Deposition of heavy metals in a Mediterranean climate area

Atmospheric Research 81 (2006) 1 – 16 www.elsevier.com/locate/atmos

Deposition of heavy metals in a Mediterranean climate area Aysen Muezzinoglu a, Sibel Cukurluoglu Cizmecioglu b,* a

Dokuz Eylul University, Department of Environmental Engineering, Kaynaklar Campus, 35160 Buca, Izmir, Turkey b Dokuz Eylul University, Graduate School of Natural and Applied Sciences, 35100 Bornova, Izmir, Turkey Received 23 June 2005; accepted 20 October 2005

Abstract Deposition samples were collected and analyzed for selected heavy metals in Izmir, Turkey on different days in October 2003–June 2004. A water surface sampler was used to collect dry deposition and rainwater samples to represent the wet deposition during the rainy period of the year. Heavy metal concentrations and fluxes of both dry and wet deposition forms were determined. Dry and wet deposition samples were filtered and both filters and filtrates were analyzed for Cr, Cd, Pb, Cu, Ni and Zn. Heavy metals were determined using atomic absorption spectrometry with a graphite furnace except for zinc which was analyzed by using a flame technique. In wet deposition samples the average total heavy metal concentrations for Cr, Cd, Pb, Cu, Zn and Ni were found as 17.2 F 8.6, 3.1 F1.6, 7.0 F 4.1, 19.7 F 25.0, 186.4 F 225.5, 7.4 F 2.6 Ag l 1. These concentrations were generally higher than the values previously measured at different sites around the world. Dry and wet deposition fluxes in total (suspended + dissolved) forms indicate that both dry and wet deposition fluxes were appreciably higher in Izmir than elsewhere. Heavy metal deposition fluxes in dry and wet forms were compared to find the importance of the atmospheric cleansing mechanisms and evaluated by taking into consideration climatic conditions existing in the study area. Wet deposition rates are found to be more significant than the dry deposition rates on a daily basis. However, dry deposition is more important than wet deposition throughout the study period. D 2005 Elsevier B.V. All rights reserved. Keywords: Soluble heavy metals; Atmospheric deposition; Mediterranean climate; Importance of dry deposition; Atmospheric cleansing; Air pollution

* Corresponding author. Tel./fax: +90 232 453 0922. E-mail address: [email protected] (S.C. Cizmecioglu). 0169-8095/$ - see front matter D 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.atmosres.2005.10.004

2

A. Muezzinoglu, S.C. Cizmecioglu / Atmospheric Research 81 (2006) 1–16

1. Introduction Izmir is a 3 million population city located on the Aegean Sea coast of Turkey and is the center of an industrialized area. Urban and intercity transport is important and the traffic is dense. The climate in the area is typically Mediterranean with occasional rains during the winter and hot–dry weather persisting from May to November. The predominant winds are from northerly directions and blow over the densely polluted industrial zones and carry the long-range air pollutants towards the city and the study site. Second to these winds are the southerly winds that blow more frequently during the winter period and bring in warm and humid air from the Aegean Sea (Dincer et al., 2003). The study site is located in the Kaynaklar Campus of the Dokuz Eylul University in a suburban area away from specific local sources of heavy metals. However, as the location is downwind of the city of Izmir and 500 m to a major highway, non-local influences are possible. A location map of the study area is given in Fig. 1. Atmospheric cleansing mechanisms involve deposition of atmospheric pollutants in wet and dry forms. According to some researchers wet deposition is the predominant cleansing mechanism to remove the ecotoxic metals from the air. The relative significance of wet versus dry deposition may change not only based on the efficiencies of the two mechanisms, but also varies with the local availability of precipitation. Groempling et al. (1997) reported more than 90% of the total amount of the metals existed in wet deposition and also concluded that heavy metals were higher in precipitation in urban–industrial areas than at rural measurement sites in reference to previous data in Germany. On the other hand, Grantz et al. (2003) suggested that, although the rates of dry deposition are orders of magnitude slower than that of wet deposition, it is a continuous and dependable process for atmospheric cleansing. Tuncer et al. (2001) indicated that the ion concentrations measured in daily wet-only samples collected at a rural station in Central Anatolia near Ankara, Turkey, during 1993–1998 were the highest in Europe among sites with similar characteristics. Yet the dry deposition as a cleansing mechanism was more important over an annual basis at this semi-arid region because of the low annual precipitation rates. Izmir area has a distinct Mediterranean climate with occasional precipitation during the winter period and rainfall below the evaporation rate on an annual basis. In Mediterranean areas like Izmir with long dry periods, dry deposition is expected to be the main mechanism of atmospheric cleansing throughout the year. The dry deposition flux rates and ambient levels of the metals were determined for bi-weekly intervals for 1 year during 2000–2001 at the same site in a previous study. Dry deposition flux rates were determined by using the greased plate method and a high volume sampler was used to determine the selected metals in suspended particulate matter. Deposition velocities were calculated by using a commonly used model. Both the dry deposition fluxes and the heavy metal concentrations in the ambient air were found to be appreciably higher than in other places in the world (Odabasi et al., 2002). This is a first report for heavy metal deposition in this part of the Mediterranean area. Major ion compositions of wet and dry deposition in the Eastern Mediterranean were studied by AlMomani et al. (1998) who indicated that dry deposition is a more important cleansing mechanism due to the prevalence of marine and crustal ions having larger particle sizes and also in view of the relative scarcity of rain events in the Eastern Mediterranean basin. Although the deposition process cleans the atmosphere, its ultimate result is the transfer of toxic atmospheric pollutants into the water and soil environments. Thus, heavy metals in

A. Muezzinoglu, S.C. Cizmecioglu / Atmospheric Research 81 (2006) 1–16

Fig. 1. Location of the suburban sampling site with the annual wind rose in the area. 3

4

A. Muezzinoglu, S.C. Cizmecioglu / Atmospheric Research 81 (2006) 1–16

atmospheric precipitation may create ecotoxic effects in the receiving water and soil environments. However, they are bioavailable only if the metal is in soluble form. Hydrolysis rates of heavy metal salts depositing from the atmosphere into the surface or soil waters depend on the anion and cation balance, redox potential and pH of these waters, and the size distribution and chemical nature of the depositing particles. The percentage of dissolved metals in the deposited matter depends on the anions that are incorporated and some other characteristics of the solution. Future research involving particular efforts to understand the impacts of these pollutants on soil microbiology has been recommended by Morselli et al. (2004). Water pH is one of the important factors determining the ratio of dissolved or suspended forms of deposited heavy metals. It is therefore important to know about the acid rain occurrences to estimate the possibility of high hydrolysis rates of the metals. Gu¨lsoy et al. (1999) indicated the frequency of acidic pH to be about one-fifth of the total rain incidences in 1996 in Istanbul, Turkey. Akkoyunlu and Tayanc¸ (2003) showed that, although SO42 concentration in precipitation was high, so was the Ca2+ as a neutralizing factor of the acidity in rain. A similar argument was reported by Al-Momani et al. (1998), based on the data indicating that the excessive alkaline material in crustal aerosols in the Eastern Mediterranean region has a strong neutralizing impact on the rainwater. Al-Momani et al. (1995) have also concluded that the neutralizing capacities of aerosols of crustal origin in precipitation are high due to the excessive limestone coverage in the Izmir area. Thus the solubility of heavy metals in rainwater is not enhanced by any acid rain effect in these areas. That is why water pH at the beginning and end of the sampling duration was recorded in order to determine the effect of pH on the solubility of the heavy metals in rainwater and receiving water of the dry deposition sampler in this study. 2. Materials and methods 2.1. Sampling Sampling was carried out on a platform 3 m above the ground in the Dokuz Eylul University Kaynaklar Campus. This location is in a growing forest about 10 km south of the city center of Izmir. An expressway exists about 500 m south of the platform. No specific heavy metal source exists in the area. In addition, no buildings or similar obstructions exist which would interfere with the operations of the meteorological instruments, water surface sampler (WSS) and the rain gauge. Dry and wet deposition samples were collected in the WSS having an open surface. This unit has a circular open surface area continuously refreshed with recirculated water coming up from the center. The water overflows from its carefully designed circular weir all around. Technical drawings and operating principles of the water surface sampler are designated in the literature (Shahin et al., 1999; Odabas¸V et al., 1999; Cakan, 1999; Tasdemir, 1997; Yi et al., 1997). In dry deposition sampling, the water surface plate was used but in wet deposition sampling this plate was removed to collect the rainwater in the holder. The water surface holder has a leading edge to minimize airflow disturbances which may be caused by collector geometry. The stainless steel surface plate has a 37.2 cm diameter and 0.65 cm depth and is placed inside the holder at a height which allows the water on the plate to be at the same level with the top of the water surface holder. The WSS has a water surface renewal system to maintain constant water level and control the surface retention time. The pump that provided the water flow through the system was an adjustable liquid pump (all wetted parts

A. Muezzinoglu, S.C. Cizmecioglu / Atmospheric Research 81 (2006) 1–16

5

covered with Teflon). A 4-l plastic bottle is used as the water reservoir. Double distilled water enters the water surface plate from the center and overflows from the weirs at the sides. The retention time of the water at the water plate surface is maintained at 2–4 min. All wetted components of the WSS and holding tank were washed and rinsed with double distilled water prior to each sampling run. Rainwater samples were collected in a narrow neck plastic bottle connected to the funnel under the WSS. Sampling was done when the WSS was empty. Bottle connection was manually opened as soon as the rain started and closed immediately after the last drops. In nearly all cases, the sampled volume of the rainwater corresponded to the total daily rain volume. And for the few days that it did not, heavy metal quantities found for the rainwater samples were corrected for the total volume of the rain on that day. Wet depositions with the rainfall were expressed in terms of the wet flux for the study day. The total quantity of rain in the study area between October 2003 and June 2004 was 406.8 kg m 2 on 31 rainy days. This is also the number of rainy days for the year 1 October 2003–1 November 2004 as it does not rain during the summer period. Out of these rain events 13 were actually sampled. For studying the wet deposition of the heavy metals, 118.4 kg m 2 of this rainfall which corresponded to nearly 30% of the annual precipitation has been evaluated on these 13 rainy days. Between October 2003 and June 2004 twelve dry deposition samples were taken for 24 h. Water pH was measured at the beginning and at the end of the sampling period during the dry deposition sample collection in the WSS. Rainwater pH was measured and recorded at the end of the sampling. A rain gauge was used to find the intensity and duration of rain. Days and durations of sampling, pH, and meteorological conditions for sampling activities are tabulated in Table 1. 2.2. Sample handling and preparation At the end of dry deposition sampling, all of the water in the WSS holding tank was transferred into a clean plastic bottle with a plastic stopper and carried to the laboratory where its volume was measured. The volumes of these samples were about 0.5–2.3 l depending on the evaporation height of the day. After thorough shaking a 250 ml aliquot was taken from these samples and filtered through a 0.45-Am Sartorius membrane filter. A 100 ml aliquot of the filtrate was acidified to pH 2 by nitric acid (Merck, 65%) and kept refrigerated in tightly covered plastic bottles until the heavy metals were analyzed to find the dissolved metal fraction. Filters were digested as explained below and analyzed for heavy metals to find the suspended fractions of the metals. Another 100 ml aliquot of the fresh unfiltered sample was acidified with nitric acid to pH 2 to be used as the btotalQ sample for heavy metal analysis, too. 2.3. Digestion Following the filtration, membrane filters were placed into clean polyethylene bottles containing 50 ml of 20% nitric acid solution prepared from a stock of high purity acid (Merck, 65%). The polyethylene bottles were placed into a shaker for 24 h operating at 60 8C and 270 rpm to dissolve the metals in the filtered particles in the nitric acid solution. Then the solution was transferred into a clean 250-ml Teflon beaker together with 2–3 rinses of distilled water. The Teflon beaker was placed on a hot plate at a temperature of about 150–180 8C to reduce the volume down to 5–10 ml. An additional 20 ml of nitric acid was added to the beaker, and

6

A. Muezzinoglu, S.C. Cizmecioglu / Atmospheric Research 81 (2006) 1–16

Table 1 Sampling dates, durations and meteorological conditions averaged over the sampling durations of sampling for DD (dry deposition) and WD (wet deposition) sampling Sample

DD-1 DD-2 DD-3 DD-4 DD-5 DD-6 DD-7 DD-8 DD-9 DD-10 DD-11 DD-12 WD-1 WD-2 WD-3 WD-4 WD-5 WD-6 WD-7 WD-8 WD-9 WD-10 WD-11 WD-12 WD-13

Datea

10.06.03–10.07.03 10.22.03–10.23.03 11.04.03–11.05.03 11.20.03–11.21.03 12.05.03–12.06.03 05.20.04–05.21.04 05.21.04–05.22.04 05.22.04–05.23.04 06.09.04–06.10.04 06.10.04–06.11.04 06.11.04–06.12.04 06.12.04–06.13.04 10.08.2003 10.09.2003 10.30.2003 10.31.2003 11.06.2003 12.31.2003 01.02.2004 01.05.2004 01.12.2004 01.22.2004 01.30.2004 01.312004 06.03.2004

Sampling durationb (min) 1339 1339 1339 1339 1339 1339 1339 1339 1245 1330 1330 1330 8 31 567 107 359 135 112 43 327 387 221 101 10

Temperature (8C) Air

Water

Avg. wind speed (m s 1)

23.6 25.2 18.6 13.9 7.2 17.9 19.8 19.2 23.6 23.6 22.7 24.4 21.2 20.2 8.0 13.7 12.9 8.7 8.9 6.5 8.4 7.1 8.1 6.0 20.7

22.4 24.7 17.4 12.5 6.7 17.6 19.5 19.1 23.3 23.5 22.1 24.1 21.1 19.8 7.4 12.9 11.7 7.9 8.5 5.8 8.3 6.7 7.9 5.8 20.3

2.9 5.2 3.0 1.3 2.8 3.0 2.3 3.0 6.2 6.2 3.6 2.1 7.4 13.8 2.5 2.7 5.1 0.3 3.7 3.1 3.1 3.3 4.9 5.1 3.1

e

Wind directionc

Rain intensity (kg m 2 day 1)

Water pHd

SE–SSE SE–SSE N–NNE ESE N–ESE NNW–WNW NW–WNW ESE–SE N N WNW–NNW ESE–WNW SSE SSE SSE SSE N S SSW WNW ENE NNE SE ESE S

– – – – – – – – – – – –

6.48/6.34 6.21/6.25 5.77/5.74 6.79/6.78 5.77/5.56 7.77/7.68 6.81/6.81 6.78/6.77 7.77/7.68 7.75/7.23 6.81/6.81 6.75/6.34 7.53 5.94 6.32 7.04 7.42 6.54 5.53 6.65 7.62 5.14 6.05 5.06 7.72

0.9 5.7 2.2 7.6 15.6 7.8 3.9 2.7 1.6 38.3 26.4 5.1 0.6

a

Beginning/end dates of sampling for DD samples, for WD samples the date given belongs to the day of the rain sampled. b For WD samples total duration of rain on that day. c Direction of the wind of maximum occurrence among all the hourly wind directions for the day. d For DD samples, WSS water pH before/after the sampling, for WD samples, pH of the collected rainwater sample. e Measured in the sampler immediately after the collection.

digestion was continued until 10–15 ml solution was left. After cooling, this solution was filtered through a clean 0.45 Am membrane filter. The filtrates were stored in clean 100-ml polyethylene bottles in a refrigerator until they were analyzed for the undissolved fraction in the environmental samples. Three unused filter blanks from the same batch were prepared in the same way to determine the contamination from the filter extraction procedure. Heavy metal concentrations in suspended form were corrected by subtracting the mean blank filter results from the amounts in the samples. 2.4. Analysis Heavy metal concentrations were measured using a Perkin-Elmer Model 700 atomic absorption spectrophotometer equipped with a graphite furnace. Background correction was

A. Muezzinoglu, S.C. Cizmecioglu / Atmospheric Research 81 (2006) 1–16

7

applied using a deuterium lamp with the two-line method. For Zn analysis, AAS with flame technique was used in the same instrument to avoid erratic results in the graphite furnace due to high concentrations in the samples. 2.5. Quality control and quality assurance Distilled water, mixtures of distilled water added by nitric acid and procedural blanks such as the covered-top WSS and the plastic tank runs were measured together with the samples. Metal concentrations were determined from calibration curves prepared with known concentrations of the heavy metals in the expected concentration ranges. Four levels of calibration standard solutions of each heavy metal within the expected concentration ranges were prepared by diluting the standard solutions with ultra pure water. These calibration standards were freshly prepared from the stock solutions of 1 mg l 1 of the soluble metal salt reagents of high purity. For all metals the linear fits of the calibration curves were with r 2 z 0.99. All samples were analyzed in triplicate. Heavy metal analysis of filters and filtrates were aiming at finding the dissolving and nondissolving fractions in the aqueous phase of the deposited material. Their sum should give the total deposition of the metal. However, to check the reliability of the filtration procedure the unfiltered samples were also analyzed to find the total heavy metals. The deviations between sum of filter and filtrate concentrations and total (unfiltered) analyses of the samples were not significant at the 99% level. For quality assurance, water blanks were analyzed. To test for the metallic interferences that might come from the stainless steel parts of the WSS, a plastic tank with similar opening size and water depth was placed next to the WSS. For several days, water blanks were obtained by simultaneous running of the WSS and the plastic tank with tops in covered position for the sampling duration of about 1 day. The background heavy metal concentrations between the closed WSS water blanks, unused water blanks and the plastic tank contents were not different at the 0.99 level of significance. Thus it was decided that using the WSS for sampling the heavy metals in dry and wet deposition samples was appropriate. All readings have been corrected with the averages obtained from triplicate analyses of at least three blank samples. Limits of detection (LOD) were found as the mean blank plus three standard deviations (Odabas¸V et al., 1999) in mass units. Ratios of mass of heavy metals in the samples to the corresponding LOD were found to vary between 1.7 (Ni in wet deposition samples) and 26.9 (Zn in dry deposition samples). Thus it was assured that all of the results were above the LODs. In order to verify the AAS readings, calibration curves were tested during the analyses by using two different tests and efficiencies were calculated as deviations from the known concentrations. A quality assurance test was applied during the analyses. At the end of each analysis run three mixed test solutions containing known amounts of the stock solution and added acidified distilled water similarly prepared as in the acidification of the samples to pH 2 were analyzed. Results were corrected for the metal concentrations in added acid–water mixture blanks. In this test, the heavy metal concentrations in the standards were chosen so that the test solution at the end is near the third calibration point. As a second test, solutions were prepared by adding known volumes of heavy metal stock solution into 500-ml volumetric flasks added to by one of the samples to make another analytical test for quality assurance. This procedure was repeated using three different samples randomly selected. The difference

8

A. Muezzinoglu, S.C. Cizmecioglu / Atmospheric Research 81 (2006) 1–16

between the measured heavy metal quantity in the sample and the analysis results of the prepared solutions should correspond to the known heavy metal quantity from the stock solution. The deviation gave information about the reliability of test results. Measurement results in both of these tests were compared with the known metal concentrations and the deviations were calculated. Results from the two quality assurance tests described above showed the deviations from the expected concentrations of the analytical results. The average agreements were 93%–97% for Cr, 100% for Cd, 100% for Pb, 93%–101.3% for Cu, 92%–96% for Zn and 97%–114% for Ni. 6

Cr

WD-13 WD-13 WD-13

12 8 4

WD-1

WD-13

WD-12

WD-11

WD-10

WD-9

WD-8

WD-7

WD-6

WD-5

WD-4

WD-3

WD-2

WD-1

WD-12 WD-12 WD-12

Ni

16

0

0

WD-11 WD-11 WD-11

150

WD-9

WD-10 WD-10

15

WD-1

micrograms per liter

300

WD-10

WD-9 WD-9

30

WD-13

WD-12

WD-11

WD-10

WD-9

WD-8

WD-7

WD-6

WD-5

WD-4

WD-3

WD-2

450

WD-8 WD-8 WD-8

45

20

600

WD-7 WD-7 WD-7

60

Zn

750

WD-6

WD-5

WD-6 WD-6

WD-3

WD-2

Cu

75

0 WD-1

micrograms per liter

WD-4

0

WD-5

3

WD-5

6

WD-4

9

WD-4

micrograms per liter

12

WD-3

WD-13

WD-12

WD-11

WD-9

WD-10

WD-8

WD-7

WD-6

WD-5

WD-4

WD-3

WD-2

WD-1

micrograms per liter

90

Pb

15

900

1 0

0

18

2

WD-3

10

3

WD-1

20

4

WD-2

30

Cd

5

WD-2

40

micrograms per liter

micrograms per liter

50

Sample number

Sample number Dissolved

Suspended

Fig. 2. Heavy metal concentrations in rainwater, dissolved and suspended forms shown separately (first standard deviations of the total metals data indicated with error bars).

A. Muezzinoglu, S.C. Cizmecioglu / Atmospheric Research 81 (2006) 1–16

9

2.6. Calculations Dry and wet deposition fluxes were calculated by dividing the amount of trace metal deposited upon the area in square meters of the holding plate during the sampling time. Eq. (1) was used to estimate the deposition fluxes. F¼

Md ð AÞ  ðT Þ

ð1Þ

where F is the trace metal deposition flux (Ag m 2 min 1), M d is the collected trace metal mass in the day of the sampling (Ag), A is the deposition sampler collection area (m2) and T is the sampling duration in minutes on that day for dry deposition sampling. In the case of wet deposition T is the duration of rainwater collection period in minutes. When it was necessary to correct the mass of the metal in the rain sample for the total volume of the rainwater for a given day, M d was calculated by multiplying the mass in the sample volume by a factor found by dividing the total volume of rain during that day by the rainwater sample volume. However, as most of the rainwater sampling duration corresponded to the total rainy period on that day, this correction was unnecessary in most of the samples. 3. Results 3.1. Heavy metal concentrations in rainwater and dry deposition samples Heavy metal concentrations in the rainwater samples are plotted in Fig. 2 showing the distribution of dissolved, suspended and total forms for the 2003–2004 rainy period. The ratio between the heavy metal concentrations in dissolved and suspended forms showed strong fluctuations as may be noted from this figure. Wet deposition is an important transport route for metals when rainy events occur. In this study the ratio of insoluble to soluble fractions is quite high for the first two samples for all heavy metals but for Cd. However, it must be noted that the first two samples are taken during rain incidences following elongated dry periods. This is in parallel with the results of Jaradat et al. (1999) who found that high concentrations were recorded after a dry period and when rain continued for several days, concentrations were lower in rain in Amman, Jordan. Table 2 Average (F r) heavy metal flux rates (Ag m 2 day 1) in dissolved and suspended fractions of the wet depositiona (n = 13) and dry deposition (n = 12) during the October 2003–June 2004 period in Izmir Metal

Cr Cd Pb Cu Zn Ni

Flux in dissolved form (Ag m 2 day 1)

Flux in suspended form (Ag m 2 day 1)

Total fluxb (Ag m 2 day 1) Wet

Wet

Dry

Wet

Dry

1413.9 F 1365.4 326.4 F 500.8 1006.4 F 1083.1 2549.8 F 6560.9 11 681.9 F 11 891.7 1334.3 F 1174.9

82.6 F 52.4 29.2 F 13.8 80.9 F 31.6 74.0 F 50.2 1938.0 F 743.6 138.4 F 63.0

268.5 F 387.0 52.5 F 105.5 38.0 F 124.7 345.0 F 984.4 1339.6 F 2330.5 Nd

13.8 F 27.5 1696.4 F 1505.9 97.4 F 69.3 12.1 F11.7 382.0 F 520.4 41.9 F 20.3 9.4 F 12.9 1098.9 F 1133.9 96.6 F 38.9 7.4 F 5.9 2915.3 F 7575.1 81.9 F 48.5 232.1 F 297.7 13 205.5 F 13 804.1 2182.1 F 610.6 Nd 1430.5 F 1215.8 152.8 F 63.6

Dry

a Total wet deposition flux per day is the rain duration in minutes multiplied by the flux per minutes for any rain event on that day. b Total flux values were not made up by summing the dissolved and suspended forms but found by separate analyses.

10

A. Muezzinoglu, S.C. Cizmecioglu / Atmospheric Research 81 (2006) 1–16

Concentrations in rainwater samples showed correlations between Pb and Cu (r 2 = 0.41; p b 0.01), Pb and Zn (r 2 = 0.59; p b 0.01), Pb and Ni (r 2 = 0.48; p b 0.01), Cu and Zn (r 2 = 0.30; p b 0.05) and Zn and Ni (r 2 = 0.53; p b 0.01). This possible indicates that these elements might have common sources. Statistical analysis for dry deposition samples showed no significant relationship between heavy metal concentrations. The statistically significant relationships were found between 2500

Cr

5000

Flux (ug/m2.day)

WD-13

WD-12

WD-11

WD-9

WD-10

WD-8

WD-7

WD-6

WD-4

WD-5

WD-11

WD-12

WD-13

WD-11

WD-12

WD-13

WD-8 WD-8

WD-10

WD-7

WD-10

WD-6 WD-6

WD-7

WD-9

WD-5 WD-5

Ni

4000 3000 2000 1000 0 WD-9

WD-13

WD-12

WD-11

WD-10

WD-9

WD-8

WD-7

WD-6

WD-5

WD-4

WD-3

WD-2

10000

WD-4

20000

WD-4

30000

WD-3

40000

WD-1

100

WD-3

Flux (ug/m2.day)

50000

0

1000

5000

Zn

Cu

10000

10

WD-13

WD-12

WD-11

WD-10

WD-9

WD-8

WD-7

WD-6

WD-5

WD-4

WD-3

WD-2

1000

WD-3

WD-1

Flux (ug/m2.day)

2000

WD-2

0

WD-13

WD-12

WD-11

WD-9

WD-10

WD-7

WD-6

WD-5

WD-4

3000

60000

Flux (ug/m2.day)

WD-3

4000

0

500

100000

Pb

WD-1

Flux (ug/m2.day)

5000

WD-2

WD-1

0

WD-8

1000

1000

WD-2

2000

1500

WD-1

3000

WD-2

4000

Cd

2000

WD-1

Flux (ug/m2.day)

6000

Sample No

Sample No Dissolved

Suspended

Note: Logarithmic scale was used for Cu. Fig. 3. Wet deposition fluxes (Ag m 2 day 1) of the heavy metals in dissolved and suspended fractions (first standard deviations of the total fluxes are indicated as error bars).

A. Muezzinoglu, S.C. Cizmecioglu / Atmospheric Research 81 (2006) 1–16

11

dissolved concentrations of Pb and Ni (r 2 = 0.60; p b 0.01) and total concentrations of Pb and Ni (r 2 = 0.73; p b 0.01) at the 95.0% confidence level, only. 3.2. Deposition fluxes Average daily wet and dry deposition heavy metal flux rates for the dissolved and suspended fractions are given in Table 2. Daily wet deposition fluxes of different heavy metals in the area are summarized in Fig. 3 indicating the range of wet deposited heavy metal fluxes found at the site for the study period. Fluctuations between individual sampling days can be noted from this figure with the dissolved and suspended heavy metal fractions in wet deposition separately shown. Dry deposition fluxes are summarized in the form of ranges of calculated values of the 300

80

Cr

Cd

160 120 80 40

DD-12

DD-11

DD-9

DD-10

DD-8

15 0 10 0 50

DD-2

DD-3

DD-4

DD-5

DD-6

DD-7

DD-8

DD-9

DD-10

DD-11

DD-12

DD-3

DD-4

DD-5

DD-6

DD-7

DD-8

DD-9

DD-10

DD-11

DD-12

DD-1

Zn

DD-2

DD-12

DD-11

DD-10

DD-9

DD-8

DD-7

DD-6

DD-5

DD-4

DD-3

DD-2

DD-1

300

3500

Ni

2500 2000 1500 1000

200 150 100

DD-12

DD-11

DD-10

DD-9

DD-8

DD-7

DD-6

DD-5

0 DD-4

0 DD-3

50

DD-2

500

DD-1

Flux (ug/m2.day)

250

DD-1

Flux (ug/m2.day)

Cu

20 0

0

0

3000

DD-7

DD-12

DD-11

DD-9

DD-10

DD-8

DD-7

DD-6

DD-5

DD-4

DD-3

25 0

Pb Flux (ug/m2.day)

Flux (ug/m2.day)

200

DD-2

DD-1

0

DD-6

0

20

DD-5

50

40

DD-4

100

DD-3

150

60

DD-2

200

DD-1

Flux (ug/m2.day)

Flux (ug/m2.day)

250

Sample number

Sample number Dissolved 2

1

Suspended

Fig. 4. Dry deposition fluxes (Ag m day ) of the heavy metals in dissolved and suspended fractions (first standard deviations of the total fluxes are indicated as error bars).

12

A. Muezzinoglu, S.C. Cizmecioglu / Atmospheric Research 81 (2006) 1–16

dissolved and suspended fractions of heavy metals in Fig. 4. The standard deviations around the average total metal fluxes are indicated as error bars in Figs. 3 and 4. This study showed that the heavy metal deposition fluxes were rather high with high fluctuations throughout the study period. However, these variations have no characteristic pattern possibly because no nearby source of the heavy metal exists in the area. 4. Discussion of results Neutral pH values were found in the WSS samples measured both before and after daily dry deposition sampling and also in the rainwater samples (Table 1). This indicates that acid rain phenomena did not exist in Izmir during the study period. This is in contrast to the high quantity of SO2 emissions in the area (Dincer et al., 2003), but in parallel with the findings of Al-Momani et al. (1995) who have described the impact of high quantities of limestone in the dust reentrained from the ground in the Izmir area. Average concentrations measured in the rain in this study are separately shown in Fig. 2 for the 13 rainwater samples during the study period. When these concentrations are compared with similar measurements in precipitation elsewhere, Zn and Cd concentrations in the rain in Izmir were higher by two orders and one order of magnitude above those cited by Takeda et al. (2000), respectively. These authors cited Zn concentration of 4.8 Ag l 1 and Cd concentration of 0.063 Ag l 1 on average in the precipitation in Hiroshima, Japan. In Izmir, the average annual precipitation for the October 2003 to June 2004 rainy season was 406.8 mm year 1 which was quite below the precipitation rate of 1433 mm year 1 given by Takeda et al. (2000). Average lead concentration in the rainwater in Izmir of 7.0 F 4.1 Ag l 1 was within the range of 0.02– 25.1 Ag l 1 given by the same authors. Izmir rainwater contained one order of magnitude more Cd, Cu, Zn and about two times more Pb than the rural station studied on the western coast of France (Deboudt et al., 2004). All heavy metal concentrations in rain in Izmir were of the same order of magnitude as in Singapore (Hu and Balasubramanian, 2003) except Cd and Zn which were higher in concentration. On the other hand, similarly high heavy metal concentrations and even higher Zn levels in the rainwater in Amman were reported by Jaradat et al. (1999). Wet deposition rates are much more significant than the dry deposition rates on a daily basis, as can be seen from the last column of Table 2. The daily wet deposition fluxes were larger than the dry deposition fluxes by 17.4 times for Cr, 9.1 times for Cd, 11.4 times for Pb, 35.6 times for Cu, 6.1 times for Zn and 9.4 times for Ni. The data also confirmed that the fluxes were controlled by the amount of precipitation except for Zn as shown by the correlation analyses between the amount of precipitation and heavy metal fluxes. This is in contradiction, except for Zn, with Takeda et al. (2000) who stated that Cr, Zn, Cu, and Ni fluxes do not correlate with the amount of precipitation but in parallel with the Pb results of Kim (1998). Strong relationships were not found between heavy metals in dry deposition samples when total and soluble fractions were separately tested (Table 2). In wet deposition, however, the deposition fluxes of different metals correlate better. For example in soluble form the correlation coefficient was as high as r 2 = 0.95 between Cd and Cu. Such strong relationships between the metals indicate that they probably come from the same emission sources and/or undergo similar reactions in the atmosphere before they were brought down by the raindrops. Table 3 indicates that the daily dry deposition fluxes were appreciably higher in Izmir compared to elsewhere. For example, daily dry depositions found in this study for Izmir were 10–50 times higher for Cd; 4–20 times higher for Pb and 5–30 times higher for Zn than in

A. Muezzinoglu, S.C. Cizmecioglu / Atmospheric Research 81 (2006) 1–16

13

Table 3 Comparison of daily dry deposition flux rates in mg m 2 day 1 cited in the literature and this study Metal Tokyoa Romeb Seoulc Chicagod Chicagoe Kunpof Earlier data at Izmirg

This study at Izmir Total heavy metals

Cr Cd

0.16

Pb

9.3

0.351

0.05

0.12

0.07

0.104

Cu

21

0.507

0.06

0.16

0.06

0.056

Zn

150

1.612

0.11

0.70

0.20

0.185

Ni a b c d e

0.423

Avg.: Range: Avg.: Range: Avg.: Range: Avg.: Range: Avg.: Range: Avg.: Range:

0.02 F 0.02 0.002–0.05 0.02 F 0.03 0.001–0.06 0.22 F 0.21 0.03–0.86 0.12 F 0.09 0.02–0.33 1.91 F 0.86 0.60–3.30 0.13 F 0.10 0.01–0.43

Wet

Dry

1.70 F 1.51 0.11–4.92 0.38 F 0.52 0.02–1.95 1.10 F 1.13 0.10–4.16 2.92 F 7.58 0.10–27.96 13.21 F13.80 1.77–50.59 1.43 F 1.22 0.18–4.53

0.10 F 0.07 0.04–0.24 0.04 F 0.02 0.01–0.07 0.10 F 0.04 0.04–0.20 0.08 F 0.05 0.04–0.20 2.18 F 0.61 0.80–2.96 0.15 F 0.06 0.06–0.24

Dry deposition in Tokyo, Japan (Sakata and Marumoto, 2004). Dry deposition in Rome, Italy (Morselli et al., 2004). 1998 dry deposition data in Seoul (Yi et al., 2001). 1991 dry deposition data for Chicago (Holsen et al., 1993). 1994 dry deposition data for Chicago (Paode et al., 1998).

Tokyo, Japan (Sakata and Marumoto, 2004). Although the methods of sampling may differ, statistical evidence showed that the sticky deposition plate and water surface sampler methods give the same dry deposition results (Odabas¸V et al., 1999; Sakata and Marumoto, 2004). Therefore, the dry deposition data generated in this study could be compared with the older data at the same study site in Izmir. The last column in Table 3 gives the average values and ranges of wet and dry deposition of heavy metal fluxes found in this study and the previous column shows the dry deposition data reported earlier at the same site (Odabasi et al., 2002). From these last two columns of Table 3, a trend may be noted over 2 years’ time. Within this 2 years, the total Cr, Cd have increased, Cu, Zn and Ni did not show a significant change whilst there has been a decrease in the average value of Pb fluxes in dry deposition at the study site. Also the range of the lead fluxes is narrower and the maximum value is nearer to the average flux from the data obtained 2 years ago. At the beginning of 2004, leaded gasoline has been largely withdrawn from the market and this may be evaluated as a first indication of the positive impact on the environmental quality of the use of unleaded gasoline in the area. The higher dry deposition fluxes of the studied metals presumably come from the soil components as the sampler was only 3 m above the ground. Odabasi et al. (2002) showed older pollution in the soil composition which may cause reentrainment of metals in the air and contribute to the deposition in the study site. When we compare metal fluxes with corresponding wind directions, higher values occurred when it blew from the north. On the contrary lower fluxes occurred when southerly winds blew. In Izmir, average estimated annual total heavy metal deposition fluxes in wet and dry forms given in Table 2 were compared with corresponding results from elsewhere. Estimations of annual deposition rates were found out by using the following methods: (a) Wet annual deposition was calculated by extrapolation from the total average daily wet deposition determined in this study from the 13 sampling days out of the 31 rain events

14

A. Muezzinoglu, S.C. Cizmecioglu / Atmospheric Research 81 (2006) 1–16

throughout the year (Table 2). The day with the studied rain event is taken as a basis in expressing the wet deposition values. Total wet deposition flux for a brainy dayQ is found by multiplying the duration of rain in minutes by the flux per minutes actually calculated for this rain event from the measurements. This definition is because although the rain may cease during the day of sampling and therefore wet deposition stops, it is assumed that there remain no particles to settle by dry deposition. To convert the wet fluxes from Ag m 2 min 1 units to Ag m 2 year 1, the ratio between the annual precipitation (406.8 kg m 2) and the total amount of rain gauged during the sampling days (118.4 kg m 2) was used. (b) As there were only 31 rainy days during the study year, the working atmospheric cleansing mechanism was dry deposition for the remaining dry weather days of 334 days of the year. Thus annual dry deposited heavy metals were estimated by taking the average daily dry deposition rate in (Ag m 2 day 1) based on 12 days of dry deposition measurement data (Table 2) and multiplying it with the total number of dry deposition days. The estimated quantity of atmospheric cleansing with dry deposition was more important than that of wet deposition throughout the study year of 1 October 2003–1 November 2004 in Izmir. The annual dry and wet deposition fluxes of heavy metals were given in Table 4. The annual wet deposition fluxes of heavy metals were compared to the annual dry deposition fluxes of heavy metals in Izmir. The ratios of dry versus wet deposition per year were 5.6 times for Cr, 10.8 times for Cd, 8.5 times for Pb, 2.7 times for Cu, 16.1 times for Zn and 10.4 times for Ni. This comparison shows the relative significance of dry deposition in this Mediterranean climate area with a low intensity and frequency of rain incidences and low humidity in the air. To discuss the solubility of the deposited heavy metal salts, the percentage mass ratio between the dissolved fraction and total heavy metals in dry and wet deposition were calculated. Solubility of a metal salt depends upon the molecular form and on the environmental conditions. Although the meteorological parameters of temperature, wind speed, and humidity were used in explaining the distribution of the soluble fraction ratios in analyzed samples, no significant relationship could be found for these ratios. Values of pH measured in rainwater and dry deposition sampler water given in Table 1 were statistically correlated and a significance test is applied to discuss the impact of pH on the solubility of the heavy metals. There is no statistically significant relationship between dissolved and total concentrations and pH values for rainwater and dry deposition samples. The statistically significant relationships were found between the WSS water pH and dissolved and total concentrations of Pb and Ni at the 95.0% confidence level, only. As can be noted from Table 5, ratios of soluble fractions of heavy metals in the WSS samples and in the rainwater are of the order of 80%–99% for the metals except for one or two samples having less soluble Cd and Cr forms in deposition. Another exception to that is Ni which has a larger range of soluble fraction ratios in wet deposition than in dry deposition. This result is Table 4 Comparison of annual heavy metal total fluxesa of dry and wet deposition samples (kg km 2 year 1) Sample fraction

DD WD a

Metal Cr

Cd

Pb

Cu

Zn

Ni

32.5 F 17.5 5.8 F 5.2

14.0 F 6.8 1.3 F 1.8

32.3 F 13.0 3.8 F 3.9

27.4 F 16.8 10.0 F 26.0

728.8 F 203.9 45.4 F 47.4

51.0 F 21.3 4.9 F 4.2

Total flux values were not made up by summing the dissolved and suspended forms but found by separate analyses.

A. Muezzinoglu, S.C. Cizmecioglu / Atmospheric Research 81 (2006) 1–16

15

Table 5 Soluble fractions of the wet and dry deposited heavy metals Metal

Cr Cd Zn Pb Cu Ni

Soluble fraction in wet deposition (%)

Soluble fraction in dry deposition (%)

This study

Morselli et al. (2003)

This study

Morselli et al. (2003)

Average

Range

Average

Average

Range

Average

0.79 0.86 0.91 0.86 0.91 0.85

0.50–0.98 0.64–0.99 0.82–0.97 0.76–0.89 0.62–0.99 0.37–0.97

0.88 0.68 0.74 0.61 – –

0.89 0.74 0.89 0.84 0.87 0.89

0.58–0.96 0.43–0.99 0.58–0.98 0.77–0.90 0.60–0.99 0.76–0.96

0.33 0.25 – – –

comparable with the results of Morselli et al. (2003) reporting lower soluble fractions in dry deposition than in wet deposition in Bologna, Italy. The difference in Cr and Ni ratios may be due to the heavy traffic near the sampling area. This discussion, in relation to the soluble fraction ratios in deposition in Izmir, implies that the studied metals are ready to impose ecotoxic impacts in the water–soil environments and on biota. This impact is enhanced by highly soluble fractions in the dry deposition, too. Sudden increases in heavy metal deposition in dissolved form can be an important input into the biochemical cycles and may create more drastic impacts. The results of this study showed the importance of metal pollution in rainwater especially in regard to the ecotoxic properties of heavy metals, rather than by dry deposition. It also showed that pH is a relatively unimportant property in creating ecotoxic solubility effects. Acknowledgements ¨ BI˙TAK 100Y104 project are acknowledged Dokuz Eylul University Research Funds and TU ¨ mit Divrikli for partial financial support. Authors are thankful for the analytical help of Dr. U from Pamukkale University Chemistry Department and the encouraging comments from Drs. Mustafa Odabasi (Dokuz Eylul University) and Emu¨r Henden (Ege University). Also the Dokuz Eylu¨l University Air Pollution Laboratory staff and especially Dr. Remzi Seyfioðlu are gratefully acknowledged for sampling and analytical help. References Akkoyunlu, B.O., Tayanc¸, M., 2003. Analyses of wet and bulk deposition in four different regions of Istanbul, Turkey. Atmospheric Environment 37, 3571 – 3579. Al-Momani, I.F., Ataman, O.Y., Anwari, M.A., Tuncel, S., Ko¨se, C., Tuncel, G., 1995. Chemical composition of precipitation near an industrial area at Izmir, Turkey. Atmospheric Environment 29, 1131 – 1143. Al-Momani, I.F., Aygun, S., Tuncel, G., 1998. Wet deposition of major ions and trace elements in the eastern Mediterranean basin. Journal of Geophysical Research [Atmospheres] 103, 8287 – 8300. Cakan, A., 1999. The direct measurements of the dry deposition of organochlorine pesticides and polychlorinated naphthalenes. PhD thesis, Illinois Institute of Technology, Chicago, IL. Deboudt, K., Flament, P., Bertho, M.L., 2004. Cd, Cu, Pb and Zn concentrations in atmospheric wet deposition at a coastal station in Western Europe. Water, Air, and Soil Pollution 151, 335 – 359. Dincer, F., Muezzinog˘lu, A., Elbir, T., 2003. SO2 levels at forested mountains around Izmir, Turkey and their possible sources. Water, Air, and Soil Pollution 147, 331 – 341. Grantz, D.A., Garner, J.H.B., Johnson, D.W., 2003. Ecological effects of particulate matter. Environment International 29, 213 – 239.

16

A. Muezzinoglu, S.C. Cizmecioglu / Atmospheric Research 81 (2006) 1–16

Groempling, A.H.J., P. Ostapczuk, P., Emons, H., 1997. Wet deposition in Germany: long-term trends and the contribution of heavy metals. Chemosphere 34, 2227 – 2236. Gu¨lsoy, G., Tayanc, M., Ertu¨rk, F., 1999. Chemical analyses of the major ions in the precipitation of Istanbul, Turkey. Environmental Pollution 105, 273 – 280. Holsen, T.M., Noll, K.E., Fang, G.C., Lee, W.J., Lin, J.M., Keeler, G.J., 1993. Dry deposition and particle size distributions measured during the Lake Michigan Urban Air Toxics Study. Environmental Science and Technology 27, 1327 – 1333. Hu, G.P., Balasubramanian, R., 2003. Wet deposition of trace metals in Singapore. Water, Air, and Soil Pollution 144, 285 – 300. Jaradat, Q.M., Momani, K.A., Jiries, A.G., El-Alali, A., Batarseh, M.I., Sabri, T.G., Al-Momani, I.F., 1999. Chemical composition of urban wet deposition in Amman, Jordan. Water, Air, and Soil Pollution 112, 55 – 65. Kim, G., 1998. Atmospheric inputs and upper ocean biogeochemistry of trace elements and radionuclide tracers in the Atlantic. PhD thesis, University of Delaware. Morselli, L., Olivieri, P., Brusori, B., Passarini, F., 2003. Soluble and insoluble fractions of heavy metals in wet and dry atmospheric depositions in Bologna, Italy. Environmental Pollution 124, 457 – 469. Morselli, L., Brusori, B., Passarini, F., Bernardi, E., Francaviglia, R., Gataleta, L., Marchionni, M., Aromolo, R., Benedetti, A., Olivieri, P., 2004. Heavy metals monitoring at a Mediterranean natural ecosystem of Central Italy. Trends in different environmental matrixes. Environmental International 30, 173 – 181. Odabas¸V, M., Sofuog˘lu, A., Vardar, N., Tas¸demir, Y., Holsen, T.M., 1999. Measurement of dry deposition and air–water exchange of polycyclic hydrocarbons with the water surface sampler. Environmental Sciences 33, 426 – 434. Odabasi, M., Muezzinoglu, A., Bozlaker, A., 2002. Ambient concentrations and dry deposition fluxes of trace elements in Izmir, Turkey. Atmospheric Environment 36, 5841 – 5851. Paode, R.D., Sofuoglu, S.C., Sivadechathep, J., Noll, K.E., Holsen, T.M., 1998. Dry deposition fluxes and mass size distributions of Pb, Cu, and Zn measured in Southern Lake Michigan during AEOLOS. Environmental Science and Technology 32, 1629 – 1635. Sakata, M., Marumoto, K., 2004. Dry deposition fluxes and deposition velocities of trace metals in the Tokyo Metropolitan Area measured with a water surface sampler. Environmental Science and Technology 38, 2190 – 2197. Shahin, U.M., Zhu, X., Holsen, T.M., 1999. Dry deposition of reduced and reactive nitrogen: a surrogate surface approach. Environmental Science and Technology 33, 2113 – 2117. Takeda, K., Marumoto, K., Minamikawa, T., Sakugawa, H., Fujiwara, K., 2000. Three-year determination of trace metals and the lead isotope ratio in rain and snow depositions collected in Higashi}Hiroshima, Japan. Atmospheric Environment 34, 4525 – 4535. Tasdemir, Y., 1997. Modification and evaluation of a water surface sampler to investigate the dry deposition and air– water exchange of polychlorinated biphenyls (PCBs). PhD thesis, Illinois Institute of Technology, Chicago, IL. Tuncer, B., Bayar, B., Yesilyurt, C., Tuncel, G., 2001. Ionic composition of precipitation at the Central Anatolia (Turkey). Atmospheric Environment 35, 5958 – 6002. Yi, S.M., Holsen, T.M., Noll, K.E., 1997. Comparison of dry deposition predicted from models and measured with a water surface sampler. Environmental Science and Technology 31, 272 – 278. Yi, S.M., Lee, E.Y., Holsen, T.M., 2001. Dry deposition fluxes and size distributions of heavy metals in Seoul, Korea during yellow-sand events. Aerosol Science and Technology 35, 569 – 576. Yun, H.J., Yi, S.M., Kim, Y.P., 2002. Dry deposition fluxes of ambient particulate heavy metals in a small city, Korea. Atmospheric Environment 36, 5449 – 5458.