Monitoring of 7Be and gross beta in particulate matter of surface air

Chemosphere 152 (2016) 481e489
Contents lists available at ScienceDirect
Chemosphere
journal homepage: www.elsevier.com/locate/chemosphere
Monitoring of 7Be and gross beta in particulate matter of surface air
from Mallorca Island, Spain
s a, Esteve Gomila b, Jose
Manuel Estela b,
Melisa Rodas Ceballos a, Antoni Borra
b, Laura Ferrer a, *
Víctor Cerda
a
b
Environmental Radioactivity Laboratory (LaboRA), University of the Balearic Islands, 07122 Palma de Mallorca, Spain
Environmental Analytical Chemistry Laboratory (LQA2), University of the Balearic Islands, 07122 Palma de Mallorca, Spain
h i g h l i g h t s
g r a p h i c a l a b s t r a c t
7Be, Ab and 210Pb activities are evaluated for the first time in Mallorca
Island.
A preliminary study of 7Be and 210Pb
in PM10 filters is presented.
Temperature strongly affects the
behavior of 7Be and Ab in PM.
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 7 September 2015
Received in revised form
19 February 2016
Accepted 6 March 2016
Available online 19 March 2016
Measurements of airbone radionuclides 7Be and gross beta (Ab) jointly with atmospheric aerosols, i.e.
particulate matter (PM) were routinely carried out for 10 years (2004e2014) at the University of the
Balearic Islands, Spain. A total of 728 filter (0.8 mm pore size) samples were collected, and in all of them
7
Be and Ab specific activities were detected. The maximum and minimum specific activities monitored
were 0.73 ± 0.41 e 15.8 ± 1.26 mBq m3 of 7Be and 0.14 ± 0.02 e 2.55 ± 0.04 mBq m3 of Ab. PM
concentrations were also determined, showing seasonal behavior with the highest concentration in
summer and the lowest one in winter. Several meteorological parameters have been considered to
explain this intra-annual variation. Principal component analysis (PCA) was applied to the dataset
indicating that it is well represented by two principal components that explain 76.6% of total variance.
Additionally, a second study with preliminary results of the specific activities of two natural radiotracers
(7Be and 210Pb) in PM10 samples was carried out. They were monitored for two years (2013e2015) in air
of Mallorca Island. 7Be and 210Pb were detected in most of the PM10 filters, in 100% and 93% of them,
respectively. The relationship between activities of both radionuclides and several relevant meteorological parameters was established at 95% confidence level. As a common result to PM and PM10 samples,
a strong positive correlation between the evaluated radionuclides and temperature was found.
© 2016 Elsevier Ltd. All rights reserved.
Handling Editor: Martine Leermakers
Keywords:
7
Be
Ab
210
Pb
PM
PCA
Mallorca Island
1. Introduction
* Corresponding author.
E-mail address: [email protected] (L. Ferrer).
http://dx.doi.org/10.1016/j.chemosphere.2016.03.021
0045-6535/© 2016 Elsevier Ltd. All rights reserved.
In the atmosphere, the radioactivity is originated from cosmogenic production, natural radioactivity decay, nuclear weapon
testing and nuclear accidents. Long-term monitoring of
482
M. Rodas Ceballos et al. / Chemosphere 152 (2016) 481e489
radionuclides in the atmosphere provides useful information in the
study of atmospheric transport processes, as a tracer of aerosol
species and to compare the environmental impact of radioactivity
~ as et al., 2009). Even
from natural and man-made sources (Duen
though Mallorca Island has a scarce industrial activity, being the
tourism its main economy sector, the atmospheric pollution is
controlled (Cerro et al., 2015). Cosmogenic radionuclides have been
widely investigated as efficient tracers in environmental science, in
order to study several basic atmospheric processes (stratosphereto-troposphere exchange, interhemispheric transport) (Tositti et al.,
~ as et al., 2011).
2014; Duen
Measurement of gross beta activity (Ab) is generally made as a
means of “screening” for unusual levels of radioactivity; and it is
important because it allows to establish trends for aerosol residence times and radionuclide concentrations in the atmosphere
~ as et al., 1999, 2004; Herna
ndez et al., 2005). Moreover the
(Duen
major sources of radioactivity in the air are radon (222Rn) and their
products decay (García- Talavera et al., 2001). Among which, the
most important beta emitters are 210Pb (t1/2 ¼ 22.3 years) and 210Bi
~ ero- García, 2014).
(t1/2 ¼ 5 days) (García- Talavera et al., 2001; Pin
7
Be is a relatively short lived (t1/2 ¼ 53.3 days) radionuclide of
cosmogenic origin, produced by cosmic ray spallation reactions
with light atmospheric nuclei of nitrogen, oxygen and carbon. Most
of 7Be in the atmosphere is present in the stratosphere, because
most of its production occurs in this layer; while the remaining part
is produced in the troposphere and particularly in the upper
troposphere (Tositti et al., 2014).
Within environmental radioactivity programmes, Ab and 7Be
activities are measurement in order to establish their temporal and
spatial variations.
Particulate matter (PM) levels have been monitored during the
past decades, especially those with aerodynamic diameter smaller
or equal to 10 mm, because they have detrimental effects in human
health and air quality (Gerasopoulos et al., 2006; Bigi and
Ghermandi, 2014). A variety of natural (sea salt, soil dust suspension, etc.) and anthropogenic sources (industrial activities, biomass
burning, etc.) are considered to cause the introduction of aerosols
in the atmosphere (Gerasopoulos et al., 2006).
The movement and transport of radionuclides in the atmosphere is mainly carried out by air PM (Atwood, 2010). Newly
formed 7Be atoms rapidly associate themselves with submicrometer (<1 mm) aerosol particulates. 7Be enters the troposphere due to the vertical exchange of material from the stratosphere, reaches the earth's surface via dry and wet depositional
events (L'Annunziata, 2012). For this reason, 7Be has been used as a
tracer of stratospheric intrusions of gases and aerosols into the
ndez et al., 2008). The variation of 7Be activity
troposphere (Herna
in surface air is usually associated with factors as solar activity level,
latitude, altitude, seasonal air mass transport and meteorological
conditions (Chao et al., 2014). Therefore, both Ab and 7Be were
monitored in PM samples (filters of 0.8 mm pore size) collected in
the surface air of Mallorca Island during 10 years (2004e2014).
Thus, the main objective of this study is to establish the variation of Ab and 7Be specific activities and their correlations with PM
concentration and meteorological parameters. Additionally, a
multivariate technique (principal component analysis, PCA) was
performed for the Ab and 7Be specific activities, the PM concentration and several factors as temperature, precipitation and relative humidity, in order to explain the influence that these factors
have in the specific activity variation.
Moreover, a preliminary study of 7Be and 210Pb specific activities
variation in PM10 (particulate matter < 10 mm) on surface air of
Mallorca Island was carried out during two years (2013e2015).
210
Pb is formed in the lower to mid-troposphere where it attaches
itself to fine aerosols once it is produced from 222Rn. Besides, 210Pb
is found in the soil closer to the surface because of wet and dry
deposition processes. 210Pb is also used to analyze and trace
pollutant from human activity (Atwood, 2010). In this preliminary
study, the correlation between the activities of these natural
radioactive tracers was determined, and the relationships with
some meteorological parameters as temperature, wind speed and
relative humidity were established.
2. Material and methods
2.1. Sampling area description
Mallorca Island is located in the Mediterranean Sea, at a distance
of 180 km from the eastern cost of the Iberian Peninsula. Mallorca
with a fixed population around 870,000 inhabitants is the largest
island of the Balearic archipelago. It has Mediterranean climate
with mean annual temperature of 19 C and mean annual precipitation of 410 mm. The coldest month is January, with 15.1 /3.5 and
the warmest August, with 31.0 /18.2 . There are approximately 51
days of rain per year and 2756 h of sunshine. The months with the
higher precipitation are October to December (Spanish Morological
Agency (AEMET), 2016).
The predominant winds in Mallorca are ENE (East-North-East).
The origin of the light winds (breeze) is somewhat conditioned by
the alignment towards the outside of the Alfabia mountains to W
(west), the Levante mountains, on a smaller scale, to E (east). These
mountains with orientation NNE/SSW shape an interior space as a
large canal behaving as a corridor through the interior valleys, favoring a convergence zone for both breezes and winds from the
component ENE and SW (southwest). Around the 9 h in the
morning, the ENE reaches its fullness with 16.4%, while sea breezes
are assimilated into this island to the winds for the third quadrant
especially the SW, which makes an appearance at around 12 h
~ oz, 2002).
(Mun
The PM samples were collected at University of the Balearic
Islands campus (39 250 4700 N; 2 390 0100 E) from 2004 to 2014. The
PM10 samples were collected at three close locations of Mallorca
Island during two years (2013e2015). One sampling station was
located at Joan March hospital (39 400 4400 N; 2 410 1600 E), another
station was in Palmanyola (39 390 2800 N; 2 390 5700 E) and the third
one was in Es Garrovers (39 370 3800 N; 2 410 5300 E), all of them are
located at the north-east of the city of Palma de Mallorca.
Fig. 1 shows the sampling stations for PM and PM10 atmospheric
aerosol samples. The locations of PM and PM10 sampling stations
were chosen to fit the necessities of several environmental monitoring plans.
2.2. Surface air sampling
A low-volume atmospheric aerosol sampler was employed in
order to collect the PM samples (728 filters) for 7Be and Ab determination. The PM sample was obtained using nitrocellulose filters
(47 mm diameter and 0.8 mm pore size), mounted on a head of air
sampler (AVS-28 A, Radeco) with a flow rate of 1.8 m3 h1. Each PM
sample was collected during one week and the filter exchange
process did not exceed 15 min. Then, filters were kept for one week
in a desiccator until their measurement of Ab activity. The gamma
measurement was performed monthly, when all filters of one
month were simultaneously measured.
A high-volume sampler was used to collect PM10 samples for 7Be
and 210Pb determination. A sensor (CAV-A/mb, MCV) and a filter
head (PM1025, MCV) with nozzle plates of 10 mm, using GF/A filters
(150 mm diameter) was employed, which in normal conditions
works at a flow rate of 30 m3 h1. Each sampling lasted 1 week. The
sampling was made when the environmental monitoring plan
M. Rodas Ceballos et al. / Chemosphere 152 (2016) 481e489
483
N
W
BALEARIC SEA
E
S
4
1
2
3
MEDITERRANEAN SEA
Fig. 1. Location of sampling stations. The symbol ( ) represents the location of the PM10 monitoring stations: 1) Joan March Hospital, 2) Palmanyola and 3) Es Garrovers. The
symbol ( ) represents the location of the PM monitoring: 4) University of the Balearic Islands campus.
allowed to use the PM10 sampler. Gamma detection was carried out
one day after sampling. Both samplers were frequently calibrated in
order to ensure the required flow rate.
2.3. Radiometric detection
The measurement of 7Be and 210Pb specific activities was performed with a high-purity coaxial Germanium detector (Canberra
GR5023-7500 SL) iron shielded, using their gamma emissions at
477 and 46.5 keV, respectively. The spectra were acquired and
analyzed by GENIE 2000 software, including the correction decay
for each radionuclide. Calibrations in energy and efficiency of the
gamma detector were periodically carried out using a gamma
cocktail standard for 7Be quantification and a 210Pb standard, both
purchased from CIEMAT (Centre for Energy, Environment and
Technology, Spain). Calibrations (backgrounds and standards) were
performed for each used geometry. The counting time used for
gamma spectrometry was 86,400 s for all samples.
The measurement of Ab specific activity was performed with a
gas-flow low background proportional counter (Tennelec LB 4100)
with four planchettes. Ab was acquired at 1380 V using APEX
software for acquisition and data processing. Calibrations in efficiency, auto absorption factor and spillover of the low proportional
counter were made using a 90Sr standard purchased from CIEMAT.
The counting time was 60,000 s for all samples.
2.4. Meteorological variables
The meteorological information corresponding to the PM sampling station was obtained from the Spanish Meteorological Agency
(AEMET) station network. The meteorological variables used were:
temperature, precipitation, pressure, relative humidity and wind
speed. In the case of PM10 samples, the data were obtained from the
meteorological station associated to the PM10 sampler. The
available meteorological variables were: wind speed, temperature,
pressure and relative humidity.
Monthly sunspots number data was obtained from Solar Influence Data Analysis Center (SIDC)-solar activity (SILSO, 2016), which
is the solar physics research department of the Royal Observatory of
Belgium.
The selection of meteorological conditions for PM and PM10
~ as et al., 2009;
were made according to previous studies (Duen
ndez et al., 2008; Chao et al., 2014;
Tositti et al., 2014; Herna
Krajny et al., 2014; Gordo et al., 2015).
2.5. Statistical analysis
Although Ab specific activity and PM concentration were
determined weekly, the reported values in this study are their
monthly averages, so they can be compared with the monthly
values of the other considered variables.
The potential correlations between the following set of variables: 7Be and Ab specific activities, and PM concentrations (characterized by a log-normal probability distribution); temperature,
atmospheric pressure, wind speed and relative humidity (characterized by a normal probability distribution), as well as precipitation and number of sunspots (characterized by an exponential
probability distribution) were studied. The correlation between
these variables was determined with Pearson correlation coefficient at 95% of confidence level. For those variables not well characterized by a normal distribution, the logarithm was taken before
computing any correlation.
PCA is a multivariate technique that analyzes a data set in which
observations are described by a high number of quantitative intercorrelated dependent variables (Abdi and Williams, 2010). The
objective of PCA is to reduce the original number of variables by
obtaining and extracting a smaller amount of new variables
(principal components) which explain most of the variance of the
484
M. Rodas Ceballos et al. / Chemosphere 152 (2016) 481e489
dataset with little loss of information. PCs are extracted in
decreasing order of importance, so the first PC contains the greatest
amount of variance in the dataset and each successive PC describes
less. Each PC is characterized by its eigenvalue, which is proportional to the variance explained by this component. An eigenvalue
greater than 1 indicates that its corresponding PC represents more
variance than one of the original variables. For this reason, usually
only those PCs with eigenvalues above 1 are kept. PCA was applied
to the 10 years data set composed by the radionuclides specific
activities, PM concentrations and the set of meteorological variables that presents a higher correlation with: 7Be and Ab specific
activities. The statistical analysis was performed using R, the language for statistical computing (R Core Team, 2015).
Moreover in the first PM10 preliminary study, the correlations
between 7Be and 210Pb specific activities, and several meteorological variables (temperature, relative humidity and wind speed) was
determined with Pearson correlation coefficient at 95% of confidence level. Once again, to ensure normality of the dataset, the
logarithm of the radionuclides specific activities was used.
3. Results and discussion
3.1. Cyclic and seasonal variation of radionuclide specific activities
in PM
The 7Be and Ab specific activities were detected above the
minimum detectable activity (MDA) in all samples at the station of
the University of the Balearic Islands campus, being the MDAs for
Ab 0.025 mBq m3, and 0.72 mBq m3 for 7Be. The maximum and
minimum specific activities monitored during these 10 years
(2004e2014) were 0.14 ± 0.02 e 2.55 ± 0.04 mBq m3 for Ab and
0.73 ± 0.41 e 15.8 ± 1.26 mBq m3 for 7Be.
The temporal distribution for PM concentration has a similar
behavior to that of the 7Be and Ab specific activities, as can be seen
in Fig. 2. All of them follow a seasonal and cyclical behavior. Every
year the maximum values were found during summer and the
minimum during winter, in agreement to previous works where a
~ as et al., 1999; Pin
~ ero García et al.,
similar pattern is found (Duen
2012; Baeza et al., 1996; Feely et al., 1989). This cyclical and seasonal trend is characteristic at northern middle latitudes (60ºN30ºN) (Preiss et al., 1996).
The higher 7Be specific activity is found between May and
August, reaching its highest activity in July (summer). Lower values
are found between September and April, with the lowest activity
being attained in December (winter). This pattern is clearly seen in
Fig. 3a, where the monthly average of 7Be specific activity during 10
years is presented. This cyclical and seasonal trend of 7Be was
~ as et al., 1999, 2009;
already reported in previous studies (Duen
~ ero García et al., 2012; Likuku, 2006). The
Tositti et al., 2014; Pin
highest 7Be specific activities in summer are due to the increase of
exchange from the stratosphere to the troposphere through the
tropopause, because of its displacement to higher atmospheric
levels during this season. Besides, the vertical transport of air
masses is augmented within the troposphere through convection
movements cause by the atmospheric low stability, due to variation
~ ero García
in temperature in the troposphere (Tositti et al., 2014; Pin
et al., 2012).
This kind of seasonal variation is also reproduced by the Ab
specific activity average, its monthly variation is shown in Fig. 3b.
Its behavior is similar to that of 7Be, namely begins to increase in
May, presenting its highest value in August and then, starts to
decrease in September, presenting its lowest value in February. A
~ as et al. (2009) in a study made
similar trend was reported by Duen
laga (Spain) with the highest value in August and the lowest
in Ma
~ as et al., 2009).
values in October and January (Duen
Moreover, the monthly variation of the PM concentration
average is presented in Fig. 3c. Similarly, the PM concentration
average starts to increase in March and starts to decrease in
September, presenting the highest value in the same month as the
Ab specific activity (August) and the lowest value in the same
month as the 7Be specific activity (February).
Inspired by the similitude in the temporal evolution of Ab and
7
Be specific activities and the PM concentration, the correlations for
these parameters are computed and presented in Fig. 4.
A positive correlation was found between 7Be specific activity
and PM concentration (R ¼ 0.59, Fig. 4a). The correlation between
Ab specific activity and PM concentration was positive as well
(R ¼ 0.73, Fig. 4b). This behavior can be explained since air particulate matter (PM) is the major responsible of the movement and
transport of contaminants in the atmosphere, and particularly of
radionuclides (Atwood, 2010). The specific activities of 7Be and Ab
are also significantly positive correlated (R ¼ 0.85, Fig. 4c) among
them. Given that 7Be and Ab specific activities in PM samples were
proportional in air throughout all seasons, it can be considered that
PM of anthropogenic origin is not a substantial contributor to the
quantity of these radionuclides in air (Chao et al., 2014). A positive
correlation between 7Be and Ab specific activities was also reported
in four Taiwanese cities studied by Chao et al. (2014), although the
correlation coefficient was lower than the one found in this study
(Chao et al., 2014).
The seasonal variation for 7Be and Ab specific activities points to
a potential effect of meteorological factors on these activities. Besides, the PM concentration was analyzed, because the increase of
the pollution dispersion or the accumulation of the air pollutants
depends on the meteorology (Pataud et al., 2010). Such relation~ as et al.,
ships have already been shown in several works (Duen
ndez et al., 2008; Krajny et al., 2014). Also the sun1999; Herna
spot number (solar activity) was evaluated, since 7Be specific ac~ ero-García and Ferro-García,
tivity has a cosmogenic origin (Pin
2013).
Fig. 2. Temporal evolution of 7Be, Ab specific activities and PM concentration monthly average at the University of the Balearic Islands during 10 years.
M. Rodas Ceballos et al. / Chemosphere 152 (2016) 481e489
485
Fig. 3. Monthly average at the University of the Balearic Islands during 10 years for a) 7Be specific activity, b) Ab specific activity, c) PM concentration.
All significant correlations (p < 0.05) between all these magnitudes are represented in Fig. 5a. Among the meteorological variables considered in this part of the study, only temperature, relative
humidity and precipitation seem to be significantly correlated, as
can be seen in Fig. 5a. A high temperature value indicates an increase of the 7Be (R ¼ 0.38) and Ab (R ¼ 0.57) specific activities and
PM concentration (R ¼ 0.52) by means of the seasonal fluctuations
of the mixing layer height, as well as the intense vertical exchange
in the lower troposphere occurring in the warm season (Tositti
et al., 2013). The effect of washout by rainfall is also clear, with
R ¼ 0.19 (Ab) and R ¼ 0.21 (7Be). For large precipitation episodes,
the activities are lower than in periods with less precipitation
(Lozano et al., 2012). Periods of high relative humidity are also
inversely correlated to episodes of high specific activities of Ab
(R ¼ 0.35) and 7Be (R ¼ 0.36).
On the other hand, atmospheric pressure and wind speed are
only significantly correlated among them and with other meteorological variables, but not with 7Be and Ab specific activities, which
are the main focus of this study.
The 7Be specific activity has been related to the solar activity
through the number of sunspots (Gordo et al., 2015). The flux of
primary galactic cosmic rays at Earth's atmosphere is influenced by
solar activity in inverse correlation to the sunspot cycle because
sun's magnetic field is stronger during sunspot maximum and
shields the Earth from cosmic rays (Gordo et al., 2015; Gorney,
1990; Lantos, 1993). When solar activity is at its maximum, a
large number of sunspots are present and then the flow of cosmic
rays reaching the Earth is the lowest in the cycle because cosmic
rays show an inverse correlation with solar activity. Because the
sun's magnetic field is very strong during the solar maximum activity, it has a shielding effect on the Earth. This process has been
used to explain the apparent relation between a decrease in the
number of sunspots and an increase in 7Be specific activity (Gordo
et al., 2015). In this study, performed for a period of time of the
same timescale of a solar cycle, no clear correlation between solar
activity and 7Be specific activity is found. Actually, the number of
sunspots is the only variable without any significant correlation
respect any other variable in the dataset. Similar results were reported previously from Poland samples, in which the solar activity
was the less important factor (Krajny et al., 2014). Nevertheless, as
can be seen in Fig. 2 there is an opposite tendency between the 7Be
specific activity and minimum of sunspots number for some
reduced periods of time, like in the summer of 2009 or the winter of
2012.
3.2. Principal component analysis
The principal component analysis (PCA) was used to identify
different sets of measurements that can be characterized by the
space that they take on a space defined by a set of new orthogonal
486
M. Rodas Ceballos et al. / Chemosphere 152 (2016) 481e489
Fig. 4. Correlations in surface air of Mallorca Island, a) 7Be vs PM concentration, b) Ab vs PM concentration, c) 7Be vs Ab.
variables called principal components (PC) (Abdi and Williams,
2010). This technique has already been successfully used to show
which environmental factor influenced the variation and classification of several radionuclides specific activities in environmental
~ ero García et al., 2012).
samples (Gordo et al., 2015; Pin
At the view of the previous section results, only those variables
with significant correlation with Ab and 7Be specific activities were
considered (PM concentration, temperature, precipitation and
relative humidity). Given the different nature and range of possible
values taken by all these variables, they were all centered and
scaled before performing the PCA. This procedure allows to properly compare the effect of each variable, because they all have
uniform variance.
Following the usual criterion, only those PCs with eigenvalues
~ ero García et al., 2012).
above one were chosen. (Pin
The screeplot in Fig. 6 shows that in our case it is enough to keep
the first two PC, which interpret 76.6% of the total dataset variance.
The rest of the components, which interpret 23.4% of the dataset
variance, are rejected.
The relation between the old and new set of variables can be
globally observed at once if the selected old variables are plotted in
terms of the two main Principal Components. In Fig. 7, a representation of the relationship of the variables with the PCs is
presented, where x-axis represents the PC1 and the y-axis the PC2.
The PC1 exhibits a negative correlation with the specific activities of
7
Be and Ab, the concentration of PM and temperature, i.e. these
variables are located in the left quadrants. Whilst PC1 has a positive
correlation with precipitation and relative humidity, i.e. they are
situated in the right quadrants. The PC2 has a negative correlation
with temperature, and positive correlations with the rest of the
variables. Three main sets of variables are clearly distinguished. PM
concentration and 7Be and Ab specific activities, all of them highly
correlated, are extremely close in the upper left quadrant. Precipitation and relative humidity are also quite close, and dominate the
upper right quadrant. Finally, temperature is located in the bottomleft quadrant, pointing in a direction almost opposite to that of
precipitation and relative humidity.
Performing a PCA allows to drastically reduce the dimensionality of the dataset while keeping most of the relevant information,
facilitating the identification of patterns and sets of measures
characterized by common properties. All this is easily achieved
representing in Fig. 7 the individual measurements in the new
space constituted by the two PC that have been kept. Considering
the meteorological variables used in this part the study, their
relation with the PCs and the typical climatology in Mallorca, we
have chosen two different sets to represent our points. The first one
M. Rodas Ceballos et al. / Chemosphere 152 (2016) 481e489
487
Fig. 5. a) Correlation matrix for 7Be (Be), Ab (Beta) specific activities, PM concentration and meteorological parameters (wind speed: WS, temperature: T, precipitation: Prec, relative
humidity: RH, sunspot number: SS and atmospheric pressure: Patm). b) Correlation matrix for 7Be, 210Pb (Pb) specific activities and meteorological parameters. The size of the circles
is proportional to the magnitude of the correlation, while its colour indicates the sign of the correlation. For all non relevant correlations (p < 0.05), its corresponding square is left
empty.
composed by those samples taken between June and September,
characterized by high temperatures and a reduced amount of
precipitation. The rest of the samples compose the second set,
which includes the months with higher precipitations (OctoberDecember), and the five left months (January-May) characterized
by mild temperatures and average amounts of precipitation. Ellipses representing the 95% confidence interval are included for both
sets. Although they partially overlap, both sets are easily identified
in this new PC space. Summer samples are mostly located in the left
quadrants. This set of measures is characterized by hot and dry
weather and relatively high Ab and 7Be specific activities and PM
concentration. The second set of measurements are mostly located
in the right quadrants, being characterized by lower temperatures
and wet weather. Also, a tendency to lower values for the specific
activities and PM concentration can be associated to this second set.
Fig. 6. PCA eigenvalues. The number above each bar represents the percentage of total
variance explain by each PC.
3.3. Radionuclide specific activities in PM10
Fig. 7. Plot with all measured samples and the projection of the variables to the PCplane.
Preliminary results for 7Be and 210Pb specific activities determination in PM10 samples during two years (2013e2015) is reported in this section. The maximum and minimum specific
activities monitored during this period were 3.4 ± 0.2
e 9.5 ± 0.5 mBq m3 for 7Be and 0.2 ± 0.1 e 1.2 ± 0.1 mBq m3 for
210
Pb.
Each sample comprises the particulate matter <10 mm obtained
during a week. Three sampling stations, namely Joan March hospital, Palmanyola and Es Garrovers, all of them located near Palma
de Mallorca city, were monitored. 7Be and 210Pb were detected in
most of the analyzed PM10 filters, 100% and 93% of them
respectively.
The test for significance of correlation coefficient was applied
and the correlation between the 7Be and 210Pb specific activities
was statistically significant (R ¼ 0.63, n ¼ 14) at 95% confidence
level. Similar behavior for these radionuclides in PM10 samples was
previously reported in other study areas by Lozano et al. (2012).
In order to study the influence of the sunspot number and some
488
M. Rodas Ceballos et al. / Chemosphere 152 (2016) 481e489
meteorological factors, temperature, precipitation, atmospheric
pressure, wind speed and relative humidity on 7Be and 210Pb specific activities, the correlation matrix among all of them is presented in Fig. 5b. Once again, only correlations at 95% of confidence
level are plotted, and the size of the circles is proportional to the
magnitude of the correlation. The temperature presents a strong
positive correlation with 7Be and 210Pb specific activities, which
demonstrates the influence of the thermal factor on the concentration of these radionuclides at the surface layer air. The air temperature is an indicator of the mixing layer height in the
atmosphere, and at higher temperatures (summer) this mixing
layer is higher (Krajny et al., 2014). Moreover, the release of the
radionuclides present in the soil is related with the rise of tem~ as
perature that increases the release of particulate matter (Duen
et al., 2009).
The wind speed present negative correlations with the specific
activities of 7Be and 210Pb. Wind disperses particles of aerosol, to
which radionuclides are attached (Krajny et al., 2014) and for this
reason tends to decrease the 7Be and 210Pb specific activities.
4. Conclusions
A monitoring of 7Be and Ab in surface air from Mallorca Island
(Spain) is reported for the first time. The results of this study, corresponding to a period of 10 years (2004e2014), present a seasonal
variation, with highest values in summer and minimums in winter.
Positive and statistically significant correlations were found between 7Be and Ab specific activities and PM concentrations. Positive
correlations were found between 7Be, Ab and PM vs temperature,
while negative ones were found between 7Be, Ab and PM vs relative
humidity and precipitation. Principal component analysis was
applied in order to study and explain the influence of the meteorological parameters on 7Be and Ab specific activities and PM. Two
PCs explained 76.6% of the total dataset variance. The results
showed a seasonal behavior of 7Be, Ab and PM, divided in two
groups: summer and the rest of the year.
The first set corresponds to warm and dry months, characterized
by relative high values of 7Be and Ab specific activities and PM
concentrations. The rest of the measurements are characterized by
mild temperatures and higher precipitation values.
In addition, a study on 7Be and 210Pb specific activities in PM10
samples from Mallorca is presented for the first time. Significant
positive correlations were found between the specific activities of
both radionuclides, indicating that they cannot be used as independent atmospheric tracers in this Island. For both radionuclides,
negative correlations with wind speed were found, showing the
dispersion of aerosol particles; while both radionuclides were
positively correlated to temperature, indicating its seasonal
behavior.
Acknowledgements
Financial support from the Spanish Ministry of Economy and
Competitiveness (MINECO) (CTM2013-42401-R) cofinanced by
European Regional Development's funds (FEDER), and from Spanish Nuclear Safety Council (CSN) are gratefully acknowledged. Authors thank the Spanish Meteorological Agency (AEMET) for the
meteorological data provided. M. Rodas acknowledges to the University of the Balearic Islands (UIB) the allowance of a grant.
References
Abdi, H., Williams, L.J., 2010. Principal component analysis. Wiley Interdiscip. Rev.
Comput. Stat. 2.
Atwood, David A. (Ed.), 2010. Radionuclides in the Environment, first ed. John Wiley
& Sons Ltd, Chichester, West Sussex, United Kingdom.
nez, A., Miro
, C., Paniagua, J.M., Rufo, M., 1996. Analysis of
Baeza, A., Delrío, L.M., Jime
the temporal evolution of atmospheric 7Be as a vector of the behavior of other
radionuclides in the atmosphere. J. Radioanal. Nucl. Chem. 207, 331e344.
Bigi, A., Ghermandi, G., 2014. Climatology of atmospheric PM10 concentration in the
Po Valley. Atmos. Chem. Phys. 14, 137e170.
Cerro, J.C., Cerd
a, V., Pey, J., 2015. Trends of air pollution in the Western Mediterranean Basin from a 13-year database: a research considering regional, suburban and urban environments in Mallorca (Balearic Islands). Atmos. Environ.
103, 138e146.
Chao, J.H., Liu, C.C., Cho, I.C., Niu, H., 2014. Monitoring of 7Be in surface air of varying
PM10concentration. Appl. Radiat. Isot. 89, 95e101.
~ as, C., Fern
Duen
andez, M.C., Liger, E., Carretero, J., 1999. Gross alpha, gross beta
activities and 7Be concentrations in surface air: analysis of their variations and
prediction model. Atmos. Environ. 33, 3705e3715.
~ as, C., Ferna
ndez, M.C., Carretero, J., Liger, E., Can
~ ete, S., 2004. Long-term
Duen
variation of the concentration of long-lived Rn descendants and cosmogenic
7
Be and determination of MRT of aerosols. Atmos. Environ. 38, 1291e1301.
~ as, C., Ferna
ndez, M.C., Can
~ ete, S., Pe
rez, M., 2009. 7Be to 210Pb concentration
Duen
ratio in ground level air in M
alaga (36.7 N, 4.5 W). Atmos. Res. 92, 49e57.
~ as, C., Orza, J.A.G., Cabello, M., Ferna
ndez, M.C., Can
~ ete, S., Pe
rez, M., Gordo, E.,
Duen
2011. Air mass origin and its influence on radionuclide activities (7Be and 210Pb)
in aerosol particles at a coastal site in the western Mediterranean. Atmos. Res.
101, 205e214.
Feely, H.W., Larsen, R.J., Sanderson, C.G., 1989. Factors that cause seasonal variation
in beryllium-7 concentrations in surface air. J. Environ. Radioact. 9, 223e249.
ndez, F., 2001. Studies on
García-Talavera, M., Quintana, B., García- Díez, E., Ferna
radioactivity in aerosols as a function of meteorological variables in Salamanca
(Spain). Atmos. Environ. 35, 221e229.
Gerasopoulos, E., Kouvarakis, G., Babasakalis, P., Vrekoussis, M., Putaud, J.-P.,
Mihalopoulos, N., 2006. Origin and variability of particulate matter (PM10) mass
concentrations over the Eastern Mediterranean. Atmos. Environ. 40,
4679e4690.
~ as, C., Ferna
ndez, M.C., Liger, E., Can
~ ete, S., 2015. Behavior of ambient
Gordo, E., Duen
concentrations of natural radionuclides 7Be, 210Pb, 40K, in the Mediterranean
coastal city of M
alaga (Spain). Environ. Sci. Pollut. Res. 22, 7653e7664.
Gorney, D.J., 1990. Solar cycle effects on the near-Earth space environment. Rev.
Geophys. 28, 315e336.
ndez, F., Hern
ndez-Aldecoa, J.C.,
Herna
andez-Armas, J., Catal
an, A., Ferna
Karlsson, L., 2005. Gross alpha, gross beta activities and gamma emiting radionuclides composition of airborne particulate samples in an oceanic island.
Atmos. Environ. 39, 4057e4066.
ndez, F., Rodríguez, S., Karlsson, L., Alonso-Pe
rez, S., Lo
pez-Pe
rez, M., HerHerna
nandez-Armas, J., Cuevas, E., 2008. Origin of observed high 7Be and mineral
dust concentrations in ambient air on the Island of Tenerife. Atmos. Environ. 42,
4247e4256.
Krajny, E., Osrodka, L., Wojtylak, M., 2014. Meteorological conditions for the variability of 7Be and 210Pb concentrations in surface layer air in Poland. J. Environ.
Sci. Eng. Technol. 2, 14e25.
L'Annunziata, M.F. (Ed.), 2012. Handbook of Radioactivity Analysis, third ed.
Elsevier, Amsterdam, The Netherlands.
Lantos, P., 1993. The sun and its effects on the terrestrial environment. Radiat. Prot.
Dosim. 48, 27e32.
Likuku, A.S., 2006. Factors influencing ambient concentrations of 210Pb and 7Be over
the city of Edinburgh (55.9 N, 03.2 W). J. Environ. Radioact. 87, 289e304.
ndez-Ceballos, M.A., San Miguel, E.G., Adame, J.A., Bolívar, J.P.,
Lozano, R.L., Herna
2012. Meteorological factors influencing the 7Be and 210Pb concentrations in
surface air from the southwestern Iberian Peninsula. Atmos. Environ. 63,
168e178.
~ oz, M.V., 2002. El ritmo estacional del viento en el arco Mediterra
neo espan
~ ol e
Mun
Islas Baleares. Papeles Geogr. 35, 171e192.
Pataud, J.P., Van Dingenen, R., Alastuey, A., Bauer, H., Birmili, W., Cyrys, J., Flentje, H.,
Fuzzi, S., Gehrig, R., Hansson, H.C., Harrison, R.M., Herrmann, H.,
Hitzenberger, R., Hüglin, C., Jones, A.M., Kasper-Giebl, A., Kiss, G., Kousa, A.,
€ schau, G., Maenhaut, W., Molnar, A., Moreno, T.,
Kuhlbusch, T.A.J., Lo
Pekkanen, J., Perrino, C., Pitz, M., Puxbaum, H., Querol, X., Rodríguez, S.,
Salma, I., Schawarz, J., Smolik, J., Schneider, J., Spindler, G., ten Brink, H.,
Tursic, J., Viana, M., Wiedensohler, A., Raes, F., 2010. A European aerosol phenomenology e 3: Physical and chemical characteristics of particulate matter
from 60 rural, urban, and kerbside sites across Europe. Atmos. Environ. 44,
1308e1320.
~ ero García, F., Ferro García, M.A., Drozd
_ zak,
_
Pin
J., Ruiz-Sambl
as, C., 2012. Exploratory
data analysis in the study of 7Be present in atmospheric aerosols. Environ. Sci.
Pollut. Res. 19, 3317e3326.
~ ero-García, F., Angeles
Pin
Ferro- García, M., Influence of air masses origin on
radioactivity in aerosols. Proceedings of the 2nd Iberian Meeting on Aerosol
Science and Technology-RICTA 2014. Tarragona, Spain.
~ ero-García, F., Ferro-García, M.A., 2013. Evolution and solar modulation of 7Be
Pin
during the solar cycle 23. J. Radioanal. Nucl. Chem. 296, 1193e1204.
lie
res, M.A., Pourchet, M., 1996. A compilation of data on lead 210
Preiss, N., Me
concentration in surface air and fluxes at the air-surface and water-sediment
interfaces. J. Geophys. Res. 101, 28847e28862.
R Core Team, 2015. R: a Language and Environment for Statistical Computing. R
Foundation for Statistical Computing, Vienna, Austria. https://www.R-project.
org/.
M. Rodas Ceballos et al. / Chemosphere 152 (2016) 481e489
SILSO, World Data Center-sunspot number and long-term solar observations, Royal
Observatory of Belgium, on-line Sunspot Number catalogue: http://www.sidc.
be/silso/datafiles. (accessed 18.02.16.).
Spanish
Meteorological
Agency
(AEMET):
http://www.aemet.es/es/
serviciosclimaticos/datosclimatologicos/valoresclimatologicos?l¼B278&k¼bal.
(accessed 15.02.16.).
489
Tositti, L., Riccio, A., Sandrini, S., Brattich, E., Baldacci, D., Parmeggiani, S.,
Cristofanelli, P., Bonasoni, P., 2013. Short-term climatology of PM10 at a high
altitude background station in southern Europe. Atmos. Environ. 65, 142e152.
Tositti, L., Brattich, E., Cinelli, G., Baldacci, D., 2014. 12 years of 7Be and 210Pb in Mt.
Cimone, and their correlation with meteorological parameters. Atmos. Environ.
87, 108e122.