Environ Monit Assess (2010) 167:493–503DOI 10.1007/s10661-009-1066-7

Trace metals solubility in rainwater: evaluationof rainwater quality at a watershed area, Istanbul

Bertan Basak · Omar Alagha

Received: 2 March 2009 / Accepted: 30 June 2009 / Published online: 15 July 2009© Springer Science + Business Media B.V. 2009

Abstract In this study, 79 bulk precipitation sam-ples were collected at two sampling sites nearBüyükçekmece Lake, one of the important drink-ing water sources of Istanbul, for the period ofOctober 2001 to July 2002. The study comprisedthe determination of trace and toxic metals con-centrations in rain water. The concentrations ofthe metals in this study were found to be higherthan those reported by other researchers aroundthe world. The solubility of toxic metals was foundin the order of Cd > Cu > V > Zn > Ni > Pb> Cr. Solubility of metals under acidic conditions(pH < 5.5) was approximately five times higherthan those under neutral conditions with Cd asthe most soluble metal (50% soluble). Statisticalevaluations including seasonal variations, crustalenrichment factors, and correlation matrix werediscussed to identify the possible sources of thesepollutants. The study revealed that anthropogenicelements were highly enriched especially for Cd> Cu > Pb which were found to be highly en-riched. Significant portion of Cu and Pb couldbe increased by the effect of local sources likecement industry in the area; however, the rest ofthe investigated trace metals could be brought to

B. Basak · O. Alagha (B)Department of Environmental Engineering,Fatih University, 34500 Büyükçekmece,Istanbul, Turkeye-mail: oalagha@fatih.edu.tr

the sampling site by long-range transport to theBüyükçekmece Lake watershed area.

Keywords Bulk deposition · Air pollution ·Trace metals · Metal solubility · Lake ecosystem

Introduction

The atmosphere is recognized as an importantgeochemical receptor and transport medium fortrace metals emitted from the ground (Churchet al. 1984). Man is primarily responsible for theenrichment of many trace elements now found inthe atmosphere caused mainly by the combustionof fossil fuels, including such additives as leadin gasoline, roasting of ores for refining metals,processing of crustal materials for manufacturingcements, and burning of waste materials (Baubelet al. 1994). Monitoring trace metals in the at-mosphere and their effect on receptor site is im-portant because of their potential adverse effectson terrestrial and aquatic ecosystems. It is alsopossible to estimate the sources of these metalsbecause certain species are emitted from particu-lar source types (Rahn and Lowenthal 1985). Be-cause metals can be dispersed over large distancesfollowing their emission, their concentrations inthe precipitation collected from remote areas arealso monitored routinely by researchers.

494 Environ Monit Assess (2010) 167:493–503

Atmospheric deposition of toxic contaminantsonto surface waters is a significant source of load-ing of these contaminants into these waters. Sub-sequent bioaccumulation of these contaminants infish and shellfish can lead to their morbidity andmortality, as well as being hazardous to humanhealth (U.S. EPA 1993). Atmospheric depositionof metals to the surface waters which are usedfor drinking may increase the exposure to thesetoxic contaminants. In addition, toxic metals maycontaminate vegetables and animal food chains,causing dietary exposure (Svensson et al. 1987).Relationship between the concentrations of met-als in surface waters and inputs of trace elementsto these water bodies were investigated by manyresearchers (Gélinas and Schmit 1998; Lawlor andTipping 2003; Ozcan et al. 2007).

The Büyükçekmece Lake, which is located30 km away from Istanbul, is one of the importantdrinking water sources of the city. Major elementsof the precipitation over the lake, relationshipbetween the acidifying species and basic species,and sources of the chemical components in theprecipitation were published before (Basak andAlagha 2004).

In this study, trace metal composition ofrainwater (Cu, Zn, Pb, Cr, Ni, V, and Cd) were in-vestigated, and seasonal variations in the concen-trations of these toxic metals and their potentialsources have been discussed. Soluble, insolublefractions were calculated and factors affecting thesolubility of trace metals were also discussed.

Experimental

Site description

Istanbul, located in the northern west of Turkey, isthe most crowded city with a population of about10 million. It is also a world-renowned touristdestination with unique historic and aesthetic siteslying on two continents. Büyükçekmece Lake islocated in a suburban area, which is located 30 kmaway from the city center. Two rainwater sam-pling stations were selected near BüyükçekmeceLake away from the direct influence of local an-thropogenic pollution sources. One of the stations(station 1) was situated in the Büyükçekmece Wa-

ter Treatment Plant of ISKI (Water and SewerageAdministrative Centre of Istanbul) (41◦02′36′′N,28◦35′26′′E), while the other one (station 2) wassituated at Fatih University Campus (41◦05′23′′N,28◦37′08′′E). The stations were about 5 and 10 kmaway from the Marmara Sea shore, respectively,and approximately at sea level. The samplers wereplaced at 1 m above the ground to prevent bounc-ing and deposition of dust particles from the topsoil.

Sampling and analysis

Bulk rain samples were collected manually on adaily event basis to reduce the amount of dustentering the sampler to the minimum. The rainsampler used was a homemade one, simple, inex-pensive, and easy to use which consists of a highdensity polyethylene (HDPE) funnel of 50 cmdiameter, mounted inside a PP (polypropylene)bucket containing a 5-l HDPE bottle to collectrain samples. The bottle was supported by Sty-rofoam to protect it from spillage. After the rainevent finished, the 5-l collection bottle inside thesamplers was replaced by a new pre-washed one.The funnel was rinsed three times with de-ionizedwater before being positioned again. During dryperiods, samples were collected weekly by wash-ing the sampler funnel with 100 ml of de-ionizedwater. Sampling bottles were closed and broughtto the air pollution control laboratory at the envi-ronmental engineering department of Fatih Uni-versity. The bottles were opened and the sampleswere processed in a laminar flow cabinet, whichprovides a particle-free environment, in order tominimize contamination and obtain low blank val-ues for better detection limits.

De-ionized water was obtained from doubledistilled water that passed through a Milliporewater deionization system model 2010. The pre-pared de-ionized water has a very good quality(resistance 18 mV).

pH were measured in unfiltered rain samplesright after the sample collection. Volume of therain was measured by comparing the sample inthe storage bottle with a pre-calibrated one tominimize the contamination. The samples werefiltered through 0.2-μm-pore-size cellulose acetatefilter paper using a Sartorius polystyrene filtration

Environ Monit Assess (2010) 167:493–503 495

apparatus in order to separate the insoluble frac-tion of rainwater. For pH measurements, a WTWInolab level 1 pH meter, a WTW pH electrodeSenTix 41, and WTW buffer solutions were used.The measurement was carried out on a separateportion (about 15 ml), which was discarded aftermeasurement. About 250 ml of the sample wasstored in polypropylene bottles at 4◦C until analy-sis. The rest of the rain sample was discarded.

The insoluble fraction of the rain on the filterwas digested by using an Ethos 900 MilestoneMicrowave digestion system and a mixture ofcommercial Suprapur grade (Merck) hydrofluoricand nitric acid. Both the soluble and insolublefractions of the deposition were analyzed for Cuand Zn using a Leeman Lab model DRE, ICP-AES. A Perkin Elmer 1100B Model, along withHGA 700 electrothermal atomizer unit, graphitefurnace atomic absorption spectrometer was usedto analyze the soluble and insoluble fractions ofPb, Ni, Cr, V, and Cd.

Filed and acid blank samples were collectedby washing the collecting funnel by 100 ml de-ionized water and analyzed in the same way as thecollected samples. This step is necessary to be surethat contamination during the sampling procedureand treatment of the samples was not significantduring both acid digestion and sampling. Resultsfrom the blank elemental concentration were usedfor blank correction. The detection limit of themethods is taken to be three times the stan-dard deviation of the blank results. Two differentstandard reference materials (SRM), GSP-2 (rocksample) and 1633b (coal fly ash), were used fordata quality assurance of the graphite furnaceAAS and the ICP-AES method. The SRMs weretreated and analyzed in the same way as the insol-uble fraction of rain.

Statistical evaluation

The collected data was statistically treated to putthem in a comprehensive manner and be easier todiscuss. For this goal, SPSS (Statistical package forSocial Sciences, version 9.0) software was used tocalculate general statistics and correlation matrix.In addition to that, enrichment factor (EF) wascalculated to give information regarding the de-gree of anthropogenic input to the observed con-

centrations. EF gives information on the presenceof non-crustal source contributions to observedlevels of elements in rainwater. The crustal en-richment factor (EFc) of an element is calculatedusing the following relation:

EFc = (Cx/CAl)rain sample /(Cx/CAl)soil ,

where (Cx)rain sample is the concentration of theelement in rainwater sample, while (Cx)soil is theconcentration of the same element in the soil. CAl

is the concentration of reference element (Al) inthe same sample and soil (Guerzoni et al. 1999;Nimmo and Fones 1997). Al, Ca, and Fe are con-sidered as typical lithophilic elements. That is tosay, soil is the major natural source for lithophilicelements, and their high concentration in soil can-not be changed easily. Consequently, any of theseelements can be used as a reference element. Alis commonly used as the reference element inthe literature, because it is the most abundantelement in the earth’s crust. Accordingly it is usedas the reference element in this study. Chemicalcomposition of the soil is obtained from Mason’sTable (Mason 1966). An EFc value close to unityindicates the only source of that element to besoil. EFc values much higher than 1 indicates theinfluence of sources other than the earth’s crust.

Results and discussion

Comparison of the two sampling stations

Precipitation samples were collected at two sam-pling stations in order to figure out if there areany local pollution sources impacting the air qual-ity of the study site as well as to minimize anypossible data lost. No significant difference (p <

0.01 at the 95% confidence level) between theconcentrations of the species sampled in two dif-ferent stations was observed. Accordingly, dataobtained from two sampling stations were treatedtogether. Comparison of the concentrations of thespecies collected at two sampling stations nearBüyükçekmece Lake is given in Fig. 1. Volume-weighted average concentrations of the specieswere given in micrograms per liter. Values weregiven on a logarithmic scale.

496 Environ Monit Assess (2010) 167:493–503

Fig. 1 Comparison ofmetals concentrations ofthe two sampling sites

0.1

1

10

100

1000

Pb Ni Cr V Cd Zn Cu

Lo

gar

ith

mic

sca

le o

f m

etal

co

nce

ntr

atio

n (

µg/L

) Plant

University

Metal concentration in precipitation

As a quality assurance measure, comparison ofthe certified values of trace elements with themeasured values is depicted in Table 1 togetherwith detection limits of the analyzed metals. Whileexamining these values, one should note that, dur-ing the whole procedure of collection, treatmentand analysis of rainwater, and insoluble fraction,no significant contamination was observed.

Volume-weighted average concentrations ofthe species were calculated and are depicted inTable 2 beside the arithmetic mean, geometricmean, standard deviation, and minimum and max-imum values expressed as micrograms per liter.As seen from Table 2, there are significant dif-ferences between arithmetic mean and geometricmean values with high standard deviations formost of the measured parameters. This is reportedfrequently by researchers such as Plaisance et al.(1996) and Hernandez et al. (1996). Fluctuations

of measured concentrations in a limited periodcan cause high standard deviations. These highstandard deviations generally indicate log-normaldistribution of the measured species.

The volume-weighted average concentrationsof metals can be ordered in descending order asfollows: Zn > Cu >> Pb > Cr > Ni > V >> Cd.The abundance of the metals may be consideredin three groups. Zn and Cu were the two mostabundant metals. Pb, Cr, Ni, and V fell into thegroup of moderately abundant metals and Cd wasthe less abundant one when compared to others.Zn was generally reported as the most abundantmetal in the precipitation (Balls 1989; Al-Momani2003; Lawlor and Tipping 2003; Avila andRodrigo 2004). However, Cu is rarely reportedas the second one (Al-Momani et al. 1997). Highvalues for Cu in bulk precipitation in our site areprobably related to local industrial emissions, par-ticularly the cement plant in that area. Althoughthe order of abundances depends on the sources

Table 1 Detection limitsand comparison of thecertified values of metalsfor two standardreference materials

Metals Detection limits Found values Certified values Recovery %

GSP-2 1633b GSP-2 1633b GSP-2 1633b

Cu 11.51 45.75 43.00 106.39Zn 11.48 126.74 120.00 105.61Pb 0.39 35.57 59.88 42.00 68.20 84.70 87.80Ni 0.37 137.88 120.60 114.33Cr 0.05 21.02 168.41 20.00 198.20 105.11 84.97V 0.25 52.50 282.50 52.00 295.70 100.96 95.54Cd 0.20 0.70 0.78 89.82

Environ Monit Assess (2010) 167:493–503 497

Table 2 Summary statistics for precipitation samples inBüyükçekmece

AMW AM GM SD MIN MAX

Cu 118.24 151.12 136.22 74.55 77.23 370.44Zn 182.52 303.51 173.99 312.42 34.11 1,073.90Ni 25.68 66.34 20.12 139.09 0.76 986.88Cr 27.31 66.92 22.05 85.82 0.26 342.49Pb 30.83 52.52 28.05 47.39 1.01 184.37V 25.58 53.41 17.65 77.93 0.30 492.63Cd 0.92 1.70 1.00 1.62 0.22 6.37

AMW arithmetic-weighted mean, AM arithmetic mean,GM geometric mean, STD standard deviation, MIN min-imum, MAX maximum

contributing the precipitation, a similar order ofabundances can be observed in different samplingsites around the world. Cd was generally reportedas the less abundant metal in the literature (Migonand Caccia 1993; Grömping et al. 1997; Rodrigoet al. 1999).

Comparison of the species with the datafrom literature

Observed concentrations of the measured para-meters were compared with the data from Turkey,Middle East, Europe, and worldwide and resultsare given in Fig. 2. Except Cd, all of the tracemetal concentrations measured in Büyükçekmecewere found to be higher compared with those in

the literature. The Cu concentration was foundto be approximately 20 times higher when com-pared with those in the other stations exceptAmasra/Turkey. Zn concentration was found tobe five times higher than the worldwide Znconcentration. Although higher Zn values wereobserved previously in Antalya and Amasra sta-tions in Turkey, the highest Zn concentrationwas observed in Büyükçekmece. The highestNi, Cr, Pb, and V concentrations were alsoobserved in Büyükçekmece among the stationstaken into account for comparison. However, dif-ferences between Ni, Cr, Pb, and V concentrationsobserved in Büyükçekmece and those in differ-ent stations were closer. Although Cd concen-trations observed in Büyükçekmece was higherthan worldwide concentration, it was significantlylesser than those observed in other stations inTurkey. This comparison indicates significant aer-ial input of these pollutants to the watershed area.

Seasonal variation of the metals

The ratios of average concentrations of the speciesin summer to the corresponding values in winterare depicted in Fig. 3. Most of the elements havehigher concentrations in the summer. Scavengingof pollutants from air by rainwater is an impor-tant process affecting seasonality of the elemental

Fig. 2 Comparison ofconcentration of metalsmeasured inBüyükçekmeceprecipitation withliterature data

498 Environ Monit Assess (2010) 167:493–503

Fig. 3 Seasonal variations in concentration of metals inwinter and summer

concentrations. The study area is receiving highamount of precipitation between June and De-cember (Basak and Alagha 2004). Because thereis more extensive scavenging in winter, the higherconcentrations of any species measured in winterindicates that they are emitted to the atmospherefrom sources close to the sampling site. Averagesummer concentrations of the metals, except Cu,were observed to be more than two times higherthan their winter concentrations. Average winterconcentration of Cu is significantly lower than theaverage concentration of this metal during sum-mertime. This indicates the effect of local sourcesthat also contribute to observed Cu concentra-tions in the study area rainwater.

Solubility of the metals

The solubility was calculated by comparing themeasured concentrations of each species in rain-water (filtrate or soluble part) and particles inrain (residue on filter or insoluble part). The cal-culation was based on volume-weighted averageconcentrations, and average solubility of differentmetals in rainwater is given in Fig. 4. The solubilityof the elements in deposition follows the order:Cd > Cu > V > Zn > Ni > Pb > Cr. Kaya andTuncel (1997) and Morselli et al. (2003) have alsoreported a similar trend of Pb and Cr solubility asthe least soluble metal in the precipitation. On the

Fig. 4 Solubility of the metals in dry and bulk depositionsamples

other hand, Cd was reported as the most solublemetal in the wet deposition by Kaya and Tuncel(1997) and in the dry deposition by Morselli et al.(2003). Average solubility of the metals in depo-sition was within 0.2–10.6% of the total metalsin deposition. Generally higher solubility valueswere reported in the literature when compared toBüyükçekmece (Morselli et al. 2003; Elagha et al.2001; Çizmecioglu and Müezzinoglu 2008). Thesolubility of element is a complex process, whichdepends on the pH of the rainwater as well astype of particles. If H+ ion concentration in rain-water is not enough to dissolve all the elements,

Fig. 5 Solubility of the metals as a function of pH

Environ Monit Assess (2010) 167:493–503 499

increasing concentrations of the elements may re-sult in decreasing solubility ratios. Another para-meter affecting the solubility of metals is their ownconcentrations.

Relationship between the solubility of the met-als and pH of the precipitation is depicted inFig. 5. Significant differences observed betweenthe solubility of the metals in precipitations having

Fig. 6 Solubility of the metals as a function of elemental concentration

500 Environ Monit Assess (2010) 167:493–503

pH values lower than 5.5 and those having pHvalues equal to or higher than 5.5. The smallestdifference was observed for Cu. Solubility of Cuin acidic rainwater samples was only two timeshigher than that of precipitation having higherpH. Probably Cu concentrations were so high andeven higher H+ ion concentrations in rainwaterwas not enough to dissolve more Cu. The greatestdifference was observed for Pb. Solubility of thePb in acidic rainwater was more than 16 times thatin the neutral one. Effect of pH on solubility ofZn, Ni, Cr, V, and Cd was similar. Solubility ofthese metals under acidic conditions was approxi-mately five times higher than those under neutralconditions. The most soluble metal was Cd andmore than 50% of it was observed to be solubleunder acidic conditions.

Plots of solubility versus concentration ofthe metals are given in Fig. 6. A similar pat-tern observed for all metals indicates that thereis a converse relationship between the concen-trations of the metals and solubility of them inBüyükçekmece precipitation. The knowledge onsoluble fraction of the trace metals in rainwaterprovides information on the impact of rainwaterto the receptor ecosystem. Organisms can uptakethese toxic species when they are in soluble form.However, high metal concentrations are also dan-gerous for the ecosystem even if they are in insol-uble form. Because Fig. 6 indicates that, possiblythey are going to have a chance to be solved whenthey reach the water body.

Correlation between the metals

The correlation matrix is a useful technique todetermine the relationship between the speciespresent in rainwater. In a correlation matrix, thep value is used to test the statistical significanceof the estimated correlation. Values of p below0.05 indicate statistically significant non-zero cor-relations at the 95% confidence level. Accord-ingly, the correlation matrix has been calculatedby using SPSS. Correlation between the totaldeposition metals were high and all the metalscorrelated to each other. A similar situation wasalso reported by Morselli et al. (2003). All thestudied trace metals are known to have anthro-

pogenic sources. In addition to that, they maybe emitted from the same industrial facilities andreach the sampling area. Kaya and Tuncel (1997)concluded that elements that are emitted fromdifferent sources may show correlations due tosimilar physical properties of the particles thatcarry them.

A correlation matrix of soluble fraction of met-als is depicted in Table 3. Cu and Pb are sig-nificantly correlated in the soluble fraction. It iswell known that the major source of Pb in theatmosphere is motor vehicles. It was previouslysuggested that correlation between Pb and Cumay be ascribed for by the road traffic (Monaciet al. 2000; Faus-Kessler et al. 2000). Maybe somepart of the Cu observed in Büyükçekmece wascaused by road traffic. Strong correlations werealso observed between Zn, Ni, Cr, V, and Cd.Maybe they show similar behavior under certainphysical conditions such as wind speed, air tem-perature, and rainfall amount. The presence of acement factory that uses coal as source of energyin the rotary kiln within a 15-km area near thesampling site may be responsible for most of theseobserved elements. It was observed that duringLodos rain storm (southern wind) these metalshave higher concentrations than Poyraz (northernwinds).

Enrichment factor

Average enrichment factors for elements inBüyükçekmece rainwater are shown in Fig. 7.Total concentrations (soluble fraction + insolublefraction of rainwater) of elements were used inenrichment factor calculations. The group of el-ements Ni, Cr, and V have EFc values less than

Table 3 Correlation matrix of soluble deposition metals inBüyükçekmece

Cu Zn Pb Ni Cr V

ZnPbNi 0.65Cr 0.33 0.30 0.47V 0.65 0.73 0.46Cd 0.61 0.55 0.36

Environ Monit Assess (2010) 167:493–503 501

Fig. 7 Crustal enrichment factors (EFc) of rainwaterspecies

10; accordingly, they are considered to be poorlyenriched in the Büyükçekmece rainwater. Zn, Pb,and Cu can be considered as the moderately en-riched group which have EFc values between 50and 70. Cd was found to be highly enriched inBüyükçekmece rainwater, of which EFc valuesexceed 100. The enrichment factors of the metalsin Büyükçekmece rainwater follow the order: Cd> Cu > Pb ∼ Zn >> Ni > Cr > V. A similarorder of enrichment was reported for both thecoastal and open western Atlantic sites (Churchet al. 1984), and for also other remote areas of theworld (Mackenzie et al. 1979). Church et al. (1984)indicated that the same order of enrichment sug-gests common sources and causes for global en-richment of trace metals in remote aerosol andprecipitation. Enrichment of Cu was reported tobe less than Pb and Zn by Church et al. (1984).Higher enrichment of Cu in Büyükçekmece pre-cipitation when compared with global similar sitesmay also indicate that the local sources for Cu areaffecting air quality over Büyükçekmece.

Deposition fluxes to the lake watershed

Although numbers of dry deposition samples werevery limited in this study, the ratio of dry depo-sition fluxes to the total deposition fluxes ranged

between 0.13 for Cu and 0.46 for Ni, which in-dicates that dry deposition is the minor pathwayfor removal of heavy metals from atmosphere atthe study area. Rainfall occurs approximately 30–40 days a year while dry deposition is a continuousprocess. Generally, crustal elements, which occuras large particles, tend to be removed with dry de-position, while anthropogenic ones, which occuras fine particles or gases, tend to be removed aswet deposition.

Surface area of Büyükçekmece Lake, which isone of the important water reservoirs of Istanbul,is about 36 km2, and total drainage area of the lakeis 620 km2. To obtain the total amount of metalsdeposited to Büyükçekmece Lake, mean averagefluxes were multiplied by the surface area of thelake, and results are depicted in Table 4. Besidesthe lake being an important source of drinkingwater in Istanbul, the streams feeding the lake areused as recreational area and for fishing.

Chemical composition of BüyükçekmeceLake is being observed periodically by IstanbulMetropolitan Municipality. Up to the presentstudy, no significant increase has been recordedfor heavy metal concentrations of BüyükçekmeceLake.

Conclusion

The current study reveals a significant data baseregarding the trace metal concentrations in theprecipitation in the vicinity of the most importantwatersheds for Istanbul city. Rainwater sampleswere collected at two sampling sites; one is lo-cated at the fringe of the Büyükçekmece Lake andthe other site is located at the campus of FatihUniversity. The collected samples were analyzedfor trace metals Cu, Zn, Pb, Ni, Cr, V, and Cdby GFAAS technique after necessary filtrationto differentiate between soluble and insolublefractions. The study revealed that there were no

Table 4 Deposition fluxes to the Büyükçekmece Watershed

Deposition flux Zn Cu Mn Pb Ni Cr V Cd

mg m−2 year−1 317.3 252.2 248.8 76.93 78.28 80.88 54.65 2.11kg year−1 (36 km2) 11,424.7 9,079.2 8,958.4 2,769.7 2,818.2 2,911.8 1,967.7 76.16kg year−1 (620 km2) 196,758.9 156,364.2 154,283.8 47,701.1 48,536.9 50,148.2 33,888.7 1,311.7

502 Environ Monit Assess (2010) 167:493–503

significant differences observed between the con-centrations of the metals collected at two samplingsites which suggests that any serious contamina-tion has not occurred during the collection andanalysis of the samples. The volume-weighted av-erage concentrations of trace metals observed inBüyükçekmece rainwater followed the order: Zn> Cu >> Pb > Cr > Ni > V >> Cd. Concentra-tions of the metals in Büyükçekmece precipitationwere generally found to be higher than those ob-served in other sampling sites in Turkey, MiddleEast, Europe, and worldwide. The greatest dif-ference was observed for Cu which showed thehighest concentration. These trace metals prob-ably have significant local and regional sourcesbecause, with the exception of Cu, concentrationsof the metals observed in summer were morethan two times higher than their winter concen-trations. The average volume-weighted pH valuewas found to be 4.81, which points out that therain is slightly acidic (Basak and Alagha 2004).The solubility of the metals in deposition followsthe order: Cd > Cu ∼ V > Zn > Ni > Pb > Cr.The solubility of the metals in acidic precipitationwas significantly higher than those observed in theneutral one. This shows the effect of acid rainevents on increasing the solubility of toxic metalsin rainwater. The smallest difference in solubil-ity was observed for Cu. A converse relationshipbetween metal concentrations and their solubilitywas revealed by the study. Soluble fractions of Cuand Pb were found to be significantly correlatedonly with each other, while the other metals cor-relate with each other significantly. EFc values ofCd, Cu, Pb, and Zn suggest that these elements arehighly enriched in rain samples. A significant partof Cu and Pb may be caused by local sources inthe vicinity of the sampling area due to emissionsfrom the cement industry; however, the rest ofthese toxic metals are probably emitted from coalcombustion in wintertime or from industrial zonessurrounding the region such as Istanbul, Çorlu,and Gebze. Dispersion of these toxic metals atlong distances and their subsequent accumulationin surface waters may contribute unfavorably tothe global water problem in a long period. Thelocal effect of coal combustion could be observedfrom the volume-weighted average pH values of4.81 (Basak and Alagha 2004). The observed av-

erage pH is considered acidic, especially in winter,where most of the acidic events were observed.This raises the question of the effect of local ac-tivities like coal burning for domestic heating andlong-range transport effects.

The authors suggest the monitoring of rain wa-ter quality in the region since there is an ongoingproject by the municipality of Istanbul to supplythe region with natural gas for domestic heating.Finally, since the region is facing a dryness period,the effect of air quality on the water resourcescould be significant; accordingly, gas phase pollu-tants need to be monitored as well.

References

Al-Momani, I. F. (2003). Trace elements in atmosphericprecipitation at Northern Jordan measured by ICP-MS: Acidity and possible sources. Atmospheric Envi-ronment, 37, 4507–4515.

Al-Momani, I. F., Güllü, G., Ölmez, I., Eler, Ü., Örtel,E., Sirin, G., et al. (1997). Chemical composition ofeastern Mediterranean aerosol and precipitation: In-dications of long-range transport. Pure and AppliedChemistry, 69(1), 41–46.

Avila, A., & Rodrigo, A. (2004). Trace metals fluxes in bulkdeposition, throughfall and stemflow at two evergreenoak stands in NE Spain subject to different exposureto the industrial environment. Atmospheric Environ-ment, 38, 171–180.

Balls, P. W. (1989). Trace metal and major ion composi-tion of precipitation at a North Sea coastal site. At-mospheric Environment, 23, 2751–2759.

Basak, B., & Alagha, O. (2004). The chemical compositionof rainwater over Büyükçekmece Lake, Istanbul. At-mospheric Research, 71, 275–288.

Baubel, R. W., Fox, D. L., Turner, D. B., & Stern,A. C. (1994). Acidic deposition. In Fundamentals ofair pollution (pp. 149–152). San Diego: Academic.

Church, T. M., Tramontano, J. M., Scudlark, J. R., Jickells,T. D., Tokos, J. J., Jr., & Knap, A. H. (1984). Thewet deposition of trace metals to the western AtlanticOcean at the mid-Atlantic coast and on Bermuda. At-mospheric Environment, 18(12), 2657–2664.

Çizmecioglu, S. C., & Müezzinoglu, A. (2008). Solubilityof deposited airborne heavy metals. Atmospheric Re-search, 89(2008), 396–404.

Elagha, O., Tuncel, G., & Tusun, S. (2001). Air qualityof the Black Sea region: Local and long range trans-ported pollutants. Eurasian Chemtech Journal, 3, 273–279.

Faus-Kessler, T., Dietl, C., Tritschler, J., & Peichl, L.(2000). Correlation patterns of metals in the epiphyticmoss Hypnum cupressiforme in Bavaria. AtmosphericEnvironment, 35(2), 427–439.

Environ Monit Assess (2010) 167:493–503 503

Gélinas, Y., & Schmit, J. P. (1998). Estimation of the bulkatmospheric deposition of major and trace elementsto a rural watershed. Atmospheric Environment, 32,1473–1483.

Grömping, A. H. J., Ostapczuk, P., & Emons, H. (1997).Wet deposition in Germany: Long term trends and thecontribution of heavy metals. Chemosphere, 34, 2227–2236.

Guerzoni, S., Molinaroli, E., Rossini, P., Rapazzo, G.,Quarantotto, G., De Falco, G., et al. (1999). The roleof desert aerosol in metal fluxes in the Mediterraneanarea. Chemosphere, 39/2, 229–246.

Hernandez, E., Gimeno, L., Sanchez, M., Rua, A., &Mendez, R. (1996). Relationship between rain compo-sition in Spain and its sources. Journal of GeophysicalResearch, 101, 23381–23387.

Kaya, G., & Tuncel, G. (1997). Trace element andmajor ion composition of wet and dry deposition inAnkara, Turkey. Atmospheric Environment, 31, 3985–3998.

Lawlor, A. J., & Tipping, E. (2003). Metals in bulk de-position and surface waters at two upland locationsin northern England. Environmental Pollution, 121,153–167.

Mackenzie, F. T., Lantzy, R. J., & Peterson, V. (1979).Global trace metals cycles and predictions. Journal ofmathematical Geology, 11, 99–142.

Mason, B. (1966). Principles of geochemistry (3rd Ed.).New York: Wiley.

Migon, C., & Caccia, J.-L. (1993). Estimation of anthro-pogenic and natural heavy metals in the northwesternMediterranean rainwater and total atmospheric depo-sition. Chemosphere, 27, 2389–2396.

Monaci, F., Moni, F., Lanciotti, E., Grechi, D., &Bargagli, R. (2000). Biomonitoring of airborne metals

in urban environments: New tracers of vehicle emis-sion, in place of lead. Environmental Pollution, 107,321–327.

Morselli, L., Olivieri, P., Brusori, B., Passarini, F. (2003).Soluble and insoluble fractions of heavy metals in wetand dry atmospheric depositions in Bologna, Italy. En-vironmental Pollution, 124(3), 457–469.

Nimmo, M., & Fones, G. (1997). The potential pool ofCo, Ni, Cu, Pb, and Cd, organic complexing ligands incoastal and urban rain waters. Atmospheric Environ-ment, 31/5, 693–702.

Ozcan, H. K., Demir, G., Nemlioglu, S., Sezgin, N.,& Bayat, C. (2007). Heavy metal concentration ofatmospheric ambient deposition dust in Istanbul–Bosphorus Bridge Tollhouses. Journal of Residual Sci-ence and Technology, 4(1), 55–59.

Plaisance, H., Coddeville, P., Roussel, I., & Guillermo, R.(1996). A qualitative determination of the source loca-tions of precipitation constituents in Morvan, France.Environmental Technology, 17, 977–986.

Rahn, K. A., & Lowenthal, D. H. (1985). Pollution aerosolin the Northeast: Northeastern–Midwestern contribu-tions. Science, 228, 275–284.

Rodrigo, A., Avila, A., & Gómez-Bolea, A. (1999). Tracemetal contents in Parmeli caperata (L.) Ach. com-pared to bulk deposition, throughfall and leaf-washfluxes in two holm oak forests in Montseny (NESpain). Atmospheric Environment, 33, 359–367.

Svensson, B. G., Björnham, Å., Schütz, A., Letteval, U.,Nilsson, A., & Skerfving, S. (1987). Acidic depositionand human exposure to toxic metals. The Science ofthe Total Environment, 67, 101–115.

U.S. EPA (1993). Deposition of toxic air pollutants to theGreat Waters. Report to Congress. Washington, D.C.:U.S. Environmental Protection Agency.