garcia, J. A. Santanello Jr., M. A.
Tischler, M. S. Moran, and Y. Wu, 2008: Role of precipitation
uncertainty in the estimation of hydrologic soil properties using
remotely sensed soil moisture in a semiarid environment. Water
Resour. Res., 44, W05S18, doi:10.1029/2007WR005884.
Powell, S. W., R. A. Houze Jr., A. Kumar, and S. A. McFarlane,
2012: Comparison of simulated and observed continental
tropical anvil clouds and their radiative heating profiles.
J. Atmos. Sci., 69, 2662–2681, doi:10.1175/JAS-D-11-0251.1.
Rasmussen, K. L., and R. A. Houze Jr., 2012: A flash flooding
storm at the steep edge of high terrain: Disaster in the
Himalayas. Bull. Amer. Meteor. Soc., 93, 1713–1724, doi:10.1175/
BAMS-D-11-00236.1.
——, S. L. Choi, M. D. Zuluaga, and R. A. Houze Jr., 2013: TRMM
precipitation bias in extreme storms in South America. Geophys. Res. Lett., 40, 3457–3461, doi:10.1002/grl.50651.
Rodell, M., and Coauthors, 2004: The Global Land Data Assimilation System. Bull. Amer. Meteor. Soc., 85, 381–394, doi:10.1175/
BAMS-85-3-381.
Romatschke, U., and R. A. Houze Jr., 2011: Characteristics of
precipitating convective systems in the South Asian monsoon.
J. Hydrometeor., 12, 3–26, doi:10.1175/2010JHM1289.1.
——, S. Medina, and R. A. Houze Jr., 2010: Regional, seasonal, and
diurnal variations of extreme convection in the South Asian
region. J. Climate, 23, 419–439, doi:10.1175/2009JCLI3140.1.
Santanello, J. A., C. D. Peters-Lidard, and S. V. Kumar, 2011:
Diagnosing the sensitivity of local land–atmosphere coupling
via the soil moisture–boundary layer interaction. J. Hydrometeor., 12, 766–786, doi:10.1175/JHM-D-10-05014.1.
Skamarock, W. C., and Coauthors, 2008: A description of the
Advanced Research WRF version 3. NCAR Tech. Note
NCAR/TN-4751STR, 113 pp. [Available online at http://www.
mmm.ucar.edu/wrf/users/docs/arw_v3.pdf.]
Smull, B. F., and R. A. Houze Jr., 1987: Rear inflow in squall lines with
trailing stratiform precipitation. Mon. Wea. Rev., 115, 2869–2889,
doi:10.1175/1520-0493(1987)115,2869:RIISLW.2.0.CO;2.
Tao, Z., J. A. Santanello, M. Chin, S. Zhou, Q. Tan, E. M. Kemp, and
C. D. Peters-Lidard, 2013: Effect of land cover on atmospheric
processes and air quality over the continental United States—A
NASA unified WRF (NU-WRF) model study. Atmos. Chem.
Phys. Discuss., 13, 5429–5475, doi:10.5194/acpd-13-5429-2013.
Warner, T. T., 2011: Quality assurance in atmospheric modeling. Bull. Amer. Meteor. Soc., 92, 1601–1610, doi:10.1175/
BAMS-D-11-00054.1.
Yuan, J., R. A. Houze Jr., and A. Heymsfield, 2011: Vertical
structures of anvil clouds of tropical mesoscale convective
systems observed by CloudSat. J. Atmos. Sci., 68, 1653–1674,
doi:10.1175/2011JAS3687.1.
Zaitchik, B. F., J. A. Santanello, S. V. Kumar, and C. D. PetersLidard, 2013: Representation of soil moisture feedbacks during
drought in NASA Unified WRF (NU-WRF). J. Hydrometeor.,
14, 360–367, doi:10.1175/JHM-D-12-069.1.

Similar documents

1 Simulation of a Flash Flooding Storm at the Steep Edge of the

1 Simulation of a Flash Flooding Storm at the Steep Edge of the Cloud microphysics: Milbrandt 2-moment, seven class microphysics, with separate graupel and hail species (Milbrandt and Yau 2005)

More information

A Flash-Flooding Storm at the Steep Edge of High Terrain: Disaster

A Flash-Flooding Storm at the Steep Edge of High Terrain: Disaster convective cells, which are minimal in horizontal extent. The occurrence of mesoscale convective systems (i.e. systems with complexes of convective cells and associated stratiform regions) over the...

More information