Author’s Accepted Manuscript

Chemical composition of rainwater in the northeastern Romania, iasi region (2003-2006) CeciliaArsene, Romeo Iulian Olariu, Nikolaos Mihalopoulos PII: DOI: Reference:

S1352-2310(07)00769-8 doi:10.1016/j.atmosenv.2007.08.046 AEA 7767

To appear in:

Atmospheric Environment

Received date: Revised date: Accepted date:

20 June 2007 28 August 2007 28 August 2007

www.elsevier.com/locate/atmosenv

Cite this article as: Cecilia Arsene, Romeo Iulian Olariu and Nikolaos Mihalopoulos, Chemical composition of rainwater in the north-eastern Romania, iasi region (2003-2006), Atmospheric Environment (2007), doi:10.1016/j.atmosenv.2007.08.046 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting galley proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Chemical composition of rainwater in the north-eastern Romania, Iasi region (2003-2006) Cecilia Arsene1,2,*, Romeo Iulian Olariu1 and Nikolaos Mihalopoulos2

1

Department of Inorganic and Analytical Chemistry, Faculty of Chemistry, Al.I.Cuza University of Iasi, Carol I

11, 700506 Iasi, Romania 2

Environmental Chemical Processes Laboratory, Department of Chemistry, University of Crete, Voutes 71003

t p

Heraklion, Greece *

i r c

Corresponding author: E-mail address: [email protected] (C. Arsene). Tel.: +40-232-201354; Fax: +40-232-

201313

s u

Abstract

n a

Chemical composition of rainwater was studied in the north-eastern Romania, Iasi region, and the concentrations of major inorganic and organic ions were measured in samples collected

m

between April 2003 and December 2006. The pH of the rainwater is 5.92 (volume weighted

d e t p

mean average, VWM) suggesting a sufficient load of alkaline components neutralizing its acidity. On average, 97% of the acidity in the collected samples is neutralized by CaCO3 and

e c

NH3. Clear seasonal variations were observed for some of the identified ions (e.g., SO42-, NO3-, Ca2+, NH4+). The data obtained during this work revealed that both concentrations and

c A

fluxes of anthropogenic source related ions (e.g., SO42-, NO3-, and NH4+) are among the highest reported for European sites. It is shown that meteorology and long-range transport processes may concur to their high levels.

Keywords: Rainwater, Ionic Composition, Deposition Flux, Romania

1

1. Introduction Rain is an efficiently scavenging process for pollutants present in the atmosphere both in the gas and aerosol phases and provides a path for essential nutrients to reach terrestrial and aquatic ecosystems. Wet deposition has several harmful effects on various environmental compartments, with acidification as a major concern (Brimblecombe, 2001). Systematic investigation on the rainwater is chemical composition, and an assessment of the different emission sources that control it, which are prerequisite knowledge for the reduction of the

t p

consequences of pollution on ecosystems. Various factors can affect the chemical

i r c

composition of precipitation at a given site, including local emissions, regional scale pollutant

s u

transport processes, sea level elevation and the meteorological conditions (Tang et al., 2005; Rocha et al., 2003; Sickless and Grimm, 2003).

n a

Over the last twenty years, rainwater chemistry has been subject to intense research and

m

many studies on the chemical composition and long-term temporal trends of precipitation have been conducted worldwide (Tu et al., 2005; Rogora et al., 2004; Sickles and Grimm,

d e t p

2003; Avila and Roda, 2002; Seto et al., 2002; Tuncer et al., 2001; Loye-Pilot et al., 1986). However, for eastern Europe there are very few works concerning the chemical composition

e c

of rainwater, mainly in Poland and the Czech Republic (Polkowska et al., 2005; Hunova et al.,

c A

2004; Hlawiczka et al., 2003; Bridges et al., 2002). Moreover, despite the possible environmental consequences of wet deposition, for Romania there is a single report concerning the chemical composition of rainwater in a mountain area (Bytnerowicz et al., 2005). The present study reports on data of the chemical composition of precipitation in Iasi, a large urban area in north-eastern Romania. Sampling has been conducted from April 2003 to December 2006. This is the first attempt to investigate the chemical composition of rainwater in the area. The levels of identified ionic species, their temporal variation, possible sources

2

and deposition fluxes are presented and discussed. The data significantly contributes to the very limited knowledge on rainwater quality in the east Europe.

2. Experimental 2.1. Sampling and analysis Iasi, a large urban city with about 400,000 inhabitants, lies in the north-eastern part of Romania (47o20ǯN latitude and 27o60ǯE longitude) at an altitude of ~100 m above sea level

t p

(Figure 1). The climate of the region is cold in winter and hot in summer (transitional between

i r c

temperate and continental) with uniformly distributed precipitation throughout the year. After 1989, there is very little industrial activity, mainly situated at about 10-15 km southwest to

s u

Iasi. During winter, traffic and domestic heating appear to be the most important local

n a

emission sources. The Black Sea is located southeast of Iasi at a distance of about 400 km.

2.1.1. Sample collection and site description

d e t p

m

From April 2003 to December 2006, sampling was undertaken at a location fulfilling the selection criteria for the chemical composition of rainwater monitoring at an urban site

e c

(WMO Report 160, 2004). Rainwater, wet only, was sampled at the Al.I Cuza University of Iasi, about 30 m above ground. As the topography of the city is dominated by an important

c A

number of hills, and the University itself is located north-east from the city centre on one of them, the sampling location is at a high elevation compared with many other surroundings. A self-made collector, manually operated, was installed at 1.5 m above the roof and far from any possible local contamination sources. Neither vents nor other exhausts are located in the near vicinity of the sampling system. Large diameter polypropylene funnels (d = 21 cm) fitted to polyethylene bottles (two distinct systems have been deployed to collect the rain) were used for the collection of the

3

rainwater. Before and after each sampling, the entire collection and storage equipment was thoroughly washed and rinsed several times with triple distilled water. To prevent/minimize any possible contamination from dry deposition, the collection surface was exposed just prior to the onset of the precipitation event. A sampling event is defined as the sample collected entirely from the onset till the end of the precipitation. The precipitation samples were collected immediately after cessation of the rainfall. In the laboratory, each sample was split in two aliquots for the pH measurements and the ionic

t p

composition analysis. The pH values were measured using an earlier calibrated pH-meter

i r c

(MeterLab ION450). The rainwater samples were filtered through 0.45 Pm pore size (47 mm diameter) Millipore cellulose filters, treated with chloroform to prevent microbial

s u

decomposition and, finally, stored in a refrigerator at 4 oC for chemical analysis.

n a

2.1.2. Sample analysis

m

Rainwater samples were analyzed for the main inorganic and organic ions using ion

d e t p

chromatography (Dionex DX 500) as described by Economou and Mihalopoulos (2002). Along with the samples were analysed several blanks kept under similar conditions as the

e c

rainwater samples. Blank matrices contained pure and chloroform treated triple distilled water, a last rinse of the collector components and a preserved dry deposition blank obtained

c A

by washing the funnel after 12 h of exposure. The concentrations of the measured ions are reported after subtraction of their blank values (less than 5% of the mean value).

2.1.3. Trajectory analysis Air mass origin was assigned on the base of the 5 days backward trajectories calculated by using the HYSPLIT 4 model of the Air Resources Laboratory of NOAA and FNL database (Draxler and Rolph, 2003). In the calculation, levels of 1000 and 3000 m were used. They

4

correspond to ~880 and ~700 hPa, respectively, barometric levels at the end point of the trajectories. Based on backward trajectory analysis, eight sectors of air mass origin were distinguished. The contribution of each sector, in terms of air masses and rain, is depicted in Figure 1. The prevalence of the northern (NE-N-NW) sector which contributes to almost 36% of the rain events on an annual basis is clearly seen. The contribution of the southern (S-SE-SW) sector is about 35%, meanwhile western and eastern sectors account for about 15% each.

t p

i r c

3. Results and discussion

A total of 178 samples were collected from April 2003 to December 2006 (2003, n = 37;

s u

2004, n = 53; 2005, n = 44; 2006, n = 44). Of the total number of samples, 10% has been

n a

discarded from final analysis (Table 1) due to either very low volume (~2-4 mL which does not allow a complete chemical characterization) or poor quality according to WMO Report

m

160 (2004) for flagging precipitation results with poor ion balances (> 20%). A study of

d e t p

seasonal variability of rainfall height showed that rain is uniformly distributed throughout the year, with the exception of summer (from May to August), when monthly rain height exceeds

e c

the mean value by a factor of up to 2.3. Annual rainfall ranged between 350 mm in 2006 to about 545 mm in 2005, and an average of 420 mm has been determined over the entire period.

c A

The Romanian National Administration of Meteorology (RNAM) reports an annual precipitation average of about 521 mm at Iasi. A direct comparison between the annual rainfall height collected during this project in 2004 (500 mm) and that reported by the RNAM for the same period (578 mm) reveals a 15% difference between the two values, indicating an almost complete collection of rain events at our site. The difference is mainly observed during the warm season and it might be due to the variability in rain intensity between the two sampling locations (our site and the RNAM site).

5

3.1. Chemical composition Statistical information, minimum and maximum concentrations and average VWM levels (Peq L-1) of the measured anions (Cl-, Br-, NO3-, PO43-, SO42-, HCO2- (formate), C2H3O2(acetate), C2O42- (oxalate), C3H5O2- (propionate), C3H3O3- (pyruvate), CH3SO3- (methane sulphonate)) and cations (Na+, NH4+, K+, Mg2+ and Ca2+) are summarized in Table 1. Table 1 also presents data of the indirectly derived HCO3- (see discussion below).

t p

The arithmetic mean of the VWM values over 2003-2006 both for the 6cations (232±131)

i r c

-1

and for the 6anions (200±93) (±V indicate an excess of 32 Peq L in cations, which suggests the existence of an unmeasured species in the analyzed samples (most probably CO32-/HCO3-,

s u

as organic anions were measured in the present work).

n a

To check this assumption, the ratio between the determined cations and anions has been investigated. Slopes higher than unity indicate a significant anion deficit in the ionic balance.

m

The linear regression of Ȉcations vs. Ȉidentif. anions gives a slope of 1.24 (r2 = 0.89; Figure 2a). In

d e t p

Figure 2b the variation of (Ȉcations – Ȉidentif. anions) is depicted as a function of the measured pH and a clear increase in the anionic deficit is observed for the samples in the alkaline range

e c

(~78% of the rainwater samples exhibited pH values in the alkaline range). Figure 2b depicts also the theoretical profile for HCO3- derived from the measured pH, equilibria constants and

c A

CO2 partial pressure. Indeed, both the theoretical HCO3- and the experimentally derived anionic deficit follow the same pattern, an observation, which increases our confidence that the assumed missing anion is in the form of HCO3-/CO32-. Previous studies suggest that these ions are most likely associated with Ca2+ and Mg2+ (Bardouki et al., 2003). In the present work (6cations-6identif.

anions)

vs. 6(Ca2++Mg2+) (VWM

equivalent concentrations) showed a significant correlation (r2 = 0.74) with a slope of 0.44 (Figure 2c). Such a significant correlation allows the estimation of the missing ion (HCO3-

6

/CO32-) from the measured Ca2+ and Mg2+ concentrations and a significant improvement in the ionic balance is thus achieved with a ratio of Ȉcations/Ȉanions of 1.00 and r2 = 0.95 (Figure 2d). The data in Table 1 shows that concentrations of the major ionic species follow the order Ca2+(nss-Ca2+) > SO42-(nss-SO42-) > NH4+ > NO3- > Na+ § Cl- > HCO3- > Mg2+ § K+ § PO43- > C2H3O2- > HCO2- § C2O42- (other constituents are very minor). Figure 3 presents the percentage distribution of the ionic composition in rainwater at Iasi. Calcium is the most important contributor to the ionic mass (23.1%) and bicarbonate accounts for 7.0%.

t p

Anthropogenic related ions (SO42-, NO3- and NH4+) account all together for 37.7%. The

i r c

+

contribution of the H ion is only 0.4% and this is probably a result of extensive acidity neutralization. Although the monitoring site is surrounded by agricultural fields, known as

s u

important sources of organic acids in the atmosphere, their contribution to the rainwater ionic

n a

composition is very small (less than 5%).

Equivalent SO42-/NO3- and SO42-/Cl- ratios were also estimated. The values of the average

m

SO42-/NO3-, and SO42-/ Cl- equivalent ratios (VWM) are 2.54 and 3.49, respectively. Both

d e t p

determined values would suggest that the acidity measured at the monitoring site is mainly in the form of H2SO4 rather than HNO3, HCl or organic acids. The latest are assumed to contribute ~ 1% to the total acidity and the equivalent ratio of SO42-/6organics is 11.23.

e c

An interesting observation arises from the yearly variability of the ratio SO42-/NO3-

c A

showing a decreasing tendency from 2003 to 2006. Indeed, the ratio is 2.65 for 2003, 2.07 for 2004, 2.61 for 2005 and 1.53 for 2006, a variability which might indicate control in SO2 emissions. However more data is clearly needed to reach a definite conclusion. Table 2 compares the chemical composition of rainwater as determined in the present study, compared with values reported for different regions in Europe. The only available study for Romania, reported by Bytnerowicz et al., (2005), deals with rainwater composition in the Retezat Mountains “characterized by the relatively clean air, acidic precipitation, and

7

healthy, growing forests”. An introspection of the data in Table 2 reveals that between our site and the Retezat Mountains, there are differences in the chemical composition of the analyzed rainwater samples. Higher SO42- and K+ concentrations are reported for Retezat compared to our site, an observation which might reflect a more important influence from local pollution sources (e.g. SO42-) or vegetation (e.g. K+). On the other hand, lower values for NO3- and Ca2+ reported for Retezat can be due to very little traffic and higher soil coverage at this mountain site.

t p

From Table 2, significant differences can also be observed between our data and other

i r c

reports. For organic anions comparison is very difficult since data is either scarce (HCO2- and -

-

-

-

C2H3O2 ) or missing (C3H5O2 , C3H3O3 and CH3SO3 ). For anthropogenic (SO42-, NO3-) and

s u

soil (Ca2+) related ions, our data is among the highest compared to those reported for sites at

n a

about the same altitude above sea level and similar geographical characteristics (Poland is a representative case; Polkowska et al., 2005). For NH4+, our value is close to that reported for

m

Turkey (Tuncer et al., 2001) and this might be a good indication of common agricultural

d e t p

practices in the two countries.

Although for Greece, a high concentration of Ca2+ is reported, the acidity appears to be

e c

much higher compared to Romania and Turkey. However, should be noted that for Greece the concentration of NH4+, another alkaline species that is involved in neutralization of the

c A

atmospheric acidity, is among the lowest reported for any other sites in Europe (Glavas and Moschonas, 2002).

3.2. pH distribution and acidity neutralization Figure 4a shows the seasonal variation both of pH and H+ as an average of the monthly volume weighted mean (VWM) values determined all over the investigated period (April 2003-December 2006). The monthly average pH was estimated from the VWM concentration

8

of H+ observed from April 2003 to December 2006 (pH = -logH+(VWM)). A value of 5.92 has been determined for the pH, which is slightly higher than the value of 5.6 expected for equilibrium of cloud water with atmospheric CO2. Figure 4a clearly shows that the seasonal variation of acidity has well defined maxima over both the summer and the winter period. The maximum value during the winter is most probably due to the intensification of coal burning processes for heating purposes, while the high value during the warm season might be a result of more intense anthropogenic activities (traffic exhaust, agricultural).

t p

The frequency distribution for the pH is depicted in Figure 4b. Of all of the considered

i r c

raining events only ~13% were in the acidic range (bellow 5.6). The more frequent alkaline rainwater probably results from high loading of cations in the air masses reaching the site.

s u

Lower pH values were observed in all those events occurring after previous frequent rains.

n a

This observation is not too surprising as it is expected that after such events, the dust in the atmosphere settles lowering the content of the alkaline ions.

m

The most acidic precipitation event (pH of 4.26) occurred on July 17, 2006. This was the

d e t p

result of an air mass originating from NE Europe (crossing Russia, Belarus and Ukraine) and followed a former precipitation event that occurred on July 16, 2006. That event characterized

e c

by a higher pH value (5.95) was associated with an air mass originating from Greenland, crossing over Norway, Sweden, Lithuania, Belarus and Ukraine. Before reaching the site, the

c A

air mass front associated with the event of July 17, 2006, travelled entirely under 500 m (only close to the monitoring site elevated at 1000 m). During this event, the chemical composition of the rainwater consisted of the highest NO3- concentration while the levels of other ions, such as SO42-, Cl- and Ca2+, were among the lowest over the year. These examples indicate that the chemical composition of the rainwater at the investigated site is strongly affected by the transport of atmospheric chemical constituents from distant source regions. Although the rainfall height of the July 16, 2006, event was five times higher

9

than that of July 17, 2006, the dilution effect cannot account for the observed difference. Information on the magnitude of the rainwater neutralization can be obtained from the (H+)/(SO42-+NO3-) equivalent ratio, which is expected to be close to 1 if the acidity caused by H2SO4 and HNO3 is not neutralized. In the present study, an average (H+)/(SO42-+NO3-) ratio of about 0.03 was estimated, which would suggest ~97% neutralization of the acidity on an annual basis. The large extensive neutralization of the acidity is most probably due to transported alkaline dust particles and other alkaline chemical constituents with a buffering

t p

capacity in the rainwater. The present determined value is in very good agreement with the

i r c

95% acidity neutralization reported by Tuncer et al. (2001) for Central Anatolia. However, as 2-

+

-

Figure 5 shows, in the present study the (H )/(SO4 +NO3 ) equivalent ratio showed a very

s u

clear seasonal variability with maxima during the winter (January) and summer time (July).

n a

Figure 5 also presents the seasonal variability of the neutralization factors (NF) for Ca2+, NH4+, and Na+, derived as an average of the monthly VWM values, using the equation

NFX

SO 4

2

X   NO 3

d e t p

m

where X denotes the chemical component of interest with all concentrations in Peq L-1. The highest NF, with a value of 0.81±0.38 corresponds to Ca2+, followed by NH4+ (0.47±0.21) and

e c

Na+ (0.35±0.19). Potassium and magnesium have very low NF with values of 0.12±0.06 and

c A

0.13±0.06, respectively. However, for Ca2+ there seems to be a seasonal variability in the magnitude of the NF with maxima especially during months of frequent high intensity winds and a reduced number of precipitation events.

3.3. Marine toward soil/anthropogenic contribution

To obtain an information on the possible sources of Cl-, SO42-, K+, Ca2+ and Mg2+, enrichment factors (EF) relative to Na+ were calculated using the following formula:

10

EF

X

Na  rain X Na  seawater

where X is the ion of interest. The results are presented in Table 3. Apart from Cl-, all the investigated compounds are enriched relative to Na+ indicating significant influence from anthropogenic and soil related sources rather than sea-salt. The values of the EF follow the order EF(Ca2+) >> EF(SO42-) § EF(K+) >> EF(Mg2+) > EF(Cl-). The value of the EF for Cl- is slightly smaller compared to 1

t p

(0.89) and suggests either Cl losses during long range transport or the existence of a small

i r c

local source of Na+ in addition to sea-salt, possibly from soil.

s u

3.4. Temporal variability of measured concentration

Figure 6 presents the seasonal variability of the average of the monthly VWM concentration

n a

values, over the entire period (April 2003-December 2006), for selected inorganic anions (a),

m

cations (b) and organic species (c). SO42-, a component related to intense coal burning, shows

d e t p

a clear minimum during the warm season and a clear maximum during the cold period. NO3-, derived mainly from vehicle exhaust and industrial activities, exhibited a similar trend to SO42-. Moreover, the N-NO3-/S-SO42- molar ratio also presents seasonal variability with a

e c

clear maximum during the warm period. This observation may highlight the role of increasing

c A

NOx emissions from mobile sources during summer time (Figure 6a). Most of the cationic species also present a seasonal variability. Ammonium for instance, shows maxima during months associated with the most intense traditional agricultural activities (February, June and September). Elements of natural origin (Ca2+ mainly related to soil and dust particles) have lower concentrations during the cold season, when an excessive decrease in temperature may contribute to soil freezing and prohibit the particle suspension process (Figure 6b). The lower Ca2+ value during the warm season could be the result of more frequent raining events during this period of the year.

11

From the identified organic components, C2H3O2- and HCO2- seem to present clear seasonal variability (Figure 6c). For C2H3O2- and HCO2-, their seasonal distribution suggest important contributions during summer, either from more intense photochemistry leading to the formation of acetic acid and/or more intense emission from soil and vegetation. During the cold period, biomass/wood combustion may also increase their levels, beside the automobile exhaust emissions. Figure 6c shows that CH3SO3-, (MS), a biogenic tracer of marine emissions, presents a specific pattern with a clear summer to fall maximum. This pattern is

t p

similar to the one showed by the main precursor of MS, dimethyl sulphide (CH3SCH3, DMS)

i r c

emitted from marine plankton (Turner et al., 1988). Such an observation indicates that despite the distance from Black Sea, a marine influence is clearly detected at our location.

s u

Data in Figure 6, however, suggest that the amount of precipitation can influence the levels

n a

of the main ions in rainwater. The higher the precipitation amount, the smaller the ionic level, especially due to the dilution effect. In the present study, a correlation of ion concentrations

m

with rainfall height has been investigated using log-log plots. Although a negative correlation

d e t p

has been observed between the rainfall height and the concentration of all identified ions, correlation coefficients were found to be statistically significant only for Cl-, Na+ and Mg2+

e c

(~0.50), suggesting that the precipitation amount either plays a minor role or cannot account as the only factor controlling the measured concentrations of ionic species.

c A

3.5. Statistical investigation of ion correlations. Principle component analysis (PCA)

The relationship between various pairs of identified ionic species has been investigated using multiple linear regression analysis (Table 4). SO42- correlates mainly with ions such as NH4+ and Ca2+, which are important neutralizing agents of rainwater alkalinity. NO3- correlates well with NH4+, which implies common sources from agricultural activities, but also with Ca2+ and Mg2+, which would underline the role of these cations in neutralizing HNO3 acidity in

12

rainwater. The significant correlation between NO3- and SO42- toward HCO3- would also suggest a common soil-related source. As for the cationic species, the pair Ca2+ vs. Mg2+ shows the highest correlation, an observation which would imply contribution from a common source, most probably from soil and unpaved roads. From the identified organic species only the HCO2- vs. C2H3O2- pair showed a good correlation (0.93), meanwhile CH3SO3- (MS-) correlates to some extent only with Cl- and Na+ possibly suggesting their common marine origin.

t p

Correlation analysis revealed that for a few pairs of ion, the slope is highly dependent on 2+

the air mass origin. The case of the nss-Ca

2-

i r c

-

vs. 6(nss-SO4 +NO3 ) is the most interesting

(Figure 7). The highest value corresponds to air masses originating from the N and E sectors

s u

(air masses crossing large continental areas). The lowest value for the NE sector is probably a

n a

result of air masses with less loading of alkaline soil related components due to heavily forestation in the area.

m

Factor analysis with extracted principal component after normalized Varimax rotation was

d e t p

used to differentiate between anthropogenic and natural sources affecting the chemical composition of the analyzed rainwater. After applying the Varimax rotation, loading factors

e c

greater than 1.00 were considered as significant in the interpretation of the obtained results. From the performed analysis, 4 factors were extracted, explaining 73.6% of the observed

c A

variability (Table 5). The first factor (F1), which explains 29.5% of the total variance, shows high loading factors for components of soil and suspended particle origin (HCO3-, Ca2+, Mg2+). Included in the same factor are some anthropogenic source related ions (SO42-, NO3-), which may indicate absorption of their gaseous precursors from alkaline media. The second factor (F2), explaining 19.8% of the total variance, mainly includes the organics apart oxalate, which is distributed among F2, F3 and F4. The sources of organics might be both of natural and anthropogenic origin (vegetation, exhaust or combustion; Pun et

13

al., 2000). A third factor (F3) explains 13.5% of the total variance and contains sea-salt related elements (Cl-, Na+). The fourth factor (F4, with a 10.9% contribution to the total variance) includes PO43-, K+ and C2O42-, most probably as a result of anthropogenic activities related to the usage of fertilizers and combustion processes in incinerators.

3.6. Wet deposition fluxes

Wet deposition fluxes for the quantified ions were calculated for each event by multiplying

t p

their concentration with the precipitation height. For the 2004-2006 average period, values for

i r c

the deposition fluxes are presented in Table 6. The fluxes for ions of anthropogenic origin 2-

-

indicate very high deposition in the following order: SO4 > NO3 > NH4+. For comparison

s u

purposes, values reported for a few other sites in Europe are also included in Table 6. The

n a

fluxes, as determined in the present study, are in good agreement with reported values for similar sites. Regarding ions related to anthropogenic or soil suspended particles, our

m

determined fluxes are among the highest but, generally, in good agreement with values

d e t p

reported for other north European sites (see for example, data in Table 6 for Poland, the Czech Republic and even Spain, as the investigated site is located in the near vicinity of

e c

Barcelona).

To highlight the areas with the most important contribution to the calculated fluxes, Figure

c A

8 depicts the sectors’ percentage contribution in terms of rain height and deposition fluxes for selected ions. For Ca2+ the highest contribution in terms of fluxes relative to rain height is associated with air masses crossing very large continental areas (N and E sectors). A first introspection of the data presented in Figure 8 suggests a more important contribution from the NW and W sectors regarding anthropogenic related elements (SO42-, NO3-, NH4+). However, higher fluxes of these elements for the NW and W sectors, when compared with other geographical sectors, is most probably also a result of important local

14

influences, as the sampling site is located east of the Oriental Carpathians chain that could constrain, to some extend, the pollution plume transported from western Europe.

4. Conclusions

This work reports on variation of the ionic composition of precipitation in Iasi area (NE Romania) during 2003-2006. To our knowledge, this is the first effort to estimate the factors controlling the ionic composition of precipitation at an urban location in Romania. High

t p

levels of ions related to natural and anthropogenic sources were determined in analyzed

i r c

rainwater samples. Although the concentrations of the ions responsible of the acidity were high, the average pH was 5.92, which indicates extensive neutralization of the acidity

s u

(roughly by 97%). Neutralization occurred throughout the year but it was more pronounced

n a

during the periods with a reduced number of precipitation events.

The ratios of various ions in rainwater relative to their seawater values were higher for

m

most of the studied ions, indicating significant influence from anthropogenic and soil related

d e t p

sources. Natural emission sources associated with dust from soil erosion, through their main 2+

2+

-

2-

constituents (Ca , Mg , HCO3 /CO3 ), appear to play a significant role in controlling the

e c

chemical composition of the rainwater in Iasi region. Anthropogenic source related components (SO42-, NO3-, NH4+) may also play an important role with contributions to the

c A

chemical composition of the rainwater highly dependent on the characteristics of the geographical sectors.

Some of the identified ions show clear seasonal variations (with a similar pattern from one year to another). The most representative component of anthropogenic sources (intensive coal burning), i.e. sulphate (SO42-), showed deviating behaviour, with a clear minimum during the warm period and a clear maximum during the cold season. Soil-related components, especially Ca2+, present maximum levels in rain events associated with air masses crossing

15

very large continental areas. For the major identified ions in the rainwater at Iasi, high average concentrations and deposition fluxes were determined. These observations, corroborated with meteorological conditions, suggest that the impact of pollution transport from distant emission sources may be important at the investigated site, as local pollution alone cannot account for the high levels of ion concentrations in the analyzed rainwater.

t p

Acknowledgment

i r c

Pavlos Zarbas is gratefully acknowledged for his support in the ion chromatographic analysis. NOAA Air Resources Laboratory (ARL) for the provision of the HYSPLIT transport and

s u

dispersion model is acknowledged (http://www.arl.noaa.gov/ready.html). We are grateful for

n a

the instructive comments from the two anonymous referees that considerably helped to improve the paper. This work has been supported by GRST-Greece and MCT-Romania in the

m

frame of the Greek-Romanian bilateral collaboration.

d e t p

e c

c A

16

References

Avila, A. and Roda, F, 2002. Assessing decadal changes in rainwater alkalinity at a rural Mediterranean site in the Montseny Mountains (NE Spain). Atmospheric Environment 36, 2881-2890. Bardouki, H., Liakakou, H., Economou, C., Sciare, J., Smolik, J., Zdimal, V., Eleftheriadis, K., Lazaridis, M., Dye, C., Mihalopoulos, N., 2003. Chemical composition of size-

t p

resolved atmospheric aerosols in the eastern Mediterranean during summer and winter.

i r c

Atmospheric Environment 37, 195–208.

Bridges, K.S., Jickells, T.D., Davies, T.D., Zeman, Z., Hunova, I., 2002. Aerosol,

s u

precipitation and cloud water chemistry observations on the Czech Krusne Hory plateau

n a

adjacent to a heavily industrialized valley. Atmospheric Environment 36, 353-360. Brimblecombe, P., 2001. Acid rain 2000±1000. Water, Air, and Soil Pollution 130, 25-30.

m

Bytnerowicz, A., Badea, O., Popescu, F., Musselman, R., Tanase, M., Barbu, I., Fraczek, W.,

d e t p

Gembasu, N, Surdu, A., Danescu, F., Postelnicu, D., Cenusa, R., Vasile, C., 2005. Air pollution, precipitation chemistry and forest health in the Retezat Mountains, southern

e c

Carpathians, Romania. Environmental Pollution 137, 546-567. Draxler, R.R. and Rolph, G.D., 2003. HYSPLIT (Hybrid Single-Particle Lagrangian

c A

Integrated

Trajectory)

Model

access

via

NOAA

ARL

READY

Website

(http://www.arl.noaa.gov/ ready/hysplit4.html), NOAA Air Resources Laboratory, Silver Spring, MD. Economou, C. and Mihalopoulos, N., 2002. Formaldehyde in the rainwater in the eastern Mediterranean: occurrence, deposition and contribution to organic carbon budget. Atmospheric Environment 36, 1337-1347. Glavas, S. and Moschonas, N., 2002. Origin and observed acidic-alkaline rains in a wet-only

17

precipitation study in a Mediterranean costal site, Patras, Greece. Atmospheric Environment 36, 3089-3099. Hlawiczka, S., Dyduch, B., Fudala, J., 2003. Long-term changes of particulate emission in the industrial region of upper Silesia (Poland) and their effect on the acidity of rainwater. Water, Air, and Soil Pollution 142, 151-163. Hunova, I., Santroch, J., Ostatnicka, J., 2004. Ambient air quality and deposition trends at rural stations in the Czech Republic during 1993–2001. Atmospheric Environment 38,

t p

887– 898.

i r c

Loye-Pilot, M.D., Martin, J.M., Morelli, J., 1986. Influence of Saharan dust on the rain acidity and atmospheric input to the Mediterranean. Nature 321, 427–428.

s u

Moschonas, N. and Glavas, S., 2002. Weak organic acidity in a wet-only precipitation study

n a

at a Mediterranean coastal site, Patras, Greece. Atmospheric Research 63, 147-157. Polkowska, Z., Astel, A., Walna, B., Malek, S., Medrzycka, K., Gorecki, T., Siepak, J.,

m

Namiesnik, J., 2005. Chemomeytric analysis of rainwater and through fall at several sites

d e t p

in Poland. Atmospheric Environment 39, 837-855. Pun, B.K., Seigneur, C., Grosjean, D., Saxena, P., 2000. Gas-Phase Formation of Water-

e c

Soluble Organic Compounds in the Atmosphere: A Retro-synthetic Analysis. Journal of Atmospheric Chemistry 35, 199-223.

c A

Rocha, F.R., Fracassi da Silva, J.A., Lago, C.L., Fornaro, A., Gutz, I.G.R., 2003. Wet deposition and related atmospheric chemistry in the Sao Paulo Metropolis, Brazil: Part 1. Major inorganic ions in rainwater as evaluated by capillary electrophoresis with contactless conductivity detection. Atmospheric Environment 37, 105-115. Rogora, M., Mosello, R., Marchetto, A., 2004. Long-term trends in the chemistry of atmospheric deposition in northwestern Italy: the role of increasing Saharan dust deposition. Tellus 56, 426-434.

18

Seto, S., Nakamura, A., Noguchi, I., Ohizumi, T., Fukuzaki, N., Toyama, S., Maeda, M., Hayashi, K., Hara, H., 2002. Annual and seasonal trends in chemical composition of precipitation in Japan during 1989-1998. Atmospheric Environment 36, 3505-3517. Sickles II, J.E. and Grimm, J.W., 2003. Wet deposition from clouds and precipitation in three high-elevation regions of the Eastern United States. Atmospheric Environment 37, 277288. Tang, A., Zhuang, G., Wang., Y., Yuan, H., Sun, Y., 2005. The chemistry of precipitation and

t p

its relation to aerosol in Beijing. Atmospheric Environment 39, 3397-3406.

i r c

Tu, J., Wang, H., Zhang, Z, Jin, X., Li, W., 2005. Trends in chemical composition of precipitation in Nanjing, China, during 1992-2003. Atmospheric Research 73, 283-298.

s u

Tuncer, B., Bayar, B., Yesilyurt, C., Tuncel, G., 2001. Ionic composition of precipitation at

n a

the Central Anatolia (Turkey). Atmospheric Environment 35, 5989-6002. Turner, S.M., Malin, G., Liss, P.S., Harbour, D.S., Holligan, P.M., 1988. The seasonal

m

variation of dimethyl sulfide and dimethylsulfoniopropionate concentrations in nearshore

d e t p

waters. Limnology and Oceanography 33, 364-375. WMO Report 160, 2004. Manual for the GAW precipitation chemistry programme.

e c

Guidelines, data quality objectives and standard operating procedures, edited by Allan, M.A., downloaded from http://stinet.dtie.mil/oai/ in January 2007.

c A

19

N

9

12

12

12

24

36

45

Year

2003

2004

2005

2006

2004-2005

2004-2006

2003-2006

500

545

350

53 (48)

44 (39)

44 (40)

420

VWM

VWM

VWM

5.92

5.85

5.78

5.24 7.32 6.38 0.56 6.14 4.31 7.28 6.33 0.66 5.94 4.52 7.42 6.06 0.67 5.61 4.26 7.37 6.18 0.58 6.00

c A 0.05 5.75 0.97 1.52 1.12 0.05 48.98 2.43 7.93 1.77 0.04 30.20 2.90 6.06 2.35 0.04 55.21 2.49 8.84 1.66

NO35.06 261.64 46.69 47.67 36.12 7.61 188.71 56.90 43.68 43.35 3.50 195.08 45.32 42.35 35.02 3.00 255.54 84.93 66.29 79.46*

Cl4.20 299.94 67.63 65.00 46.08 2.02 240.64 58.55 54.65 35.01 2.77 238.35 42.89 48.50 25.74 1.82 113.43 40.90 26.07 42.84

1.73

1.93

2.06

37.42

34.53

30.38

e c

465

min max average V VWM min max average V VWM min max average V VWM min max average V VWM

H+

d e t p

nss-SO4215.60 215.67 74.13 56.01 72.34 15.31 266.11 82.01 59.06 74.58 11.52 209.12 72.13 50.59 61.09 0.63 261.21 77.13 62.64 74.79

SO4218.43 236.25 83.18 59.77 79.10 16.06 272.67 89.52 62.59 78.96 12.71 233.40 77.45 53.61 63.96 3.20 273.39 82.38 63.78 79.83

48.49

39.19

39.19

75.46

74.25

71.46

m

523

287

37 (33)

pH

70.70

70.15

67.84

n a

97 (88) 151 (127) 178 (160)

h (mm)

n

Inorganic anions and cations

0.01 0.16 0.06 0.04 0.05 0.01 0.32 0.08 0.06 0.06 0.03 0.17 0.08 0.04 0.07 0.07 0.30 0.10 0.05 0.07

2.51 268.85 30.27 38.07 25.71 2.78 171.94 18.20 22.99 15.27 1.83 165.00 14.87 28.33 14.18 1.53 214.11 18.08 27.64 13.22 14.73

0.07

Br-

PO43-

17.10

14.22

0.06

0.07

13.74 15.38 15.31

6.26 319.02 60.74 58.73 53.86 7.37 210.03 60.56 55.25 43.17 1.26 191.73 58.57 47.72 48.51 6.00 255.02 80.84 58.28 63.69 45.84 51.79 52.31

13.86 252.15 72.43 60.03 54.08 5.97 284.65 63.30 66.58 35.02 1.67 210.30 42.52 50.53 22.92 3.34 100.24 41.97 27.18 40.28

38.08

32.74

28.97

1.20 101.00 22.87 15.89 15.08 1.23 102.98 23.56 24.22 14.62 1.12 78.31 17.64 14.99 12.86 0.72 125.72 21.62 15.45 18.66

NH4+

Na+

K+

16.74

16.14

12.48

4.96 93.65 22.78 21.77 18.53 2.35 63.15 17.90 16.27 12.69 2.25 142.18 14.71 23.43 12.26 0.79 68.17 25.19 19.69 23.46

Mg2+

107.46

109.34

102.66

10.14 583.96 117.53 115.44 101.82 6.10 644.58 147.75 157.31 109.08 1.77 550.98 107.56 118.41 96.23 8.97 480.34 147.28 122.51 122.72

Ca2+

21

105.83

107.94

101.42

9.17 576.88 114.42 113.94 99.49 5.54 636.10 145.03 155.79 107.58 1.09 548.34 105.73 117.56 95.25 8.08 479.82 145.48 121.96 120.98

nss-Ca2+

Table 1: Statistical data and volume weighted mean (VWM) concentrations of the measured ionic species (concentrations in Peq L-1 entirely with the exception of pH). 1a

s u

i r c

t p

9

12

12

12

24

36

45

2003

2004

2005

2006

2004-2005

2004-2006

2003-2006

545

44 (39)

420

VWM

VWM

VWM

32.40

32.86

31.75

c A

465

4.03 270.89 53.84 53.73 31.10 2.44 287.73 66.69 61.08 33.00 0.75 244.24 49.88 46.59 32.50 3.56 233.28 70.90 60.13 35.10

0.01 0.58 0.09 0.12 0.06 0.01 1.97 0.24 0.40 0.18 0.02 1.36 0.27 0.34 0.15 0.02 5.06 0.75 1.05 0.67

0.01 1.55 0.15 0.29 0.05 0.01 1.03 0.18 0.22 0.13 0.02 0.86 0.21 0.20 0.16 0.01 1.08 0.26 0.21 0.17

0.01 3.45 0.56 0.74 0.42 0.01 4.72 0.87 1.28 1.09 0.01 19.12 1.45 2.29 1.14 0.02 30.94 3.43 3.05 2.71

0.04 20.03 6.35 4.74 5.48 0.10 120.55 6.92 9.83 2.53 0.06 15.13 2.66 3.22 1.93 0.21 35.22 6.93 7.54 6.42

0.32 119.13 14.38 21.66 6.07 0.25 81.96 14.28 20.65 9.99 0.16 87.15 17.57 23.14 12.70 3.84 102.15 26.31 20.01 18.69

0.12 1.83 0.26 0.12 0.27 0.12 0.93 0.27 0.18 0.31 0.12 28.65 4.35 7.34 3.56 0.45 51.95 16.15 12.76 12.17

4.08

5.35

1.93

11.86

13.79

11.34

4.09

3.63

2.23

1.34

1.65

1.12

0.13

0.15

0.15

0.27

0.33

0.17

C3H3O3-

CH3SO3-

C3H5O2-

C2O42-

C2H3O2-

Organics HCO2-

e c

523

min max average V VWM min max average V VWM min max average V VWM min max average V VWM

HCO3-

Bicarbonate

m

97 (88) 151 (127) 178 (160)

350

500

53 (48)

44 (40)

287

37 (33)

n

h (mm)

n a

N-number of months (in 2003 the sampling has been initiated in April); n-number of events; h-rainfall height. Concentrations are reported in Peq L-1 units. The numbers given in the parenthesis of the column for n represent the number of valid cases used in the final analysis. * If for November 2006, the VWM concentration for NO3- (appear as an exceptional case) is not taken into account, then the annual VWM for it becomes 58.34 Peq L-1.

N

Year

1b

d e t p s u

i r c

t p

22

Table 2: Volume weighted mean concentrations of the measured species related to other studies in Europe. Parameter h (mm) pH H+ SO42nss-SO42NO3ClHCO2C2H3O2Na+ NH4+ K+ Mg2+ Ca2+ nss-Ca2+

This worka

Romaniab

Polandc

Czechd

Turkeye

Greecef,g

Spainh

2003-2006 420

2000-2002 256

1996-1999 870

1993-2001 -

1993-1998 450

2000-2001 -

1983-2000 901

5.92 1.73 75.46 70.70 48.49 37.42 4.08 11.86 38.08 52.31 15.31 16.74 107.46 105.83

5.35 115.43 7.92 19.38 12.87 38.27 55.92 18.33 51.92 -

5.14 31.04 52.90 38.87 19.56 9.48 17.50 116 -

~35.62

6.13 3.0 56 28 18 16 64 8.4 11 74 -

5.16 7.0 46.1 34.7 19.4 114.3 3.77 4.81 90.2 16.3 6.6 30.4 98.5 94.6

5.27 41.2 21.6 27.9 22.1 22.8 3.7 9.2 52.5 -

~40.32 -

t p

i r c

-1

Concentrations are reported in Peq L . a - site at ~100 m above sea level, plain area, Iasi, Romania. b - Bytnerowicz et al. (2005): site at 800 m in the Retezat Mountain, Romania, about 400 km far of Iasi. c - Polkowska et al. (2005): for 1999 site at 122 m in Stara-Pila, Rumia, precipitation, Poland. d - Hunova et al. (2004): site at about 800 m, Czech. e - Tuncer et al. (2001): site at 1000 m in the Anatolian plateau, Ankara, Turkey. f - Glavas and Moschonas (2002): Mediterranean coastal site, Patras, Greece. g - Moschonas and Glavas (2002): Mediterranean coastal site, Patras, Greece. h - Avila and Roda (2002): site at 700 m in the Montseny Mountain, Spain.

s u

n a

d e t p

m

e c

c A

23

Table 3: Rainwater ratios and enrichment factors for equivalent ratios of selected ionic species vs. Na+ compared with seawater ratios. Seawater ratios Rainwater ratios (this work) Enrichment factors (EF)

Cl-/Na+ 1.16 1.03 0.89

SO42-/Na+ 0.12 3.85 32.08

K+/Na+ 0.02 0.66 31.42

Ca2+/Na+ 0.04 5.10 118.60

Mg2+/Na+ 0.23 0.64 2.78

- the seawater concentrations for the selected ions are from WMO Report 160, 2004. - rainwater ratios and enrichment factors are reported as VWM values.

t p

i r c

s u

n a

d e t p

m

e c

c A

24

-

Cl1.00 0.25 0.50 0.05 0.63 0.41 0.51 0.44 0.60 0.63 0.53 0.72 0.45 0.19 0.61 0.64 1.00 0.12 0.87 0.62 0.66 0.86 0.52 0.44 0.32 0.71 0.89 0.20 0.90 0.85

NO3-

2-

1.00 0.06 -0.07 -0.13 -0.08 -0.05 -0.14 -0.11 -0.09 -0.14 0.65 -0.06 0.07

PO43-

-

1.00 0.63 0.66 0.75 0.56 0.39 0.36 0.61 0.80 0.25 0.86 1.00

HCO3-

c A

-

1.00 0.54 0.14 0.64 0.39 0.33 0.30 0.06 0.03 0.08 0.40 0.77 0.38 0.47 0.65

SO42Py

1.00 0.43 0.22 0.20 0.42 0.37

MS

1.00 0.85 0.45 0.26 0.20 0.52 0.34

Pr

1.00 0.75 0.56 0.64 0.33 0.14 0.47 0.58

Ox

1.00 0.55 0.49 0.33 0.54 0.66 0.10 0.86 0.72 -

m

Py - C3H3O3 .

1.00 0.59 0.62 0.77 0.74 0.54 0.57 0.15 0.67 0.66

Ac

CH3SO3-;

1.00 0.93 0.52 0.53 0.65 0.64 0.43 0.60 0.06 0.55 0.64

Fo

e c

Fo - HCO2 ; Ac - C2H3O2 ; Ox - C2O4 ; Pr - C3H5O2 ; MS -

Cl SO42NO3PO43HCO3Fo Ac Ox Pr MS Py Na+ NH4+ K+ Mg2+ Ca2+

-

Table 4: Correlation between the ions measured in rainwater during this study (r).

d e t p 1.00 0.65 0.10 0.68 0.64

Na+

1.00 0.27 0.75 0.80

NH4+

t p

i r c

s u

n a

Mg2+

1.00 0.83

K+

1.00 0.22 0.25

1.00

Ca2+

25

Table 5: Factor loadings after Varimax rotation normalized applied to the data set of the ion concentrations in the investigated rain events. Factor Variance (%) ClSO42NO3PO43HCO3Fo Ac Ox Pr MS Py Na+ NH4+ K+ Mg2+ Ca2+

F1

F2

F3

F4

29.48

19.80

13.45

10.89

0.50 0.87 0.65 0.34 0.93 0.10 0.06 0.15 -0.04 0.19 0.08 0.32 0.61 0.61 0.73 0.92

-0.01 -0.06 0.38 -0.03 0.11 0.73 0.59 0.45 0.80 0.75 0.80 0.04 0.20 -0.06 0.20 0.10

0.80 0.24 0.18 0.00 0.15 -0.16 0.26 0.45 0.01 -0.04 0.16 0.89 0.10 0.36 0.36 0.17

0.06 0.07 0.25 0.82 0.13 0.05 -0.28 0.54 0.10 0.15 -0.04 0.08 0.43 0.55 0.26 0.11

t p

i r c

s u

Fo - HCO2-; Ac - C2H3O2-; Ox - C2O42-; Pr - C3H5O2-; MS - CH3SO3-; Py - C3H3O3-.

n a

d e t p

m

e c

c A

26

Table 6: Wet deposition fluxes of measured ions. Parameter h (mm) H+ SO42NO3ClHCO2C2H3O2C2O42C3H5O2CH3SO3C3H3O3Na+ NH4+ K+ Mg2+ Ca2+ HCO3-

This worka

Romaniab

Polandc

Czechd

Turkeye

Greecef

Spainh

2004-2006 465

2000-2002 256

1996-1999 835

1993-2001 -

1993-1998 450

2000-2001 -

1983-2000 901

0.9 1657 1130 570 112 378 74 56 7 13 350 433 279 90 1016 932

1418 126 176 76 176 558 56 265

1296 2853 1200 391 322 183 2018 -

~1500* ~1770* -

1.1 1000 680 240 140 420 130 48 560 -

720 420 654 89 79 115 600 -

1782 1209 893 458 370 131 101 949 -

-

-

i r c

t p

Deposition fluxes are reported in mg m-2 y-1; the reported rainfall height is the average of all investigated period. a - Iasi, Romania: Deposition fluxes estimated only for the period 2004-2006 (2003 incomplete year). b - Bytnerowicz et al. (2005): Fluxes estimated with the data from Table 2. c - Polkowska et al. (2005): Fluxes estimated with the data from Table 2. d - Hunova et al. (2004): sites at about 800 m, Czech; *data correspond for the year 2001 in the diagrams. e - Tuncer et al. (2001). f - Glavas and Moschonas (2002): Mediterranean coastal site, Patras, Greece. h - Avila and Roda (2002): site at 700 m in the Montseny Mountain, Spain.

s u

n a

d e t p

m

e c

c A

27

Figure Legends

Figure 1: Sampling location and representative 5 day backward trajectories of the air masses arriving at the monitored site (at altitudes of 1000 m). Included also are example of local assumed air masses. Figure 2: Linear regression between Ȉcations and Ȉidentif. anions for the analysed rainwater samples (a); pH dependence of the difference Ȉcations-Ȉidentif. anions (b); Regression between the difference (Ȉcations-Ȉidentif.

anions)

and 6(Ca2++Mg2+) (c); Linear regression between

Ȉcations and Ȉanions when the contribution of HCO3- to the anionic budget is taken into account (d).

t p

Figure 3: Percentage distribution of the ionic composition in the rainwater at Iasi, Romania. Figure 4: Seasonal variability both of the pH and of the calculated acidity, H+, (a), and

i r c

frequency distribution of the pH during the sampling period (b). The pattern in the seasonal variability is the result of an average of the VWM values over the entire

s u

period (April 2003-December 2006).

Figure 5: Seasonal variability of the (H+)/(SO42-+NO3-) ratio and of the neutralization factors

n a

for various cations. The pattern in seasonal variability is the result of an average of the monthly VWM values over the entire period (April 2003-December 2006).

m

Figure 6: Distribution of the average of the VWM concentration values, over the entire

d e t p

period (April 2003-December 2006), of selected inorganic anions (a), cations (b) and organic species (c). For NO3- the value for November 2006 was not included in the average.

Figure 7: Distribution of nss-Ca2+ vs. 6(nss-SO42-+NO3-) as a function of the air masses

e c

origin.

c A

Figure 8: Contribution of each sector to the deposition fluxes and rainfall height (%).

28

Figure 1 N: 18% air masses 20% rain NE: 12% air masses 6% rain

NW: 24% air masses 10% rain

E: 8% air masses 15% rain

W: 18% air masses 15% rain

t p

SE: 5% air masse 12% rain SW: 10% air masses 12% rain

i r c

S: 5% air masses 11% rain

s u

n a

d e t p

m

e c

c A

29

Figure 2 -1 6cations - 6identif. anions (Peq L )

-1 6cations (Peq L )

1200 (a) m = 1.24 r2 = 0.89

900

600

300

0

0

200

400 600 6identif. anions (Peq L-1)

800

1000

(b) experimental HCO3 theoretically derived

300 200 100 0

-100

3

4

m = 0.44 2 r = 0.74

300 200 100 0 -100

5

6 pH

7

8

0

200 400 600 6(Ca2+ + Mg2+) (Peq L-1)

800

d e t p

(d) m = 1.00 r2 = 0.95

900

600

0

n a

0

i r c

s u

300

300

9

t p

1200 (c) -1 6cations (Peq L )

-1 6cations - 6identif. anionsPeq L )

400

400

600 6anions (Peq L-1)

900

1200

m

e c

c A

30

Figure 3 Organics (4.7 %)

2-

SO4 (16.3 %)

HCO3(7.0 %) H+ (0.4 %) Other ions (3.7 %)

NO3(10.4 %)

NH4+ (11.3 %) Mg2+ (3.6 %)

Cl(8.1 %)

t p

K+ (3.3 %)

i r c

Ca2+ (23.1 %)

Na+ (8.2 %)

s u

n a

d e t p

m

e c

c A

31

Figure 4 6

6.5 pH Frequency distribution (%)

(b) alkaline range

acid range 30

6.0

+

-1

VWM H (Peq L )

H pH

4

2

0

40

+

(a)

5.5

l t Jan Feb Mar Apr May Jun Ju Aug Sep Oc Nov Dec

0.0

20

10

0 4 .0

5 -4.

4 .5

-5 .

0

5.0

-5 .

Month

6

5 .6

0 -6.

6 .0

-6 .

5

6.5

-7 .

0

7 .0

5 -7.

pH

t p

i r c

s u

n a

d e t p

m

e c

c A

32

Figure 5

NFCa

1.5 NFNa

0.06

1.2

NF

0.9

2-

-

H /(SO4 +NO3 )

0.08

H+/(SO42-+NO3-) NFNH4

0.04

+

0.6

0.02

0.00

0.3

t p

0.0

l t r Jan Feb Mar Ap May Jun Ju Aug Sep Oc Nov Dec

i r c

Month

s u

n a

d e t p

m

e c

c A

33

Figure 6 120

(a)

SO4

NO3 -

N-NO3 /S-SO4

100

2-

20

-

-1

40

-

40

60

N-NO3 /S-SO4

0.4

2-

60 80

VWM NO3 (Peq L )

2-

0.6

80

-

-1

VWM SO4 (Peq L )

2-

0.2

20 0

0

-1

60

+

100

40 50 20

5

0

e c

r r Jan Feb Ma Ap May Jun Month

c A

Jul Aug Sep Oct Nov Dec

4

2

0

-1

6

0 0.6

0.4

-

-

10

m

-1

d e t p

15

1

-

-

-1

VWM C2H3O2 (Peq L )

CH3SO3

0 8

-

VWM HCO2 (Peq L )

C2H3O2

20

HCO2

-

s u

n a

0

(c)

t p

i r c

-1

2

80 VWM NH4 (Peq L )

2+

0.0

100

-1

VWM Ca (Peq L )

150

+

Rainfal height (mm day )

2+ NH4 Ca Rainfal height

VWM CH3SO3 (Peq L )

(b)

0.2

0.0

35

Figure 7 600 m =1.29 2 r = 0.86

m =0.78 2 r = 0.83

400

2+

-1

nss-Ca (Peq L )

N-E NE other sectors

m =0.41 2 r = 0.79

200

0

0

100

200 300 400 6(nss-SO42+ + NO3) (Peq L-1)

t p

500

i r c

s u

n a

d e t p

m

e c

c A

36

Figure 8 N 24.0 NW

SO4

N

2-

NE

16.0

NW

8.0

0.0

W

E

SW

0.0

W

SE

SE

PO4

30.0

N

3-

NE

20.0

NW

16.0

S N

NH4

+

d e t p

27.0

NW

NE

18.0 9.0

e c

0.0

c A S

SE

n a

SW

SE

E

s u

0.0

W

E

SW

SW

i r c

NE

8.0

0.0

W

+

Na

24.0

10.0

NW

t p

S

N

W

E

SW

S

NW

NE

16.0

8.0

-

NO3

24.0

W

m

E

SE

S

N

2+

Ca

30.0 NE

20.0 10.0 0.0

E

SW

SE

S

% deposition fluxes

% rain

37