journal of applied geophysics - kocaeli...

Journal of Applied Geophysics 70 (2010) 46–57

Contents lists available at ScienceDirect

Journal of Applied Geophysics

j ourna l homepage: www.e lsev ie r.com/ locate / jappgeo

Magnetic susceptibility and heavy-metal contamination in topsoils along theIzmit Gulf coastal area and IZAYTAS (Turkey)

M. Canbay a,1, A. Aydin b,⁎, C. Kurtulus a,1

a Kocaeli University, Faculty of Engineering, Department of Geophysics, 41300, Izmit, Turkeyb Pamukkale University, Faculty of Engineering, Department of Geophysics, 20017, Kinikli, Denizli, Turkey

⁎ Corresponding author. Fax: +90 258 296 3370.E-mail addresses: [email protected] (M. Ca

(A. Aydin), [email protected] (C. Kurtulus).1 Fax: +90 262 303 3109.

0926-9851/$ – see front matter © 2009 Elsevier B.V. Adoi:10.1016/j.jappgeo.2009.11.002

a b s t r a c t
a r t i c l e i n f o
Article history:Received 7 July 2008Accepted 6 November 2009

Keywords:Magnetic susceptibilityHeavy metalsSoil contaminationIzmitTurkiye

The study on topsoil contamination due to heavy metals was carried out by using the Magnetic susceptibility(MS) measurements in Izmit industrial city, northern Turkey. We attempted to investigate correlationsbetween the concentration of selected heavy metals and the MS from 41 sample sites around Izmit Gulf.These investigations let us quantify and standardize the MS method, which may have consequences for longterm monitoring of anthropogenic pollution, especially in urban areas. The MS surfer contour map based onthe topsoil measurements was compiled with a randomly ranged distance density. The soil samples collectedthroughout the industrial areas, the parks, road sides and residential areas were also analyzed by AtomicAbsorption Spectrometer. Heavy metals Cu, Ni, Cr and Pb show strong correlations with MS, while Zn and Coshow a weak correlation with MS. Moreover, the Tomlinson pollution load index (PLI) shows insignificantcorrelation with the MS.The MS was examined vertically (0–30 cm) with respect to anthropogenic and/or lithogenic influences at thefourteen sample sites. The maximum values were mostly observed in depths of 2–5 cm and the MS values onthe depth profiles vary between 10×10−8 m3 kg−1 and 203×10−8 m3 kg−1. The study revealed that MS isan inexpensive, fast and non-destructive method for the detection and mapping of contaminated soils.

nbay), [email protected]

ll rights reserved.

© 2009 Elsevier B.V. All rights reserved.

1. Introduction

Taking magnetic measurements on soil sediments and identifiedpollution sources has become a widespread practice as a reliable,efficient and sensitive method for evaluating polluted samples. Thismethod is promising as an alternative to conventional chemicalanalysis. Magnetic susceptibility (MS) values depend on the compo-sition, grain size of magnetic minerals and their source, such aslithogenic, pedogenic and anthropogenic origins. In this paper, wedemonstrate an application of MS measurements to evaluate thedegree of pollution along the Izmit Gulf coastal area, Turkey.

MS operates as efficiently as chemical measurements on soil andsediment samples for monitoring anthropogenic pollutions. In thiscontext, several studies revealed important insights about pollutionin soil, air and water samples through heavy metal and industrialactivities. Led by Le Borgne's (1955, 1956) pioneeringwork, researchersidentified important sources of contamination such as seasonal changesin climate (Tite and Linington, 1975), lithological differences insediments due to changes in their depositional environments (Mullinsand Tite, 1973) and magnetic enhancement of topsoils due to natural

processes (Mullins, 1977). MS was shown to be highly effective forinvestigating industrial pollutants, traffic and other atmosphericpollutants (Thompson and Oldfield, 1986; Thornton, 1991; Strzyszcz,1993; Hay et al., 1997; Strzyszcz and Magiera, 1998; Durza, 1999;Hoffmann et al., 1999; Morris et al., 1995; Kapicka et al., 1999).

Many researchers reported that MS measurements are very usefulin investigating industrial discharges and exhaust gasses in urbanregions. It appears that atmospheric deposition is one of the majorsources of contamination in topsoil samples (Shu et al., 2000; Lecoanetet al., 2001; Goddu et al., 2004; Knab et al., 2001; Hanesch andScholger, 2005; Lu et al., 2007). Hanesch and Scholger (2002) also usedMSmeasurements to estimate anthropogenic sourced pollution levelsof the soils and sediments in urban area. Further, several authorsemphasized the value of MS studies in determining heavy-metalpollutions in Turkey (Aydin andGelisli, 1996; Gelisli et al., 1996;Gelisliand Aydin, 1998; Okay et al., 2001). Tolun et al. (2001) and Ozkul(2003) studied the effects of industrialism on heavy-metal concentra-tions in Izmit and surrounding areas through laboratory testing ofcollected samples. According to this study, Izmit Gulf region showshigh concentrations of Cu, Pb, Zn, Ni and Co.

To identify the type of industrial polluting sources, someresearchers used magnetic parameters. In this context, MS mappingof soils and sediments has become one of the most important tools forestimating anthropogenic pollution (Oldfield et al., 1985; Heller et al.,1998; Petrovsky et al., 2001; Schibler et al., 2002; Kapicka et al., 2003;

mailto:[email protected]



http://dx.doi.org/10.1016/j.jappgeo.2009.11.002

http://www.sciencedirect.com/science/journal/09269851

47M. Canbay et al. / Journal of Applied Geophysics 70 (2010) 46–57

Jordanova et al. 2004; Lu et al., 2005a,b). These studies illustrate therelationships between heavy-metal contents and the MS measure-ments in soil, road dust, airborne particle and sediment from sea, riverand lake samples. The use of magnetic measurement as a heavy-metalpollution indicator is based on the fact that origin of heavy metals andmagnetic particles is genetically related. By focusing on environmen-tal and magnetic factors, such studies demonstrated the relationshipbetween heavy-metal contents and magnetic properties, includinglithological and pedological properties in soils (Jordanova et al., 2004;Schmidt et al., 2005). Several studies confirmed a direct correlationbetween the MS of contaminated soils and the presence of hydro-carbons and certain heavy metals such as Pb, Zn, Cr, etc. (Morenoet al., 2003; Desenfant et al., 2004; Lu and Bai, 2006).

In this paper, we would like to compare the results obtainedfrom chemical analyses of heavy metals. Our main purposes are to(a) examine the relationship between the MS and heavy-metalcontamination along the Izmit Gulf coastal area, (b) obtain robustinformation on the relative degree of industrial pollution, and(c) characterize the anthropogenic magnetic particles in road dustmaterials in the industrial section of the region.

Fig. 1. Location and geological maps showin

2. Material and methods

2.1. Study area

Izmit is a coastal city located in the eastern Marmara Sea, nearIstanbul (Fig. 1), and is one of the most populated and industrializedprovinces of the Marmara Region (40°75′N, 29°91′E). It is one of thecountry's most important gateways by rail, road, sea and air. Startingfrom the 1960s and during the 1980s, some of industrial facilitieswereinitiated along the shores of the sheltered sea of the northernMarmaracoast and Izmit Bay. Today these areas are the “hot spots” in relation tothe environmental constraint of the industrial corporation. Its locationand the rapid industrialization resulted in accelerated urban develop-ment. Izmit manufacturing industry shares 10% of investment goods,2.8% of consumer product goods, and 27% chemicals within themanufacturing industry in Turkey. This region includes majorindustries, such as those that deal with iron production, and, refining,processing and transporting crude oils. Some of the industries andtheir sources of pollution are shown in Fig. 1. The city center has apopulation of 248,424 during the 2007 census. Neighboring Istanbul,

g the study area and sampling stations.

48 M. Canbay et al. / Journal of Applied Geophysics 70 (2010) 46–57

one of theworld's largestmetropolitan centers, Izmit is one of themostdensely populated cities of Turkey with 344 people per squarekilometer, with a population growth of 20% per year (Tepecik andAnderson, 2006). It is located on the crossroad between Asia andEurope, as well as the Izmit Gulf, which is a natural harbour and a busymaritime line. Climatic conditions of the Izmit Gulf region are usuallymild, whereas higher regions have more severe weather conditions.Izmit's climate can be seen as a transition betweenMediterranean andBlack Sea climates. While summers are hot with scarce rainfalls,winters are rainy and occasionally cold. The average annual temper-ature is 15 °C and average rainfall is 800 mm at Izmit.

One of the dominant landmarks in this area includes the Samanlimountain, with 1.330 m in height starting from the west side of theSakarya river, and reaching Pamukova and North of the Iznik Lakeuntil Bozburun. Izmit lies on one of Turkey's active tectonic fault lines.

Table 1Distribution of Cu, Pb, Zn, Ni, Cr, Cd and Co concentrations in the topsoil samples (0–5 cm),dependent susceptibility χFD% and lithology of the topsoil samples in Izmit, Turkey. MCIcommercial area, McI; Industrial polluted area around the factory and McP; Unpolluted and

Sample site Sampleidentification

Lon. Lat. Cu Pb Zn Ni Cr

(mg kg−1)

1 MCI-1 29.771 40.759 45.1 42.0 78.0 51.0 192 MCI-7 30.004 40.761 39.0 42.0 78.0 50.0 173 MCI-8 30.028 40.762 41.0 45.0 68.0 48.0 294 MCI-9 29.983 40.727 38.0 38.2 102.0 43.0 335 MCI-15 29.785 40.751 40.8 44.0 72.0 57.0 226 MCI-21 30.011 40.743 45.0 45.0 72.0 59.0 277 MCI-22 30.061 40.758 42.0 44.0 98.0 45.0 328 MCI-23 30.001 40.716 33.0 36.2 92.0 40.0 349 MCP-2 29.844 40.767 25.7 23.7 78.2 34.0 4110 MCP-13 29.894 40.795 17.9 14.3 74.3 21.0 7011 MCP-14 29.941 40.802 25.5 8.0 62.2 29.0 5512 MCP-16 29.865 40.754 20.2 24.1 77.2 24.0 3813 MCRC-3 29.899 40.769 36.0 35.5 75.0 42.0 3014 MCRC-4 29.969 40.789 34.0 14.4 69.7 31.0 5315 MCRC-6 29.952 40.758 42.4 39.1 67.2 47.0 3816 MCRC-11 29.897 40.717 40.7 33.3 64.6 38.0 4017 MCRC-12 29.782 40.722 31.8 29.2 102.2 36.0 3218 MCRC-17 29.923 40.761 33.4 37.2 72.0 39.1 2619 MCRC-18 29.978 40.776 37.1 15.6 65.7 30.0 5520 MCRC-20 29.961 40.762 39.1 41.1 64.7 45.0 3521 MCRC-25 29.861 40.723 37.5 31.5 63.1 36.0 3822 MCRC-26 29.761 40.721 34.6 25.7 87.3 35.0 3423 MCRC-27 29.839 40.709 38.1 33.5 64.8 34.0 3724 MCRC-28 29.821 40.712 31.2 26.8 83.3 36.0 3125 MCRS-5 30.012 40.804 28.0 17.1 61.0 29.0 4026 MCRS-10 29.951 40.723 42.4 40.0 64.0 53.2 2027 MCRS-19 30.003 40.811 22.4 15.1 60.0 24.0 3728 MCRS-24 29.961 40.723 47.2 42.5 74.0 55.8 19Range 18–47 8–45 60–102 21–59 17Mean±SD 25±8 32±11 74±12 40±10 35Median 37 34.5 72.0 39 34Standard error 1.5 2.1 2.3 1.9 2.395% confidence interval 2.9 4.3 4.6 4.0 4.729 McP-1 30.025 40.796 76.5 214.1 40.2 3.4 –

30 McP-2 30.023 40.796 54.9 48.1 64.7 2.3 –

31 McP-3 30.031 40.790 53.3 39.3 80.4 1.8 –

32 McP-8 30.055 40.789 74.4 53.4 64.1 3.8 –

33 McP-9 30.002 40.782 49.0 101.1 89.9 4.0 –

34 McP-10 30.014 40.773 26.4 74.6 40.8 3.5 –

35 McP-11 30.017 40.776 30.1 79.9 16.2 6.0 –

36 McP-12 30.028 40.776 34.9 90.5 43.7 5.4 –

37 McP-13 30.047 40.778 39.7 124.1 37.8 2.3 –

38 McI-4 30.028 40.787 59.2 12.8 91.0 5.3 –

39 McI-5 30.033 40.784 65.3 41.1 95.1 2.6 –

40 McI-6 30.031 40.782 86.6 55.2 68.1 5.2 –

41 McI-7 30.020 40.790 160.7 62.3 61.0 18.5 –

Range 26–160 13–214 16–95 2–19 –

Mean±SD 62±34 77±50 61±24 5±4 –

Median 55 62 64.1 4 –

Standard error 9.6 14.0 6.7 1.1 –

95% confidence interval 21 30 14.6 2.6 –

a Abbreviations are given in the geology map and locations of the sampling stations are

As a result, this area has often been destructively affected by severeearthquakes. Izmit was heavily damaged by the Marmara Earthquakeon August 17, 1999. More than 20,000 people died, over 70% of thebuildings collapsed, and Turkey's largest energy producing establish-ments were damaged (Tuysuz and Genc, 1999).

2.2. Geology

Fig. 1 shows the modified geological map of the region. Scholarswho have studied this area (Ozdemir et al., 1982; Yilmaz et al., 1995;Gedik et al., 1999; Yigitbas et al., 1999; Ozalaybey et al., 2008)have found the following patterns: Triassic aged the Cakraz formation(c-T2); Recrystallized limestone and pebbles, red colored sandstone,siltstone, claystone, Campanian–Maestrichtian aged the Peksimetformation (c-T2); sandstone dominant conglomerate, limestone and

Tomplinson pollution index (PLI), magnetic susceptibilities (χLF and χHF), frequency-; Industrial area, MCRS; Roadside, MCP; Park and green area, MCRC; Residential andpark area around the factory.

Cd Co PLI χLF χHF χFD% Forma

(10×−8 m3 kg−1)

.4 0.32 18.0 1.70 176 169 4.0 Q Izmit Gulfcoastalarea samples

.1 0.29 14.0 1.55 203 198 2.5 Q

.3 0.12 15.0 1.48 119 115 3.4 Q

.9 0.08 21.0 1.51 78 73 6.4 Q

.8 0.28 15.0 1.66 164 154 6.1 Q

.0 0.31 17.0 1.79 186 180 3.2 Q

.1 0.09 11.0 1.44 135 129 4.4 Q

.1 0.10 22.0 1.49 88 83 5.7 Q

.0 0.19 13.0 1.35 16 15 6.3 Q

.8 0.07 17.9 1.08 10 9 10.0 c-T2

.1 0.10 10.1 1.00 10 9 10.0 c-T2

.0 0.22 17.0 1.30 21 19 9.5 Q

.1 0.29 13.8 1.57 79 77 2.5 Q

.2 0.21 16.0 1.37 60 58 3.3 c-T2

.0 0.25 18.0 1.72 91 87 4.4 Q

.2 0.30 25.0 1.74 67 63 6.0 Q

.0 0.33 17.3 1.63 59 55 6.8 Q

.1 0.27 14.7 1.51 68 65 4.4 Q

.9 0.23 21.0 1.47 55 51 7.3 c-T2

.0 0.23 13.0 1.57 78 73 6.4 Q

.2 0.35 23.0 1.70 45 42 6.7 Q

.0 0.27 18.0 1.55 65 61 6.2 Q

.2 0.28 22.0 1.64 45 42 6.7 Q

.0 0.25 20.0 1.52 65 58 10.8 Q

.3 0.23 17.4 1.31 35 32 8.6 Q

.2 0.10 13.0 1.33 193 187 3.1 Q

.8 0.24 16.0 1.19 44 39 11.4 Q

.4 0.07 14.0 1.34 155 140 9.7 Q–71 0.07–0.35 10–25 1–1.8 10–203 9–198 3–11±12 0.21±0.09 17±3 1±0.2 86±57 81±56 6±3

0.23 17 1.5 68 64 60.02 0.7 0.03 11 11 0.50.03 1.4 0.08 22 21 10.07 4.8 1.56 15 15 5.8 Q IZAYTAS

topsoilsamples

0.06 2.2 1.00 14 14 0.7 Q0.04 17.0 1.26 17 17 5.5 c-T20.07 21.0 1.74 13 14 14.3 c-T20.07 21.9 1.94 16 15 8.8 c-T20.05 8.2 1.14 12 12 4.1 Q0.08 18.7 1.37 20 20 0.0 Q0.06 40.7 1.81 13 11 10.3 Q0.06 22.5 1.49 15 15 0.7 Q0.06 21.6 1.45 86 78 9.3 c-T20.05 10.2 1.37 76 70 7.5 c-T20.08 29.6 2.06 53 49 8.5 Q0.07 37.6 2.88 55 49 10.6 c-T20.04–0.08 2–41 1–3 12–86 11–78 0–140.06±0.01 20±12 1.6±1 31±26 29±24 7±40.06 21 1.5 16 15 7.50.001 3.2 0.13 7.3 6.7 1.20.007 7.1 0.3 16 14 2.6

in Fig. 1.


marl, Campanian–Lutesiyen (Ludavian) aged the Akveren formation(c-T1); clayey limestone, marl and sandstone, Campanian–Lutesiyen(Ludavian) aged the Caycuma formation (c-T2); sandstone, claystone,marl, limestone and pebblestone, Holocene aged Alluvium (Q); gravels,sand, silt and clay (Sariaslan, 1999; Ozalaybey et al., 2008). Formationtypes of sample sites are provided in Table 1.

2.3. Sampling

In this study, the sampling stations were considered to pursue theprojectwhich covers the Izmit Gulf coastal area, and a small area aroundthe IzmitWaste and Residue Treatment, Incineration and Recycling, Co.,Inc. (IZAYDAS) (Fig. 1). Sampling stations were subdivided into twomain groups: Stations 1 to 14 include both the topsoil and verticalprofileMSmeasurements taken fromanabout300 km2 area in the IzmitGulf coastal area, whereas stations 15 to 41 cover only topsoil samplesaround a factory of about 15 km2 area. The vertical MS measurementswere taken with 1-cm interval from 0 to 30 cm depth. Samplingstations were carefully selected to cover different land use zones asfollows: Industrial area (MCI), Roadside (MCRS), Park and green area(MCP), Residential and commercial area (MCRC), Industrial pollutedarea around the factory (McI) and Unpolluted and park area around thefactory (McP).

The sampling interval of all the topsoil was 0–5 cm. At eachsampling site, a sample was collected using a stainless steel trowel,and stored in a plastic bag. In addition, the samples were dried andstored in bags for use.

2.4. Magnetic measurements

To investigate magnetic features of the sediment samples, variouskinds of MS measurements were carried out both at laboratory and inthefield.MSmeasurementswere carriedout byusinganMSmeterMS2,Bartington Instruments Ltd., linked to MS2B dual-frequency sensor.Bartington MS2H down hole probe was used as a sub-surface probe inconjunction with the MS2 susceptibility meter for profiling the MS ofstrata in25 mmnominal diameter augerholes.MSmeasurements in thefield were carried out with a step of 1 cm along the vertical profiles. Thevolume susceptibility (κ)wasmeasured andmass-specific susceptibility(χ) was computed. Low frequency MS (χLF) and high frequencyMS (χHF) were measured using a dual-frequency Bartington MS2 MS

Fig. 2. Magnetic measurement results of the topsoil samples along the Izmit Gulf coastal areMcP around the factory.

meter. Frequency-dependent MS was defined as χFD(%)=[(χLF−χHF)/χLF]×100, here χLF and χHF represent susceptibility valuesmeasured at 0.47 and 4.7 kHz, respectively. Each data point wasmeasured 3–4 times in order to check reproducibility and to avoidmeasurement errors. The error of themeasurements ofMSwas found tobe smaller than 3%, using repeat-measurements.

2.5. Soil chemical analysis

Heavy-metal concentrations were determined only for topsoilsamples (0–5 cm) by atomic absorption spectrometry. The sampleswere dried and prepared following the protocol for acid digesting ofsediments. Soils were digested in an oven using a mixture solutionof the concentrated acids of HNO3-HF-HClO4. After digestion, thesolution was cooled and diluted with 10–15 mL of distilled water, andfiltered. The filtrate was then made up to 50 mL with distilled waterand then analyzed for Cu, Pb, Zn, Ni, Cr, Cd and Co using atomicabsorption spectrometer (Shimadazu model AA-6601F Shimatsu Ltd.,Australia). Analytical variability was tested by repeated analysis onevery ten samples. Experiments were performed in triplicate and thestandard deviation was computed and indicated in the plots as errorbars with a confidence limit of 95%. The optimum conditions ofthe instrument were wave length 283.3 nm, slit width 0.5 nm andlamp current 8–300 mA. The wave lengths and slit widths used were213.9 nm and 0.7 nm for Zn, 232.0 nm and 0.2 nm for Ni, 357.9 nmand 0.7 nm for Cr, 228 nm and 0.7 nm for Cd and 240 nm and 0.7 nmfor Co. The standard solutions of the elements were prepared in orderto obtain the calibration curves.

2.6. Data analysis

Statistical analyses of the magnetic measurements and chemicalcompositional data were carried out using Microsoft Excel 2002SP3. Correlation coefficients were obtained in order to establish therelationship between heavy-metal concentrations and magneticparameters in the topsoil samples. Mapping of the MS distributionwas created with the Surfer 4.0 software.

The Tomlinson pollution load index (PLI) was calculated toexamine its correlation with the MS. The linear correlation betweenthe PLI and MS originally was predicted by Angulo (1996). Chan et al.(2001) used the PLI to show howmuch a sample exceeds the contents

a, Turkey. MCI, MCRS and MCRC samples were collected around the Izmit Gulf. McI and


of heavy-metal background of the natural environments. The PLIwhich is a result of the contribution of several heavy metals is definedas the root of the multiplication of the concentration factors (CFHMk),

PLI =

ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi∏n

k=1CFHMk

n

sð1Þ

where CFHMk was the ratio of the concentrations of each heavy metal(CHM) to the amv given in Table 1 which means the mean values.

CFHMk =CHM

amv: ð2Þ

A good correlation between PLI and χLF suggested that the com-bined heavy-metal pollution degree could be assessed by the intensityof MS.

Fig. 3. Surfer contour maps of the χLF values (×10−8 m3 kg−1) and heavy-metal contamina(excluding IZAYTAS samples). The data used is given in Table 1.

3. Results

3.1. Topsoil samples

3.1.1. Characterization of magnetic propertiesTable 1 shows the results of the chemical analysis and magnetic

measurements (PLI, magnetic susceptibilities (χLF and χHF) andfrequency-dependent susceptibility (χFD)) on the topsoil samples. TheMS (χLF) measurements show that they vary from 10×10−8 m3 kg−1

to 203×10−8 m3 kg−1 with an average of 86×10−8 m3 kg−1. As illus-trated in Fig. 2, the χLF values can be divided into four different areas;MCI, McI (mean χLF: 143×10−8 m3 kg−1, ranged 78–203×10−8 m3

kg−1, mean χLF: 68×10−8 m3 kg−1, ranged 53–86×10−8 m3 kg−1),MCRS (mean χLF: 107×10−8 m3 kg−1, ranged: 15–193×10−8 m3

kg−1), MCRC (meanχLF: 65×10−8 m3 kg−1, ranged 45–91×10−8 m3

kg−1) and MCP, McP (mean χLF: 14×10−8 m3 kg−1, ranged 10–21×10−8 m3 kg−1, meanχLF: 15×10−8 m3 kg−1, ranged 12–20×10−8 m3

kg−1).

tions, created using Surfer 4.0, for the topsoils along the Izmit Gulf coastal area, Turkey


The χLF values depend on geology of the area, sedimentation andanthropogenic dope additive materials (Dearing et al., 1996; Lu et al.,2007). Lu et al. (2007) showed that the χLF value of natural ‘non-polluted’ topsoil samples depends on five major soil-forming factors:parent material, climate, geomorphology, vegetation and time. TheχLF of MCP non-polluted samples from the study area vary between10×10−8 m3 kg−1 and 21×10−8 m3 kg−1, which is much lowerthan the χLF measured for the polluted topsoil samples. The high χLFvalues in industrial areas (MCI-XX; XX shows the number of sample)and especially near the pollution units around the IZAYDAS can beexplained by the anthropogenic input of various origins although thesamples have similar soil types and geology.

Figs. 3 and 4 show the surfer contour maps of the χLF and heavy-metal content distributions for the topsoil samples. As illustrated, theχLF values differ significantly in different stations. It appears that thehigher χLF values are concentrated mostly in the industrial plantareas (heavy machine manufacture, petrochemical works, trafficemissions etc.). In Fig. 3, the high χLF values are concentrated mostlyin the eastern and western parts of the area where a lot of MCI arelocated. In the MCRC, the magnetic minerals may be derived from thetraffic emissions. The topsoil samples (McI and McP) taken from nearthe IZAYTAS yield high χLF values and heavy-metal contents. As seenin Fig. 4, the low χLF values and heavy-metal contents are foundmostly in the MCP compared to other areas. On the other hand, thehigh χLF values and heavy-metal concentrations are encountered in

Fig. 4. Surfer contour maps of χLF (×10−8 m3 kg−1) and heavy-metal contaminations prepused is given in Table 1.

the topsoil samples taken from the anthropogenic polluted areas asseen in the surfer contour maps (Figs. 3 and 4).

The frequency-dependent susceptibility (χFD%) values are givenin Table 1. Dearing (1999) showed the relation betweenχFD% and theparamagnetic component. χFD% of topsoil samples range from 3% to11% with a mean value of 6%. The mean values in the urban samplesincrease from the MCI (mean=4.7%) towards MCRC (mean=5.9%)and the MCRS (mean=8.2%) regions. Such a low value of χFD%indicates that the magnetic properties of the samples are predomi-nantly contributed by the coarse multidomain grains, rather than bythe super paramagnetic particles (Lu et al., 2007). Dearing's χFD%model (Dearing, 1999) predicted two types of magnetic clusters in thesamples. A few of samples with a χFD%b2% are dominated by coarsemulti domain grains. Fifteen χFD% values were placed between 2%and 6%, which belong to an intermediate group (Fig. 2). These valuescorrespond to a mixture of multi domain and stable single domain.Most samples were evaluated to be dominated by the frequency-dependent grains (χFD%N6%); hence the ferromagnetic grains (i.e.Fe) are present in high concentration (Table 1).

3.1.2. Characterization of heavy-metal contentsTable 1 shows the concentrations of the Pb, Zn, Ni, Cu, Cr, Cd and Co

values for the topsoil samples and their median, standard deviation,minimum and maximum values. We examined only seven heavymetals since they have sufficient potential to show possible pollution

ared by using Surfer 4.0, for the topsoil samples around IZAYTAS area, Izmit. The data


level changes. The topsoil samples contain high mean concen-trations around the gulf: Cu (25 mg kg−1), Pb (32 mg kg−1), Zn(74 mg kg−1), Cr (35 mg kg−1), Ni (40 mg kg−1), Cd (0.21 mg kg−1),and Co (17 mg kg−1) and around the IZAYTAS: Cu (62 mg kg−1),Pb (77 mg kg−1), Zn (61 mg kg−1),Ni (5 mg kg−1), Cd (0.06 mg kg−1),and Co (20 mg kg−1). Metal concentrations also show large standarddeviations alluding a great heterogeneity of heavy-metal pollution. Themeasured high concentrations of the heavy metals and the MS valuesof the MCI and MCRS topsoil samples indicate the presence of aconsiderable amount of magnetic particles accompanying the heavy-metal emission. Cu, Pb and Zn concentrations in the MCI and MCRS aresignificantly higher than those in the MCP. These metal concentrations

Fig. 5. The contents of Pb, Cu, Zn, Cr, Co, and Ni and the χLF for topsoil samples. Equations andthe graphs.

may be affected by the dust deposition whichmay be resulted from thevarious industrial activities. It has been well established that Cu is acommon element in automobile thrust bearing, brake lining, and otherparts of the engine (Lu et al., 2005a,b), therefore the corrosion of metalwear in the automobile engine may cause releasing of heavy metalsparticularly Cu and Pb to the environment and consequently accumu-lating on the topsoil. Pb has been used as an effective indicator of traffic-induced pollution caused by automobile exhausts (Lu et al., 2007).

3.1.3. Relationship between MS and heavy-metal contentChan et al. (1997) showed that certain heavymetals, such as Pb, Zn

and Cu are preferentially adsorbed onto the exterior surface of the fly

correlation coefficients (R2) between the χLF and heavy-metal contents were given on


ash and aerosols from the industrial emissions, which often containsignificant amounts of Fe oxides. A strong and positive correlationbetween heavy-metal content with χLF is therefore expected. Gelisliand Aydin (1998) claimed that the presence of common heavy metalsin the traffic emissions also accounts for the strong correlationsbetween MS and heavy metals in the urban topsoil side of the heavytraffic roads in Trabzon, Turkey. They emphasized that anthropogenicpollution was the dominant factor influencing topsoil susceptibility.Thus, they suggested that the high correlation coefficients betweenχLF measurements and heavy-metal content can be used as anindicator of pollution.

The scatter plots of MS versus Cu, Pb, Cr, Co, Ni and Zn contents aregiven in Fig. 5 where the correlation coefficients between the heavy-metal contents and MS are also given. In general, all heavy metalsexcept Co and Zn show good associations with χLF. The highestcoefficient between the MS and the heavy metal Ni is R2=0.82.Correlations between the MS and the concentration of other heavymetals are as follows: Pb (R2=0.61), Cr (R2=0.59), Cu (R2=0.60),Co (R2=0.06) and Zn (R2=0.0).

The correlations between χLF with heavy-metal contents werealso studied for the topsoil samples. As shown in Table 1 and Fig. 5, apositive correlation (R2N0) exists between the MS and the Pb, Ni, Cucontents and Cr (negative correlation) but not Co or Zn. The relationsbetween χLF and χFD%, and χLF and PLI were also shown for thesame samples (Fig. 6). As can be seen from Fig. 6, relatively very lowcorrelation coefficient is found between PLI and χLF (R2=0.20). Also,very low negative correlation between χFD% and χLF is found.

Fig. 6. Relationship between χHF, χFD%, PLI and χLF for the topsoil samples. Equationsand correlation coefficients (R2) between χLF and heavy-metal contents were given onthe graphs.

It is necessary to show the background value of MS in the studyarea to define the anthropogenic input or natural enhancement ofmagnetic minerals. Hanesch and Scholger (2002) showed that geo-chemical techniques could be used to distinguish existing naturalbackground from anthropogenic anomalies. These results were usedin order to reveal the sources of pollution. The value of the layer (10–30 cm) was used as an approximation for the soil background level,because the soiling process acts within this layer but anthropogenicpollution influences are to the upper horizons. Based on randomlyplaced profiles around the industrial region and rural areas (Fig. 7), thepresent study revealed the vertical variation of χLF. In Fig. 8, themeanχLF values get high levels of 85×10−8 m3 kg−1 at 5 cm; then thevalues drop off abruptly to a nearly constant level 40×10−8 m3 kg−1

at 10–30 cm. As a result, wewere able to use theχLF value at the depthof 10–30 cm as the background value.

As illustrated in Figs. 3 and 4, the surfer contourmaps showed highcorrelations of the heavy metals and χLF. As given in Fig. 3, there is ageneral increase in the χLF values near the probable pollution units ofMCI-XX and MCRS-XX. This is consistent with the distribution ofheavy metals. The high χLF values and heavy-metal contents arerecorded at the same points. Fig. 4 shows the distribution of thepattern of the surfer contour χLF values and heavy-metal contentsaround the IZAYTAS. The pattern shows a general increase in the χLFvalues and heavy-metal contents near the IZAYTAS. Both surfercontour χLF maps in Figs. 3 and 4 are used based on their capability tovisualize spatial relationships. In Fig. 3, higher values are concentratedin the southeastern part where there are a lot of industrial plants, forthe other parts of the city given the MS anomalies viewing by MCRC-XX where the magnetic minerals may be derived from the trafficemissions. The values of χLF are relatively smaller in MCP-XX com-pared to MCI-XX around the city and IZAYTAS, as shown in Figs. 3and 4. The magnetic enhancement of topsoil is mainly ascribed toanthropogenic inputs of magnetic minerals, compared to the back-ground χLF value. These results confirm the enhanced values of χLFmeasurements in the topsoil, which are weakly affected by lithology.These results also gave us information about the main magneticcarriers in these layers using the χFD and background χLF values forpedogenic or anthropogenic origins.

The relationship between the χLF and heavy-metal content for allthe topsoil samples is grouped in defined clumps. Clumps MCI andMCRS show higher Pb, Zn, Ni, Cu and Cr concentrations than theircorresponding clump of MCP and MCRC.

3.2. Characterization of the vertical MS profiles

The vertical profile variations of the χLF measurements belongingto the four different areas are shown in Fig. 7. The χLF values for eachprofile are given in Table 2. TheχLF vertical variation values of the soilsfrom the MCI-XX, ranged from 0 to 320×10−8 m3 kg−1 throughoutthe profiles (Fig. 7). These profiles are divided into four groups. Thegraphs of all soil profiles out of two profiles were taken roadside codedMCRS-XX from study area. The χLF values gradually increase from0 cm to 3–4 cm in its surface horizons (0–5 cm), and steeply decreasedfrom 5 cm to 10 cm horizon. After this depth the χLF values continueto decline. In addition, values reach their maximum levels, exceptthe profile of MCRS-XX where the maximum value is at 10 cm. Thesebehaviors explain clearly the contribution of ferrimagnetic minerals ofanthropogenic origin. MCP-XX profiles show bimodal χLF distribu-tions within the 0–5 cm interval where the χLF values were changingbetween 0 and 20×10−6 SI. The soil profiles of MCP-XX show lowcontributions of ferrimagnetic minerals of anthropogenic origin after5 cm. In particular, the upper 0–5 cm and further depths of soil profilefrom industrial area (MCI-XX) exhibit ten to fifteen times largerenhancement of χLF, compared to those from the background area.This indicates that urban sediments accumulated higher amounts ofanthrogenic materials.

Fig. 7. The vertical profile of the measured χLF values for the soil samples from the MCI, MCRS, MCP and MCRC polluted areas.


Soil profile from theMCRS-XX exhibits the largest enhancement ofχLF as its values are fifteen to twenty times higher than those from thebackground area. This indicates that these sediments most likelyaccumulated with a higher amount of anthrogenic materials partic-ularly sourced from the heavy traffic. In the MCRC-XX, MS valuescontain lower input of anthropogenic magnetic materials due to closeassociation to MCI-XX and MCRS-XX. These soil profiles exhibit lowermagnetic enhancement on top horizon but upper magnetic enhance-ment due to MCP-XX. The mean χLF values of all the vertical profilesof the each area are redrawn as shown in Fig. 8 which supports theabove mentioned interpretations. The two profiles (MCP13 andMCP14) for the Lower-middle Eocene aged are made up of thesandstone, claystone and conglomerate unit and at top the Quaternaryaged alluvium (i.e. gravel, sand, silt and clay). These sample sites arelocated far from anthrogenic pollution sources, and placed in a verylow χLF soil medium such as limestone. Therefore the χLF valueschange at low levels (0–20×10−8 m3 kg−1).

4. Discussion

The lateral and vertical changing of magnetic particles and heavymetals were observed in the sampling area along the Izmit Gulf coastalarea. Our results showed that the magnetic particles of samples inurban and residential areas formed multi domain grains originatedfrom anthropogenic sources. The sources of anthropogenic particleswere reported to increase the χLF of the urban topsoils. They wereidentified as emissions from fossil–fuel combustion processes, otherparticles from vehicles, building and road surfaces, and incorporatedinto disturbed soil surfaces (Lu et al., 2007). It was concluded that χLFhad been successfully used to detect fly ash and ash particles byautomobile emissions in the study area. On the other hand, onlylimited information was available about the value of the MS ofmetallurgical dusts. However, Lu et al. (2007) reported that Pb, Zn, Cdand Cu also account for the strong correlation between χLF and thesetwo metals in the soils, we find high correlations between MS and Pb,

Fig. 8. Mean χLF variations along the vertical profiles of soil samples from the MCI, MCRS, MCP and MCRC polluted areas.


Cu, Ni and Cr. For Zn and Cd, correlations were low. Lu et al. (2007)reported that the correlation between magnetic concentrations andheavy-metal contents reveals a causal relation between the ferrimag-netic oxide and the heavy metals in urban topsoils.

Pb, Cu, Zn and Cd are preferably adsorbed on the exterior surface offly ashes and aerosols from industrial emissions. Our results fromfour different areas supported Lu's results. These metals give a strongcorrelation with χLF (Lu et al., 2005a,b), and so we took the same

Table 2Vertically (0–30 cm; measurements were done for every 1 cm) measured χLF value distributions for 14 stations from the MCI, MCRS, MCP and MCRC areas. Locations of the stationsare shown in Fig. 1.

Sampling stations 1 9 13 14 26 15 2 3 4 25 16 17 10 11

Depth (cm) Magnetic susceptibility (χLF) (×10−8 m3 kg−1) Mean χLF (×10−8 m3 kg−1)

0 124.7 0.0 6.9 – – 17.6 0.0 140.0 120.4 – 64.0 0.0 – – 52.61 138.8 10.0 6.9 5.2 8.8 61.7 9.0 174.7 136.1 22.2 66.1 9.9 7.8 6.5 47.42 163.5 19.3 33.4 25.1 13.3 102.2 152.1 131.6 81.4 153.3 68.5 39.0 11.9 9.1 71.73 176.5 16.3 79.0 57.7 35.4 90.7 202.8 118.9 72.5 186.7 67.1 58.7 10.4 10.2 84.54 148.2 9.7 100.5 59.8 53.1 59.0 169.7 107.4 72.5 195.6 62.6 53.8 6.6 10.5 79.25 135.3 7.3 118.2 32.3 50.9 37.9 67.2 92.6 75.8 251.1 56.7 44.1 4.8 11.2 70.46 121.2 6.7 117.8 18.5 46.5 13.2 33.1 78.9 79.2 324.4 43.5 34.4 3.0 10.7 66.57 111.8 7.0 112.4 12.2 50.9 7.0 20.7 66.3 80.3 315.6 38.1 27.4 2.4 9.3 61.58 107.1 7.7 101.7 8.9 79.6 5.3 18.6 55.8 82.5 311.1 34.1 22.3 2.4 9.3 60.59 104.7 8.0 70.2 7.3 115.0 7.9 17.6 34.7 83.6 326.7 32.7 18.3 1.2 9.1 59.810 102.4 8.3 66.0 4.7 112.8 6.2 15.5 17.9 87.0 304.4 32.9 16.1 1.2 9.3 56.111 109.4 9.0 63.7 4.5 119.5 6.2 13.4 7.4 95.9 188.9 33.2 16.5 1.2 10.2 48.512 118.8 9.7 59.1 3.3 172.6 6.2 13.4 5.3 95.9 73.3 33.4 16.8 1.2 10.2 44.213 125.9 10.0 64.1 2.3 123.9 6.2 12.4 12.6 85.9 40.0 34.4 17.6 1.2 10.2 39.014 135.3 10.0 69.8 3.0 115.0 6.2 15.5 31.6 80.3 31.1 35.1 19.4 1.2 11.9 40.415 138.8 11.3 73.7 2.6 108.4 6.2 11.4 41.1 79.2 53.3 35.5 22.3 1.2 12.6 42.716 132.9 10.0 74.0 2.3 101.8 6.2 11.4 33.7 81.4 33.3 36.0 24.9 0.0 9.3 39.817 125.9 10.0 74.0 1.9 121.7 6.2 10.3 22.1 80.3 31.1 33.9 26.9 1.2 10.5 39.718 123.5 8.3 74.4 1.9 146.0 6.2 9.3 12.6 80.3 46.7 32.0 27.6 1.2 9.1 41.419 117.6 11.3 71.0 6.3 163.7 6.2 7.2 14.7 79.2 86.7 29.9 27.8 0.0 7.7 45.020 110.6 11.3 65.2 6.1 210.2 6.2 6.2 14.7 79.2 55.6 28.2 27.6 0.0 6.0 44.821 103.5 11.3 61.0 2.8 272.1 6.2 0.0 14.7 64.7 28.9 26.6 27.6 0.0 5.8 44.722 94.1 10.0 56.0 2.6 267.7 6.2 0.0 22.1 65.8 24.4 25.4 27.4 0.0 7.2 43.523 89.4 11.3 51.8 2.6 276.5 6.2 0.0 13.7 70.3 17.8 24.0 27.4 0.6 7.2 42.824 84.7 9.0 48.3 2.6 287.6 6.2 5.2 11.6 61.3 13.3 22.6 27.4 0.0 7.9 42.025 81.2 8.7 44.5 1.6 – – 0.0 10.5 61.3 88.9 21.2 27.4 0.3 8.4 29.526 78.8 7.3 43.7 .0 – – – 11.6 61.3 88.9 19.8 27.6 0.0 7.4 31.527 75.3 8.7 42.6 .0 – – – 12.6 58.0 66.7 17.6 27.6 0.0 5.8 28.628 71.8 5.7 42.2 .0 – – – – – 0.0 16.5 28.0 0.0 5.1 18.829 70.6 – 41.4 .0 – – – – – 0.0 15.1 28.2 0.0 3.5 19.930 72.9 – 54.1 .0 – – – – – 22.2 13.6 28.5 – 3.0 27.8


results near highways. In metallurgical dusts and fly ash in the studiedarea, these researchers showed a strong correlation between χLFand Fe, Mn, Zn, Pb and Ni. In this study, we observed very strongcorrelations between χLF and four heavy elements in both samplingareas.

5. Conclusions

In this study, we tested the potential use of the χLF measurementsfor lateral and vertical distributions of heavy-metal pollution moni-toring in urban soil samples. The χLF measurements for topsoil sam-ples everywhere in the study area were higher than the backgroundlevel, whichwas deeper than 10 cm. TheχLF values show a significantcorrelation with the concentration of heavy metals except for Zn andCo. In contrast to chemical analysis techniques, the χLF measurementtechniques provide us cheaper and less time-consuming methods foridentification of probable soil pollution. Also, we showed that χLFmeasurement could be used as a proxy measure for the degree ofheavy-metal contamination, revealing the distribution of pollutionparts in such industrial areas. The following conclusions are drawn;

1. Evaluated values of χLF (N20×10−8 m3 kg−1) with higher heavy-metal concentrations of the soil samples proved to be strongindicators of anthropogenic contribution in the soil samples.

2. χLF measurements could yield useful information on the degree ofpollution. This study shows that χLF shows strong correlationswith heavy metals such as Pb, Cu, Cr and Ni; and weak correlationswith Zn and Co.

3. χLF measurements show that the main magnetic components inurban soil samples are multidomain grains of ferrimagneticminerals, which are introduced by industrial activities, automobileexhaust, and deposition of atmospheric particulates.

4. The enrichment of magnetic particles and heavy metals in thetopsoil is considerably clear in industrial and roadside areas.

5. The Tomlinson pollution load index (PLI) shows no correlationwith the MS. Since analytical measurements of heavy-metal con-tamination are expensive and laborious, a straight linear correla-tion between χLF and the mineral concentration of Cu, Cd and Pbconfirms the use of magnetic technique as a simple, rapid, and non-destructive tool for assessment of heavy-metal contamination inthe investigated region.

Acknowledgements

Funding by PAU (2006MHF006) is kindly acknowledged. Theauthors would like to warmly thank Prof. Dr. Kadir GURGEY fromGeology Department, Pamukkale University and Dr. Ulas KAPLANfrom Department of Psychology, James Madison University for theirthoughtful reviews and many helpful suggestions in preparing themanuscript and for the English language and comments. The authorswould like to thank Editor-in-chief Prof. Dr. Michel Chouteau, JournalManager Paul Pearson, anonymous reviewers and editors for thesuggestions and comments to improve the manuscript.

References

Angulo, E., 1996. The Tomlinson pollution load index applied to heavy metal “Mussel-Watch” data: a useful index to assess coastal pollution. Sci. Total Environ. 187,19–56.

Aydin, A., Gelisli, K., 1996. Magnetic studies in Saruhan–Bayburt scarn zone. Jeofizik 10(1–2), 40–49 in Turkish.

Chan, L.S., Yeung, C.H., Yim, W.W.S., Or, O.L., 1997. Correlation between magneticsusceptibility and distribution of heavy metals in contaminated sea-floorsediments of Hongkong Harbour. Environ. Geol. 36, 77–86.

Chan, L.S., Ng, S.L., Davis, A.M., Yim, W.W.S., Yeung, C.H., 2001. Magnetic properties andheavy-metal contents of contaminated seabed sediments of Penny's bay HongKong. Mar. Pollut. Bull. 42 (7), 569–583.


Dearing, J.A., 1999. Environmental magnetic susceptibility, using the Bartington MS2system, 2nd Ed. Chi Publishing.

Dearing, J.A., Dann, R.J.L., Hay, K., Lees, J.A., Loveland, P.J., Maher, B.A., O'Grady, K., 1996.Frequency-dependent susceptibility measurements of environmental materials.Geophys. J. Int. 124, 228–240.

Desenfant, F., Petrovsky, E., Rochette, P., 2004. Magnetic signature of industrialpollution of stream sediments and correlation with heavy metals: case study fromSouth France. Water Air Soil Pollut. 152, 297–312.

Durza, O., 1999. Heavy metals contamination and magnetic susceptibility in soilsaround metallurgical plan. Phys. Chem. Earth, Part A Solid Earth Geod. 24 (6),541–543.

Gedik, I., Timur, E., Alan, L., Altun, L., Pehlivan, S., Akbas, B., Duru, M., 1999. The GeologicProject of the Kocaeli Peninsula. MTA. in Turkish.

Gelisli, K., Aydin, A., 1998. Investigations of environmental pollution using magneticsusceptibility measurements. Eur. J. Environ. Eng. Geophys. (3), 53–61.

Gelisli, K., Aydin, A., Dondurur, D., Livaoglu, S., 1996. Pollution investigations near theregion of Yeniay–Çamburnu (Trabzon) by the magnetic susceptibility method.Jeofizik 10 (1–2), 50–56 in Turkish.

Goddu, S.R., Appel, E., Jordanova, D., Wehland, F., 2004. Magnetic properties of road dustfrom Visakhapatnam (India) — relationship to industrial pollution and road traffic.Phys. Chem. Earth 29, 985–995.

Hanesch, M., Scholger, R., 2002. Mapping of heavy metal loadings in soils by means ofmagnetic susceptibility measurements. Environ. Geol. (42), 857–870.

Hanesch, M., Scholger, R., 2005. The influence of soil type on the magnetic susceptibilitymeasured throughout soil profiles. Geophys. J. Int. 161, 50–56.

Hay, K.L., Dearing, J.A., Baban, S.M.J., Loveland, P.A., 1997. Preliminary attempt toidentify atmospherically-derived pollution particles in English topsoils frommagnetic susceptibility measurements. Phys. Chem. Earth 22, 207–210.

Heller, F., Strzyszca, Z., Magiera, T., 1998. Magnetic record of industrial pollution inforest soils of Upper Silesia, Poland. J. Geophys. Res. 103, 17767–17774.

Hoffmann, V., Knab, M., Appel, E., 1999. Magnetic susceptibility mapping of roadsidepollution. J. Geochem. Explor. 66, 313–326.

Jordanova, D., Hoffmann, V., Fehr, K.T., 2004. Mineral magnetic characterization ofanthropogenic magnetic phases in the Danube river sediments. Earth Planet. Sci.Lett. 221, 71–89.

Kapicka, A., Petrovsky, E., Ustjak, S., Machackova, K., 1999. Proxy mapping of fly-ashpollution of soils around a coal-burning power plant: a case study in the CzechRepublic. J. Geochem. Explor. 66, 291–297.

Kapicka, A., Jordanova, N., Petrovsky, E., Podrazsky, V., 2003. Magnetic study of weaklycontaminated forest soils. Water Air Soil Pollut. 148, 31–44.

Knab, M., Appel, E., Hoffmann, V., 2001. Separation of the anthropogenic portion ofheavy metal contents along a highway by means of magnetic susceptibility andfuzzy c-means cluster analysis. Eur. J. Environ. Eng. Geophys. 6, 125–140.

Le Borgne, E., 1955. Susceptibilité magnétique anormale du sol superficiel. Ann.Geophys. 11, 399–419.

Le Borgne, E., 1960. Influence du feu sur les propriétés magnétiques du sol et sur cellesdu schiste et du granite. Ann. Geophys. 16, 159–195.

Lecoanet, H., Lévêque, F., Ambrosi, J., 2001. Magnetic properties of saltmarsh soilscontaminated by iron industry emissions (southeast France). J. Appl. Geophys. 48,67–81.

Lu, S.G., Bai, S.Q., Cai, J.B., Xu, C., 2005a. Magnetic properties and heavymetal contents ofautomobile emission particulates. J. Zhejiang Univ. Sci. 6 (B), 731–735.

Lu, S.G., Bai, S.Q., Cai, J.B., Xu, C., 2005b. Magnetic properties and heavy metal contentsof automobile emission particulates. J. Zhejiang Uni. Sci. 6 (B), 731–735.

Lu, S.G., Bai, S.Q., 2006. Study on the correlation of magnetic properties and heavymetals content in urban soils of Hangzhou city, China. J. Appl. Geophys. 60, 1–12.

Lu, S.G., Bai, S.Q., Xue, Q.F., 2007. Magnetic properties as indicators of heavy metalspollution in urban topsoils: a case study from the city of Luoyang, China. Geophys. J.Int. 171, 568–580.

Moreno, E., Sagnotti, L., Dinares-Turell, J., Winkler, A., Cascella, A., 2003. Biomonitoringof traffic air pollution in Rome using magnetic properties of tree leaves. Atmos.Environ. 37, 2967–2977.

Morris, W.A., Versteeg, J.K., Bryant, D.W., Legzdins, A.E., Mccary, B.E., Marvin, C.H., 1995.Preliminary comparisons between mutagenicity and magnetic susceptibility ofrespirable airborne particulate. Atmos. Environ. 29, 3441–3450.

Mullins, C.E., 1977. Magnetic susceptibility of the soil and its significance in soil science:a review. J. Soil Sci. 28, 223–246.

Mullins, C.E., Tite, M.S., 1973. Magnetic viscosity, quadrature susceptibility andfrequency dependence of susceptibility in single-domain assemblages of magnetiteand maghaemite. J. Geophys. Res. 78, 804–809.

Okay, O.S., Tolun, L., Telli, K.F., Tufekci, V., Tufekci, H., Morkoc, E., 2001. Izmit Bay(Turkey) ecosystem after Marmara earthquake and subsequent refinery fire: thelong-term data. Mar. Pollut. Bull. 42 (5), 361–369.

Oldfield, F., Hunt, A., Jones, M.D.H., Chester, R., Dearing, J.A., Olsson, L., Prospero, J.M.,1985. Magnetic differentiation of atmospheric dusts. Nature 317, 516–518.

Ozalaybey, S., Zor, E., Tapırdamaz, M.C., Tarancıoglu, A., Ozalaybey, S.C., Erkan, B.,Karaaslan, A., Alpaslan, E., Ergin, M., Engintav, S., Tan, E., 2008. Site Classificationand Seismic Hazard Assessment of Kocaeli Province of Turkey. Marmara ResearchCenter of TÜBİTAK Project No: 5057105, Kocaeli. 190 pp. (in Turkish).

Ozdemir, U., Yurtsever, G., Dager, Z., Yurtsever, A., 1982. The geology of Mesozoic–Cenozoic rocks in Gebze–Hereke–Tepecik. Biostratigraphic project of Kocaeli Trias.Report no: 7195. MTA. 143 pp.

Ozkul, C., 2003. A preliminary investigation into industrial impacts on heavy metalconcentrations in the topsoil around Izmit (Kocaeli). Master thesis, KocaeliUniversity, Natural and Applied Sciences Institute, Kocaeli, Turkey (in Turkish).

Petrovsky, E., Kapicka, A., Jordanova, N., Boruvka, L., 2001.Magnetic properties of alluvialsoils contaminated with lead, zinc and cadmium. J. Appl. Geophys. 48, 127–136.

Sariaslan, M., 1999. Natural Resources and Geology of around Kocaeli. ReportNo. 10195. MTA, p. 143. In Turkish.

Schibler, L., Boyko, T., Ferdyn, M., Gajda, B., Holl, S., Jordanova, N., Rösler, W., Team, M.A.P.R.O.X., 2002. Topsoil magnetic susceptibility mapping: data reproducibility andcompatibility, measurement strategy. Stud. Geophys. Geod. 46, 43–57.

Schmidt, A., Yarnold, R., Hill, M., Ashmore, M., 2005. Magnetic susceptibility as proxy forheavy metal pollution: a site study. J. Geochem. Explor. 85, 109–117.

Shu, J., Dearing, J.A., Morse, A.P., Yu, L.Z., Li, C., 2000. Magnetic properties of dailysampled total suspended particulates in Shanghai. Environ. Sci. Technol. 34,2392–2400.

Strzyszcz, Z., 1993. Magnetic susceptibility of soils in the areas influenced by industrialemissions. Soil Monitoring, Monte Verita. Birkhäuser, Basel.

Strzyszcz, Z., Magiera, T., 1998. Magnetic susceptibility and heavymetals contaminationin soils of Southern Poland. Phys. Chem. Earth 23 (9–10), 1127–1131.

Tepecik, A., Anderson, E., 2006, Sustainable development within the region strategiesand the complex social system what Turkey and the USA can learn from Skane,Sweden. Blekinge Institute of Technology. Master thesis.

Thompson, R., Oldfield, F., 1986. Environment Magnetism. Allen and Unwin, London.Thornton, I., 1991. Metal contamination of soils in urban areas. In: Bullock, P., Gregory,

P.J. (Eds.), Soils in the Urban Environment. Blackwell, Oxford, pp. 47–75.Tite, M.S., Linington, R.E., 1975. Effect of climate on the magnetic susceptibility of soils.

Nature 265, 565–566.Tolun, L., Okay, O.S., Gaines, A.F., Tolay, M., Tufekci, H., Kiratli, N., 2001. The pollution

status and toxicity of surface sediments in Izmit Bay, Turkey. Environ. Inter. 26,163–169.

Tuysuz, O., Genc, C.S., 1999. Tectonic features of the surface rupture of August 17, 1999earthquake and its effects on the engineering structures. ITUIAHS InternationalConference on the Kocaeli Earthquake, Scientific Assessments and Recommenda-tions for Re-building. ITU, Istanbul, Turkey.

Yigitbas, E., Elmas, A., Yilmaz, Y., 1999. Pre-Cenozoic tectono-stratigraphic componentsof the Western Pontides and their geological evolution. Geol. J. 34, 55–74.

Yilmaz, Y., Genc, S.C., Yigitbas, E., Bozcu, M., Yilmaz, K., 1995. Geological evolution of theLate Mesozoic continental margin of Northwestern Anatolia. Tectonophysics 243,155–171.

journal of applied geophysics - kocaeli...

Documents