soil pollution project work methodology pdf

Global assessment of soil pollution

• News, Stories & Speeches
• Get Involved
• Structure and leadership
• Committee of Permanent Representatives
• UN Environment Assembly
• Funding and partnerships
• Policies and strategies
• Evaluation Office
• Secretariats and Conventions

• Asia and the Pacific
• Latin America and the Caribbean
• New York Office
• North America
• Climate action
• Nature action
• Chemicals and pollution action
• Digital Transformations
• Disasters and conflicts
• Environment under review
• Environmental rights and governance
• Extractives
• Fresh Water
• Green economy
• Ocean, seas and coasts
• Resource efficiency
• Sustainable Development Goals
• Youth, education and environment
• Publications & data

Global Assessment of Soil Pollution

Soil pollution is a chemical degradation process that consumes fertile soils, with implications for global food security and human health. Soil pollution hampers the achievement of Sustainable Development Goals (SDGs), including achieving zero hunger, ending poverty, ensuring healthy lives and human well-being, halting and reversing land degradation and biodiversity loss, and making cities safe and resilient. Most contaminants originate from human activities and enter into the environment because of unsustainable production chains, consumption patterns or inappropriate waste disposal practices.

In May 2018, FAO and its Global Soil Partnership (GSP), the World Health Organization (WHO), the Secretariat of the Basel, Rotterdam and Stockholm Convention and the United Nations Environment Programme (UNEP) organized the Global Symposium on Soil Pollution (GSOP18) to bring together science and policy to understand the status, causes, impacts and solutions to soil pollution. The Outcome document of the symposium, ‘ Be the solution to soil pollution ’ paved the way to the implementation of a coordinated set of actions to # StopSoilPollution .

This report considers both point source contamination and diffuse pollution, and detail also the risks and impacts of soil pollution on human health, the environment and food security, without neglecting soil degradation and the burden of disease resulting from exposure to polluted soil.

The Global Assessment of Soil Pollution report and its Summary for Policy makers will be launched on 4th June are a response to this request and as part of the World Environment Day celebrations and the launch of the UN Decade on Ecosystem Restoration. This report and its summary, coordinated by the FAO’s GSP, the ITPS, and UNEP, are the product of an inclusive process involving scientists from all regions.

• Nature Action

© 2024 UNEP Terms of Use Privacy   Report Project Concern Report Scam Contact Us

Soil Pollution: Causes and Consequences

• First Online: 03 November 2018

Cite this chapter

• Bhupendra Koul 3 &
• Pooja Taak 3  

2504 Accesses

8 Citations

There has been a rapid rise in the soil pollution over the last two decades which has posed threat to living beings and the ecosystem as well. Soil pollution is caused by both natural and anthropogenic activities. Former includes volcanic eruptions, earthquakes, tsunamis etc. while the later includes metals (trace and heavy metals), chemicals and radioactive wastes. The chemicals can be grouped into pesticides and allied chemicals, crude petroleum and its derivatives and polymers, plasticizers and other wastes. Radioactive wastes include nuclear power generation wastes and other by products released from nuclear technology (medicines and research). These are harmful substances which stay in the ecosystem for long duration during which they get accumulated and biomagnified to concentration potentially toxic to organisms at higher trophic levels in the food chain. Most of these chemicals are carcinogenic, teratogenic and mutagenic in nature. It is therefore crucial to develop tools to assess potential risks of human exposure to pollutants and to decide threshold concentrations in soils in order to protect human health.

Monitoring of soil quality is a difficult process because of the scarcity of monitoring variables and other indicators. The alarming situation of the state of soil pollution has forced the scientific community to develop innovative, reproducible strategies/technologies ( in situ or ex situ ) for treating polluted soils. This chapter summarizes various aspects of soil pollution, its causes and consequences.

This is a preview of subscription content, log in via an institution to check access.

Access this chapter

Subscribe and save.

• Get 10 units per month
• Download Article/Chapter or Ebook
• 1 Unit = 1 Article or 1 Chapter
• Cancel anytime
• Available as PDF
• Read on any device
• Instant download
• Own it forever
• Available as EPUB and PDF
• Durable hardcover edition
• Dispatched in 3 to 5 business days
• Free shipping worldwide - see info

Tax calculation will be finalised at checkout

Purchases are for personal use only

Institutional subscriptions

Abdul KSM, Jayasinghe SS, Chandana EP, Jayasumana C, De Silva PMC (2015) Arsenic and human health effects: a review. Environ Toxicol Pharmacol 40(3):828–846

Article   PubMed   CAS   Google Scholar  

Abedin MJ, Cotter-Howells J, Meharg AA (2002) Arsenic uptake and accumulation in rice ( Oryza sativa L.) irrigated with contaminated water. Plant Soil 240(2):311–319

Article   CAS   Google Scholar  

Adriano DC (2003) Trace elements in terrestrial environments; biogeochemistry, bioavailability and risks of metals, 2nd edn. Springer, New York

Google Scholar  

Afal A, Wiener SW (2014) Metal toxicity. Medscape.org

Ahmad I, Akhtar MJ, Zahir ZA, Jamil A (2012) Effect of cadmium on seed germination and seedling growth of four wheat ( Triticum aestivum L.) cultivars. Pak J Bot 44(5):1569–1574

Alexander J (2015) Selenium. In: Handbook on the toxicology of metals, 4th edn. Academic, London, pp 1175–1208

Chapter   Google Scholar  

Alloway BJ (1995) Heavy metals in soils. Blackie Academic and Professional, an Imprint of Chapman and Hall, London

Anawar HM, Akter F, Solaiman ZM, Strezov V (2015) Biochar: an emerging panacea for remediation of soil contaminants from mining, industry and sewage wastes. Pedosphere 25(5):654–665

Article   Google Scholar  

Arlai A, Nakkong R, Samjamin N, Sitthipaisarnkun B (2012) The effects of heating on physical and chemical constitutes of organic and conventional okra. Procedia Eng 32:38–44

Arya SK, Roy BK (2011) Manganese induced changes in growth, chlorophyll content and antioxidants activity in seedlings of broad bean ( Vicia faba L.). J Environ Biol 3(26):707–711

Asensio V, Vega FA, Singh BR, Covelo EF (2013) Effects of tree vegetation and waste amendments on the fractionation of Cr, Cu, Ni, Pb and Zn in polluted mine soils. Sci Total Environ 443:446–453

Article   CAS   PubMed   Google Scholar  

Asrar Z, Khavari-Nejad RA, Heidari H (2005) Excess manganese effects on pigments of Mentha spicata at flowering stage. Arch Agron Soil Sci 51(1):101–107

Barrachina AC, Carbonell FB, Beneyto JM (1995) Arsenic uptake, distribution, and accumulation in tomato plants: effect of arsenite on plant growth and yield. J Plant Nutr 18(6):1237–1250

Barsukova VS, Gamzikova OI (1999) Effects of nickel surplus on the element content in wheat varieties contrasting in Ni resistance. Agrokhimiya 1:80–85

Bashiri F, Ahmadi R, Khezri SM (2015) Remove soil contaminants by heat treatment. Int J Fund Arts Architect 1(1):8–12

Batterman S (2004) Findings on assessment of small-scale incinerators for health-care waste water, sanitation and health protection of the human environment. World Health Organization, Geneva

Becker M, Asch F (2005) Iron toxicity in rice-conditions and management concepts. J Plant Nutr Soil Sci 168(4):558–573

Berlin M, Zalups RK, Fowler BA (2015) Mercury. In: Handbook on the toxicology of metals, 4th edn. Elsevier, Amsterdam, pp 1013–1075

Bernard A (2008) Cadmium and its adverse effects on human health. Indian J Med Res 128(4):557–564

CAS   PubMed   Google Scholar  

Beyersmann D (2002) Effects of carcinogenic metals on gene expression. Toxicol Lett 127(1–3):63–68

Bonnet M, Camares O, Veisseire P (2000) Effects of zinc and influence of Acremonium lolii on growth parameters, chlorophyll A fluorescence and antioxidant enzyme activities of ryegrass ( Lolium perenne L. cv Apollo). J Exp Bot 51(34):945–953

Brevik EC (2013) Soils and human health: an overview. In: Brevik EC, Burgess LC (eds) Soils and human health. CRC Press, Boca Raton, pp 29–56

Cachada A, Rocha-Santos T, Duarte AC (2018) Soil and pollution: an introduction to the main issues. Soil pollution, pp 1–28

Cassidy DP, Srivastava VJ, Dombrowski FJ, Lingle JW (2015) Combining in situ chemical oxidation, stabilization, and anaerobic bioremediation in a single application to reduce contaminant mass and leachability in soil. J Hazard Mater 297:347–355

Catalani S, Rizzetti MC, Padovani A, Apostoli P (2012) Neurotoxicity of cobalt. Hum Exp Toxicol 31:421–437

Chary NS, Kamala CT, Raj DS (2008) Assessing risk of heavy metals from consuming food grown on sewage irrigated soils and food chain transfer. Ecotoxicol Environ Saf 69:513–524

Cook CM, Kostidou A, Vardaka E, Lanaras T (1997) Effects of copper on the growth, photosynthesis and nutrient concentrations of Phaseolus plants. Photosynthetica 34(2):179–193

Cox MS, Bell PF, Kovar JL (1996) Differential tolerance of canola to arsenic when grown hydroponically or in soil. J Plant Nutr 19(12):1599–1610

Cox A, Venkatachalam P, Sahi S, Sharma N (2016) Silver and titanium dioxide nanoparticle toxicity in plants: a review of current research. Plant Physiol Biochem 107:147–163

DOE (2002) EU Council Directive 1999/31/EC of 26 April 1999 on the landfill of waste. A consultation paper on limiting landfill to meet the EC Landfill Directive’s targets for reducing the landfill of biodegradable municipal waste. Environmental Policy Division, Department of the Environment

Doncheva S, Stoynova Z, Velikova V (2001) Influence of succinate on zinc toxicity of pea plants. J Plant Nutr 24(6):789–804

Doncheva S, Georgieva K, Vassileva V, Stoyanova Z, Popov N, Ignatov G (2005) Effects of succinate on manganese toxicity in pea plants. J Plant Nutr 28(1):47–62

Du X, Zhu YG, Liu WJ, Zhao XS (2005) Uptake of mercury (Hg) by seedlings of rice ( Oryza sativa L.) grown in solution culture and interactions with arsenate uptake. Environ Exp Bot 54(1):1–7

Du P, Xie Y, Wang S, Zhao H, Zhang Z, Wu B, Li F (2015) Potential sources of and ecological risks from heavy metals in agricultural soils, Daye City, China. Environ Sci Pollut Res 22:3498–3507

Dupuis J, Knoepfel P (2015) Episode IV: consolidation of the institutional regime for contaminated sites, and gambling on the remediation objectives, methods and funding of the bonfol site 2001–2008. In the politics of contaminated sites management. Springer International Publishing, pp 127–146

Eapen S, D’Souza SF (2005) Prospects of genetic engineering of plants for phytoremediation of toxic metals. Biotechnol Adv 23:97–114

Eapen S, Singh S, D’Souza SF (2007) Advances in development of transgenic plants for remediation of xenobiotic pollutants. Biotechnol Adv 25:442–451

Ellingsen DG, Møller LB, Aaseth J (2015) Copper. In: Handbook on the toxicology of metals, 4th edn. Elsvier/Academic Press, Amsterdam, pp 765–786

Escarré J, Lefèbvre C, Raboyeau S, Dossantos A, Gruber W, Cleyet Marel JC, Frérot H, Noret N, Mahieu S, Collin C, van Oort F (2011) Heavy metal concentration survey in soils and plants of the Les Malines mining district (Southern France): implications for soil restoration. Water Air Soil Pollut 216:485–504

Falciglia PP, Vagliasindi FGA (2015) Remediation of hydrocarbon polluted soils using 2.45 GHz frequency-heating: influence of operating power and soil texture on soil temperature profiles and contaminant removal kinetics. J Geochem Explor 151:66–73

Fowler BA, Sullivan D, Sexton MJ (2015) Bismuth. In: Handbook on the toxicology of metals, 4th edn. Academic, San Diego, pp 655–666

Fox RD (1996) Critical review: physical/chemical treatment of organically contaminated soils and sediments. J Air Waste Manag 46:391–431

Gautam M, Agrawal M (2017) Phytoremediation of metals using vetiver ( Chrysopogon zizanioides (L.) Roberty) grown under different levels of red mud in sludge amended soil. J Geochem Explor 182:218–227

Gidarakos E, Petrantonaki M, Anastasiadou K, Schramm KW (2009) Characterization and hazard evaluation of bottom ash produced from incinerated hospital waste. J Hazard Mater 172(2–3):935–942

Glick BR (2003) Phytoremediation: synergistic use of plants and bacteria to clean up the environment. Biotechnol Adv 21(5):383–393

Glick BR (2010) Using soil bacteria to facilitate phytoremediation. Biotechnol Adv 28(3):367–374

Gomes HI, Dias-Ferreira C, Ribeiro AB (2013) Overview of in situ and ex situ remediation technologies for PCB-contaminated soils and sediments and obstacles for full-scale application. Sci Total Environ 445:237–260

Gupta M, Gupta S (2017) An overview of selenium uptake, metabolism, and toxicity in plants. Front Plant Sci 7:2074

Article   PubMed   PubMed Central   Google Scholar  

GWRTAC (1997) Remediation of metals-contaminated soils and groundwater. Tech Rep TE-976-01, Pittsburgh

Hassanein RA, Hashem HA, El-Deep MH, Shouman A (2013) Soil contamination with heavy metals and its effect on growth, yield and physiological responses of vegetable crop plants (turnip and lettuce). J Stress Physiol Biochem 9:145–162

Henry JM, Cring FD (2013) Geophagy: an anthropological perspective. In: Brevik EC, Burgess LC (eds) Soils and human health. CRC Press, Boca Raton, pp 179–198

Holler JS, Fowler BA, Nordberg GF (2015) Silver. In: Handbook on the toxicology of metals, 4th edn. Elsevier/Academic Press, Amsterdam, pp 1209–1216

Hussain A, Abbas N, Arshad F, Akram M, Khan ZI, Ahmad K, Mansha M, Mirzaei F (2013) Effects of diverse doses of lead (Pb) on different growth attributes of Zea mays L. Agric Sci 4(5):262–265

CAS   Google Scholar  

Igwe JC, Nnorom IC, Gbaruko BC (2005) Kinetics of radionuclides and heavy metals behavior in soils: implications for plant growth. Afr J Biotechnol 4(13):1541–1547

Jayakumar K, Jaleel CA, Vijayarengan P (2007) Changes in growth, biochemical constituents, and antioxidant potentials in radish ( Raphanus sativus L.) under cobalt stress. Turk J Boil 31(3):127–136

Jayakumar K, Jaleel CA, Azooz MM (2008) Phytochemical changes in green gram ( Vigna radiata ) under cobalt stress. Glob J Mol Sci 3(2):46–49

Jayakumar K, Rajesh M, Baskaran L, Vijayarengan P (2013) Changes in nutritional metabolism of tomato ( Lycopersicon esculentum Mill.) plants exposed to increasing concentration of cobalt chloride. Int J Food Nut Saf 4(2):62–69

Jiang W, Liu D, Hou W (2001) Hyperaccumulation of cadmium by roots, bulbs and shoots of garlic. Bioresour Technol 76(1):9–13

Jiang X, Xiong Z, Liu H, Liu G, Liu W (2017) Distribution, source identification, and ecological risk assessment of heavy metals in wetland soils of a river-reservoir system. Environ Sci Pollut Res 24:436–444

Kabir M, Iqbal MZ, Shafiq M (2009) Effects of lead on seedling growth of Thespesia populnea L. Adv Environ Biol 3(2):184–190

Kang MY, Cho SH, Lim YH, Seo JC, Hong YC (2013) Effects of environmental cadmium exposure on liver function in adults. Occup Environ Med 70:268–273

Kapusta P, Sobczyk Ł (2015) Effects of heavy metal pollution from mining and smelting on enchytraeid communities under different land management and soil conditions. Sci Total Environ 536:517–526

Karishnamurti GSR, Naidu R (2003) Solid-solution equilibria of cadmium in soils. Geoderma 113:17–30

Khalid BY, Tinsley J (1980) Some effects of nickel toxicity on rye grass. Plant Soil 55(1):139–144

Khan MS, Zaidi A, Wani PA, Oves M (2009) Role of plant growth promoting rhizobacteria in the remediation of metal contaminated soils. Environ Chem Lett 7:1–19

Kibra MG (2008) Effects of mercury on some growth parameters of rice ( Oryza sativa L.). Soil & Environ 27:23–28

Kirpichtchikova T, Manceau A, Lanson B, Sarret G, Marcus MA, Jacquet T (2003) Speciation and mobility of heavy metals (Zn, Cu, Pb) in soil contaminated by sewage irrigation. J Phys 107:695–698

Kjær C, Elmegaard N (1996) Effects of copper sulfate on black bindweed ( Polygonum convolvulus L.). Ecotoxicol Environ Saf 33(2):110–117

Article   PubMed   Google Scholar  

Klein C, Costa M (2015) Nickel. In: Handbook on the toxicology of metals, 4th edn. Elsevier, Amsterdam, pp 1091–1111

Kong Z, Glick BR (2017) The role of plant growth-promoting bacteria in metal phytoremediation. In Adv Micro Physiol 71:97–132

Koubová K, Tlustoš P, Břendová K, Száková J, Najmanová J (2016) Lead accumulation ability of selected plants of noccaea spp. Soil Sediment Contam Int J 25(8):882–890

Koul B (2014) Expression of insecticidal toxin coded by modified full-length and truncated Bt- cry1Ac genes in transgenic tomato for assessment of their stability and efficacy against target insects. PhD thesis, Banasthali Vidyapith, Rajasthan, India

Kumar V, Chopra AK (2017) Phytoremediation potential of water caltrop ( Trapa natans L.) using municipal wastewater of the activated sludge process-based municipal wastewater treatment plant. Environ Technol 22:1–12

Kuppusamy S, Thavamani P, Megharaj M, Venkateswarlu K, Lee YB, Naidu R (2016) Pyrosequencing analysis of bacterial diversity in soils contaminated long term with PAHs and heavy metals: implications to bioremediation. J Hazard Mater 317:169–179

Langård S, Costa M (2015) Chromium. In: Handbook on the toxicology of metals, 4th edn. Elsevier, Amsterdam, pp 717–742

Li Z, Ma Z, van der Kuijp TJ, Yuan Z, Huang L (2014) A review of soil heavy metal pollution from mines in China: pollution and health risk assessment. Sci Environ 468:843–853

Lin YC, Kao CH (2005) Nickel toxicity of rice seedlings: cell wall peroxidase, lignin, and NiSO 4 - inhibited root growth. Crop Environ Bioinf 2:131–136

Ling W, Shen Q, Gao Y, Gu X, Yang Z (2007) Use of bentonite to control the release of copper from contaminated soils. Aus J Soil Res 45(8):618–623

Lison D (2015) Cobalt. In: Handbook on the toxicology of metals, 4th edn. Elsevier, Amsterdam, pp 743–763

Lucchini RG, Aschner M, Kim Y, Šarić M (2015) Manganeese. In: Handbook on the toxicology of metals, 4th edn. Elsevier, Amsterdam, pp 975–1011

Luo L, Gu JD (2016) Alteration of extracellular enzyme activity and microbial abundance by biochar addition: Implication for carbon sequestration in subtropical mangrove sediment. J Environ Manag 182:29–36

Ma Y, Oliviera RS, Nai F, Rajkumar M, Luo Y, Rocha I, Freitas H (2015) The hyperaccumulator Sedum plumbizincicola harbors metal-resistant endophytic bacteria that improve its phytoextraction capacity in multi-metal contaminated soil. J Environ Manag 156:62–69

Macek T, Mackova M, Kas J (2000) Exploitation of plants for the removal of organics in environmental remediation. Biotechnol Adv 18:23–34

Manivasagaperumal R, Balamurugan S, Thiyagarajan G, Sekar J (2011) Effect of zinc on germination, seedling growth and biochemical content of cluster bean ( Cyamopsis tetragonoloba (L.) Taub). Curr Bot 2(5):11–15

Marchand C (2017) Phytoremediation of soil contaminated with petroleum hydrocarbons and trace elements. PhD dissertation, Linnaeus University, Sweden

Marin AR, Pezeshki SR, Masscheleyn PH, Choi HS (1993) Effect of dimethylarsinic acid (DMAA) on growth, tissue arsenic and photosynthesis of rice plants. J Plant Nutr 16(5):865–880

Maslin P, Maier RM (2000) Rhamnolipid-enhanced mineralization of phenanthrene in organic-metal co-contaminated soils. Biorem J 4(4):295–308

McLaughlin MJ, Zarcinas BA, Stevens DP, Cook N (2000a) Soil testing for heavy metals. Commun Soil Sci Plant Anal 31(11–14):1661–1700

McLaughlin MJ, Hamon RE, McLaren RG, Speir TW, Rogers SL (2000b) Review: a bioavailability-based rationale for controlling metal and metalloid contamination of agricultural land in Australia and New Zealand. Aust J Soil Res 38(6):1037–1086

Mesa V, Navazas A, González-Gil R, González A, Weyens N, Lauga B, Peláez AI (2017) Use of endophytic and rhizosphere bacteria to improve phytoremediation of arsenic-contaminated industrial soils by autochthonous Betula celtiberica . Appl Environ Microbiol 83(8):11–16

MingHo Y (2005) Environmental toxicology: biological and health effects of pollutants, 2nd edn. CRC Press, BocaRaton

Moqsud MA, Omine K (2013) Bioremediation of agricultural land damaged by tsunami. In: Rolando Chamy, Francisca Rosenkranz (ed) Biodegradation of hazardous and special products. InTech https://doi.org/10.5772/56595

Moral R, Navarro Pedreno J, Gomez I, Mataix J (1995) Effects of chromium on the nutrient element content and morphology of tomato. J Plant Nutr 18(4):815–822

Moustakas M, Lanaras T, Symeonidis L, Karataglis S (1994) Growth and some photosynthetic characteristics of field grown Avena sativa under copper and lead stress. Photosynthetica 30(3):389–396

Nagajyoti PC, Lee KD, Sreekanth TVM (2010) Heavy metals, occurrence and toxicity for plants: a review. Environ Chem Lett 8(3):199–216

Nath K, Saini S, Sharma YK (2005) Chromium in tannery industry effluent and its effect on plant metabolism and growth. J Environ Biol 26:197–204

Navarro S, Vela N, Navarro G (2007) Review. An overview on the environmental behaviour of pesticide residues in soils. Span J Agric Res 5(3):357–375

Nematshahi N, Lahouti M, Ganjeali A (2012) Accumulation of chromium and its effect on growth of Allium cepa cv. hybrid. Euro J Exp Bio 2(4):969–974

Ou-Yang X, Chen JW, Zhang XG (2010) Advance in supercritical CO 2 fluid extraction of contaminants from soil. Geol B China 29(11):1655–1661

Panda SK, Patra HK (2000) Nitrate and ammonium ions effect on the chromium toxicity in developing wheat seedlings. Proc Natl Acad Sci 70:75–80

Pandolfini T, Gabbrielli R, Comparini C (1992) Nickel toxicity and peroxidase activity in seedlings of Triticum aestivum L. Plant Cell Environ 15(6):719–725

Paoli L, Fiorini E, Munzi S, Sorbo S, Basile A, Loppi S (2013) Antimony toxicity in the lichen Xanthoria parietina (L.) Th. Fr. Chemosphere 93(10):2269–2275

Pavel LV, Gavrilescu M (2008) Overview of ex situ decontamination techniques for soil cleanup. Environ Eng Manag J 7(6):815–834

Ponka P, Tenenbein M, Eaton JW (2015) Iron. In: Handbook on the toxicology of metals, 4th edn. Elsevier, Amsterdam, pp 879–902

Rani A, Souche Y, Goel R (2013) Comparative in situ remediation potential of Pseudomonas putida 710A and Commamonas aquatica 710B using plant ( Vigna radiata (L.) wilczek) assay. Ann Microbiol 63(3):923–928

Rascio N, Navari-Izzo F (2011) Heavy metal hyperaccumulating plants: how and why do they do it? And what makes them so interesting? Plant Sci 180(2):169–181

Reeves RD (2006) Hyperaccumulation of trace elements by plants. In: Phytoremediation of metal-contaminated soils. Springer, Dordrecht, pp 25–52

Rounsevell MDA, Loveland PJ (1992) An overview of hydrologically controlled soil responses to climate change in temperate regions. Seesoil 8:69–78

Ryszka P, Turnau K (2007) Arbuscular mycorrhiza of introduced and native grasses colonizing zinc wastes: implications for restoration practices. Plant Soil 298:219–229

Sachan S, Singh SK, Srivastava PC (2007) Buildup of heavy metals in soil water plant continuum as influenced by irrigation with contaminated effluent. J Environ Sci Eng 49:293–296

Sandstead HH (2015) Zinc. In: Handbook on the toxicology of metals, 4th edn. Elsevier, Amsterdam, pp 1369–1385

Semple KT, Morriss AWJ, Paton GI (2003) Bioavailability of hydrophobic organic contaminants in soils: fundamental concepts and techniques for analysis. Eur J Soil Sci 54:809–818

Sharma DC, Sharma CP (1993) Chromium uptake and its effects on growth and biological yield of wheat. Cereal Res Commun 21(4):317–322

Shaylor H, McBride M, Harrison E (2009) Sources and impacts of contaminants in soil. Cornell Waste Management Institute

Shekar CHC, Sammaiah D, Shasthree T, Reddy KJ (2011) Effect of mercury on tomato growth and yield attributes. Int J Pharma Biosci 2(2):358–364

Sheldon AR, Menzies NW (2005) The effect of copper toxicity on the growth and root morphology of Rhodes grass ( Chloris gayana Knuth.) in resin buffered solution culture. Plant Soil 278(1–2):341–349

Shenker M, Plessner OE, Tel-Or E (2004) Manganese nutrition effects on tomato growth, chlorophyll concentration, and superoxide dismutase activity. J Plant Physiol 161(2):197–202

Sheoran IS, Singal HR, Singh R (1990) Effect of cadmium and nickel on photosynthesis and the enzymes of the photosynthetic carbon reduction cycle in pigeonpea ( Cajanus cajan L.). Photo Res 23(3):345–351

Singh RP, Singh P, Arouja ASF, Ibrahim MH, Sulaiman O (2011) Management of urban solid waste: vermicomposting a sustainable option. Resourc Conser Recycl 55:719–729

Skerfving S, Bergdahl IA (2015) Manganeese. In: Handbook on the toxicology of metals, 4th edn. Elsevier, Amsterdam, pp 911–967

Smith LA, Means JL, Chen A (1995) Remedial options for metals-contaminated sites. Lewis Publishers, Boca Raton

Stenuit B, Eyers L, Schuler L, Agathos SN, George I (2008) Emerging high-throughput approaches to analyze bioremediation of sites contaminated with hazardous and/or recalcitrant wastes. Biotechnol Adv 26(6):561–575

Stockamp NW, Thompson GR (2016) Coccidioidomycosis. Infect Dis Clin N Am 30:229–246

Tylenda CA, Sullivan DW Jr, Fowler BA (2015) Antimony. In: Handbook on the toxicology of metals, 4th edn. Elsevier, Amsterdam, pp 565–579

Van der Ent A, Baker AJ, Reeves RD, Pollard AJ, Schat H (2013) Hyperaccumulators of metal and metalloid trace elements: facts and fiction. Plant Soil 362(1–2):319–334

Van der Perk M (2006) Soil and water contamination. Taylor & Francis, London

Book   Google Scholar  

Veetil DP, Mercier G, Blais JF, Chartier M, Tran LH, Taillard V (2014) Remediation of contaminated dredged sediments using physical separation techniques. Soil Sediment Contam Int J 23(8):932–953

Verbruggen N, Hermans C, Schat H (2009) Molecular mechanisms of metal hyperaccumulation in plants. New Phytol 181(4):759–776

Vieira da Silva FB, Araujo do Nascimento CW, Muniz Araujo PR, Vieira da Silva LH, da Silva RF (2016) Assessing heavy metal sources in sugarcane Brazilian soils: an approach using multivariate analysis. Environ Monit Assess 188(8):457

Wang M, Zou J, Duan X, Jiang W, Liu D (2007) Cadmium accumulation and its effects on metal uptake in maize ( Zea mays L.). Bioresour Technol 98(1):82–88

Wang P, Menzies NW, Lombi E, Sekine R, Blamey FPC, Hernandez-Soriano MC, Kopittke PM (2015) Silver sulfide nanoparticles (Ag 2 S-NPs) are taken up by plants and are phytotoxic. Nanotoxicology 9(8):1041–1049

Wei B, Yang LA (2010) review of heavy metal contaminations in urban soils, urban road dusts and agricultural soils from China. Microchem J 94:99–107

WHO (2011) Polychlorinated dibenzodioxins and dibenzofurans, in Air Quality Guidelines for Europe 2000, European Series no. 91, chapter 5.11 WHO Regional Office for Europe, Copenhagen, Denmark

WHO (2018) Health care wastes. In: Avenue Appia 20 1202 Geneva

Wild A (1993) Soils and the environment. Cambridge University Press, Cambridge

Wuana RA, Okieimen FE (2011) Heavy metals in contaminated soils: a review of sources, chemistry, risks and best available strategies for remediation. ISRN Ecol 2011:1–20

Xu F, Liu X, Chen Y, Zhang K, Xu H (2016) Self-assembly modified-mushroom nano composite for rapid removal of hexavalent chromium from aqueous solution with bubbling fluidized bed. Sci Rep 6:26201

Article   CAS   PubMed   PubMed Central   Google Scholar  

Yao Z, Li J, Xie H, Yu C (2012) Review on remediation technologies of soil contaminated by heavy metals. Procedia Environ Sci 16:722–729

Yourtchi MS, Bayat HR (2013) Effect of cadmium toxicity on growth, cadmium accumulation and macronutrient content of durum wheat (Dena CV.). Int J Agri Crop Sci 6(15):1099–1103

Zhang S, Qiu CB, Zhou Y, Jin ZP, Yang H (2010) Bioaccumulation and degradation of pesticide fluroxypyr are associated with toxic tolerance in green alga Chlamydomonas reinhardtii . Ecotoxicology 20:337–347

Zuverza-Mena N, Armendariz R, Peralta-Videa JR, Gardea-Torresdey JL (2016) Effects of silver nanoparticles on radish sprouts: root growth reduction and modifications in the nutritional value. Front Plant Sci 7:90–97

Download references

Author information

Authors and affiliations.

School of Bioengineering & Biosciences, Lovely Professional University, Phagwara, Punjab, India

Bhupendra Koul & Pooja Taak

You can also search for this author in PubMed   Google Scholar

Rights and permissions

Reprints and permissions

Copyright information

© 2018 Springer Nature Singapore Pte Ltd.

About this chapter

Koul, B., Taak, P. (2018). Soil Pollution: Causes and Consequences. In: Biotechnological Strategies for Effective Remediation of Polluted Soils. Springer, Singapore. https://doi.org/10.1007/978-981-13-2420-8_1

Download citation

DOI : https://doi.org/10.1007/978-981-13-2420-8_1

Published : 03 November 2018

Publisher Name : Springer, Singapore

Print ISBN : 978-981-13-2419-2

Online ISBN : 978-981-13-2420-8

eBook Packages : Biomedical and Life Sciences Biomedical and Life Sciences (R0)

Share this chapter

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Publish with us

Policies and ethics

• Find a journal
• Track your research

An official website of the United States government

The .gov means it’s official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

• Publications
• Account settings

Preview improvements coming to the PMC website in October 2024. Learn More or Try it out now .

• Advanced Search
• Journal List
• Int J Environ Res Public Health

Method for Assessing the Integrated Risk of Soil Pollution in Industrial and Mining Gathering Areas

1 College of Environmental Science and Engineering, Nankai University, Tianjin 300071, China; E-Mails: ten.haey@devdenarik (Y.G.); nc.ude.iaknan@tiemuj (M.J.)

Chaofeng Shao

2 Department of Soil Pollution and Control, Chinese Research Academy of Environmental Sciences, Beijing 100012, China; E-Mails: nc.gro.searc@bqug (Q.G.); nc.gro.searc@naiqgnahz (Q.Z.)

Industrial and mining activities are recognized as major sources of soil pollution. This study proposes an index system for evaluating the inherent risk level of polluting factories and introduces an integrated risk assessment method based on human health risk. As a case study, the health risk, polluting factories and integrated risks were analyzed in a typical industrial and mining gathering area in China, namely, Binhai New Area. The spatial distribution of the risk level was determined using a Geographic Information System. The results confirmed the following: (1) Human health risk in the study area is moderate to extreme, with heavy metals posing the greatest threat; (2) Polluting factories pose a moderate to extreme inherent risk in the study area. Such factories are concentrated in industrial and urban areas, but are irregularly distributed and also occupy agricultural land, showing a lack of proper planning and management; (3) The integrated risks of soil are moderate to high in the study area.

1. Introduction

Industrial and mining activities have always been leading sources of soil pollution [ 1 , 2 ]. In China, mining and industrial gathering areas have been established and rapidly developed across the country. According to several studies [ 3 , 4 , 5 ] and the National Soil Pollution Survey Bulletin [ 6 ], over ten million hectares of land in China have been threatened by soil pollution. Among these, two million hectares are threatened by mining and five million are threatened by petroleum pollution. Moreover, because of irrational planning and rough development in the initial construction stage, types, range and potential risks of the pollutants in soil of industrial and mining gathering areas are intricate and should not be underestimated.

In the interests of preventing and controlling soil pollution caused by industrial and mining activities, several studies have conducted ecological and health risk assessments, evaluation criteria grading, and spatial distributions of pollutants in different regions of China. Since the 1990s, China has developed quality standards for general soil environments [ 7 ] and for soils allocated to specific land usages [ 8 , 9 , 10 , 11 ]. Li et al. [ 12 ] summarized the published data (2005–2012) on soils polluted with heavy metals originating from mining areas in China, and then comprehensively assessed the heavy metal pollution derived from these mines based on soil pollution levels and human health risks. In previous studies, the mines and surrounding areas were identified as sources and heavy metals were considered the most serious pollutants. The spatial distribution [ 2 , 13 , 14 , 15 ], risk assessment [ 2 , 13 , 16 , 17 , 18 ], and mobility [ 19 , 20 ] of pollutants in soils also provided important information.

However, the risk assessment of soil pollutants has focused almost exclusively on heavy metal pollution. Organic pollution, especially the organic pollutants generated by the petrochemical industries, is largely ignored, despite its strong presence in the industrial and mining gathering areas and its potential for serious harm. Moreover, although most researchers have analyzed the risk sources, the survey, classification, grading, and management methods of polluting factories (the main sources of pollution in industrial and mining gathering areas) are relatively simple and not associated with the risk analysis. Finally, the difficulty of investigation and data collection has precluded a spatial analysis that combines a soil pollution study with an evaluation of the inherent risk level of polluting factories. Therefore, to understand the overall status of the soil environment in industrial and mining gathering areas, a comprehensive method that considers the soil risk caused by complex pollutants and pollution sources is necessary.

The present study proposes a comprehensive method for evaluating soil risk status. It includes the human health risk, the inherent risk level of the polluting factories and evaluated risk regionalization and characteristics of pollution sources. The method was piloted in Binhai New Area, Tianjin, China, a typical area of high mining and industrial activity.

2. Methodology

This study assesses human health risk and the inherent risk levels of polluting factories. From these results, a comprehensive method for assessing hazardous soil environments in industrial and mining gathering areas is developed.

2.1. Human Health Risk Assessment

By assessing human health risk, we can characterize the potential health hazards imposed by environmental pollution and elucidate the impacts and damage to human health [ 21 , 22 ]. The latter (including the carcinogenic and non-carcinogenic health risks of soil contamination) are revealed by the land use patterns and exposure pathways. Appendix A describes the assessment of human health risk on both organic pollutants and heavy metal contaminants. This assessment provides a scientific basis and technical support for comprehensive risk management.

2.2. Inherent Risk Assessment of Polluting Factories

Polluting factories, refer to factories engaged in industrial production or other industries, which may directly or indirectly cause large-scale environmental or ecological pollution. As mentioned in the Introduction, mining and industrial activities are major sources of soil contamination in industrial and mining gathering area. The operating conditions, pollutant emission levels, environmental management, and risk prevention levels of polluting factories are all important affecters of soil environmental risks. Therefore, to guide the soil environmental management in industrial and mining gathering areas, the risk assessment of polluting factories should be included in the soil environmental risk assessment.

2.2.1. Evaluation Index System

The polluting factories in industrial and mining gathering areas significantly differ in type, pollutant emission characteristics, and risk supervision level. Therefore, the environmental risks also differ among factories. The evaluation system is divided into three levels. The first layer, referred to as the target layer, measures the overall level of soil environmental risk posed by polluting factories. The second layer, the criteria layer, includes the inherent degree of sudden environmental risks, degree of cumulative environmental risk, and degree of environmental risk supervision by factories. The third level includes the assessment indicators. Appendix B described the construction of indicator system and the weight distribution of indicators, the results are listed in Table 1 .

Risk assessment indicators of polluting factories.

2.2.2. Scoring of Indicators and Comprehensive Assessment

Because the dimensions of each index in the index system are variable, these indices cannot be directly calculated and must instead be standardized. In this study, the indicators were scored and standardized by referencing the national standards and evaluation guidelines of related industries. The standardized indicators and calculation method of parameters are presented in Appendix C .

2.3. Spatial Analysis

To guide the functional zoning of contaminated soil environment and identify the primary areas of soil contamination management, we require spatial analysis, risk regionalization, and a comprehensive risk partitioning method. In the current study, the human health risks were quantified by sampling, surveying, and analyzing the soil pollutants, soil environment, and the integrated status of the industrial and mining gathering areas. The assessment methods are described in Section 2.1 . The results were then spatially interpolated using the inverse distance weighted (IDW) method, implemented in the ArcGis 9.3 software environment (Spatial Analyst module, ESRI, Beijing, China).

Compared with the IDW interpolation, other commonly employed methods such as Kriging and Spline interpolation also have strong ability to predict the overall trend of soil pollution. However, in purpose of identifying of the polluted areas, it is necessary to require the interpolation method to predict the local feature of soil pollution. In industrial and mining gathering areas, the concentration of pollutants in soil showed a high spatial variability, but the local maxima of soil pollution (concentration or risk value) is likely to be smoothed out by Kriging or Spline interpolation. Therefore, to reserve the local maxima and minima of soil pollution in industrial and mining gathering areas, IDW interpolation is an appropriate choice. Moreover, relevant study [ 23 ] indicated that, according to the root mean square error (RMSE) for cross validation, although Kriging and Spline interpolation are more accurate than other methods, the interpolation results of soils in polluted area estimated by Kriging are significantly smaller than the results by actual statistical results. Therefore, as a measure of overall sample prediction accuracy, RMSE cannot describe the estimated error of local extreme values.

From the interpolation results, the spatial distribution maps of human health risk were derived. Again, using the spatial interpolation, the assessment results of the polluting factories were embedded in an integrated risk regionalization map. In this way, the soil environment risks and inherent risks of the polluting factories were combined into a comprehensively partitioned classification of the soil environmental risks in the study area.

2.4. Comprehensive Analysis

The classification of land use in industrial and mining gathering areas was performed based on landscape, distribution and characteristics of population, functional requirements of lands, and protection requirements of ecological sensitive targets. Different types of land are divided into four classes: industrial land, including industrial land and mining sites, storage land, supply facilities area and so on; agriculture land, including farmland, orchards, aquaculture bases and so on; residential land, including residential areas, living areas, culture and entertainment land, education and health land, business area and so on; conservation land, including nature conservation objectives, coastal waters, wetlands and so on.

The regionalization results of human health risk ( TCR value) and the land use class of the study area were incorporated into a matrix assessment method. This method uses the comprehensive risk classifications in Table 2 to evaluate the risk status of the industrial and mining gathering areas. From the classification results, risk management and control measures can be designated.

Classification method of soil integrated risk.

2.5. Site Description

The selected study area is a typical mining and industrial gathering area in the Binhai New Area, Tianjin, China, located southeast of Tianjin, China. The study area, shown in Figure 1 , mainly covers the southern region of this area. Established in 1994, the Binhai New Area has become an important industrial and economic center in Tianjin, one of China’s largest industrial cities. The area is also the third zone especially designated for industrial economy development in China [ 24 ]. Unfortunately, industrial economic expansion and progress of the mining industry has been accompanied by increased soil contamination (mainly heavy metals). The study area covers approximately 1200 km 2 and experiences a warm temperate, semihumid continental monsoon climate. Its average annual temperature and precipitation levels are 14 °C and 600 mm, respectively [ 25 ]. As noted in reports on National Major Function-oriented Zoning, China has invested heavily in developing this international port city as an eco-city and in enhancing the northern economic center of Tianjin. Owing to its rich reserves of oil and metal resources near Bohai Bay, the Binhai New Area is of significant strategic interest. The main industries are located in the northeastern and eastern parts of the region and include petrochemical, metallurgical, and mining industries. In particular, this area is becoming an important petrochemical industry base in northern China, as outlined in the Overall Plan for the New Town in Tianjin (2006–2020).

Location and scope of the study area.

The ecological environment of this area is extremely sensitive and fragile because it borders the river, sea, and land. The pollution problem is exacerbated by the uneven distributions of the residential and industrial regions. Heavy metals introduced to the soil by human activity have contaminated large portions of this area and its vicinities [ 26 , 27 , 28 ]. Specifically, rivers, farmlands, and coastal waters have been polluted to varying extents by heavy metals discharging into water bodies over long periods.

2.6. Sampling and Analysis

Soil samples were collected from the study area in 2013. Forty-six census points were selected by systematic random grid sampling, the grid spacing of census points was 3 km. These points were separated by soil type, topographic characteristics, and the distribution of their contamination sources by a grid laying method. Moreover, to fully reflect the impact of highly aggregated mining industries on the quality of the soil environment, 68 encrypted points were selected in densely mined areas and areas with industrial activity ( Figure 2 ), the grid spacing of encrypted points was 1 km. The large empty area in the sampling point map is occupied by a water reservoir.

Soil sampling bitmaps. Note: the blank area is the North Dagang Reservoir.

The selected monitoring targets were eight common heavy metals, namely, As, Cd, Cr, Cu, Ni, Pb, Zn, and Hg; two common pollutants of mining industries, Co and V; and eight organic pollutants including Pyrene, Carbon tetrachloride, Dichloroethane, 1,2-, Trichlorothane, 1,1,1-, Benzene, Ethylbenzene, Fluoranthene, and Xylenes. The soil samples were pretreated by air-drying at room temperature. Plant roots, organic residues, and visible intrusions were removed from the samples. Finally, the samples were crushed, ground, and passed through a 0.85-mm sieve. The samples were suspended in deionized water (1:2.5 v / v ), agitated for 1 h using a Sartorius PB-10 (Sartorius, Beijing, China), and subjected to pH measurements. The concentrations of the monitoring targets (except As and Hg) were determined by inductively coupled plasma-atomic emission spectrometry. The As and Hg concentrations were measured by atomic fluorescence spectrometry (AFS-2202, Haiguang Company, Beijing, China). Volatile and semivolatile organic compounds, polycyclic aromatic hydrocarbon substances, and other organic pollutants were monitored by gas chromatography mass spectrometry.

3.1. Overview of Soil Pollutants

Table 3 showed the basic situation of various pollutants in soil. The detection rate of nine kinds of heavy metals reached 100%, and the detection rate of Hg and eight kinds of organic pollutants was less than 10%, indicating that heavy metal pollution was more serious in soil environment of the study area. Moreover, coefficients of variation of nine kinds of heavy metals except Hg were relative low, indicating that the nine kinds of heavy metals were widely distributed in the study area and the content difference of the nine heavy metals at different sampling points were relatively small.

Overview of soil pollutants.

3.2. Human Health Risk

3.2.1. overview.

The TCR and THI values of heavy metal contaminants and organic pollutants were separately calculated, as described in Section 2.1 . The THI values of organic pollutants were below 1 at all sampling points, and the total TCR values were lower than 10 −6 . Therefore, the health hazards posed by organic pollutants in the study area were generally acceptable. The THI values of heavy metals were also below 1 at all sampling points, indicating that heavy metals pose acceptable non-carcinogenic risk. However, in the carcinogenic risk category, the TCR values of Cd, As, and Cr 6+ at many of the sampling points exceeded 10 −6 . The results confirmed that heavy metals are the most important risk factors in the study area, especially considering their bioaccumulative character and non-biodegradability. Therefore, heavy metal pollution should continue to be targeted in industrial and mining gathering areas.

3.2.2. Human Health Risk of Heavy Metals

The TCR ranges of Cd, Cr 6+ and As were 1.6 × 10 −6 to 3.8 × 10 −4 , 2.2 × 10 −8 to 8.3 × 10 −6 , and 6.0 × 10 −6 to 7.6 × 10 −3 , respectively, indicating that the carcinogenic risks of these three heavy metals exceeded the acceptable level by varying degrees. At 114 of the sampling points, the TCR of Cr 6+ was lower than 1 × 10 −5 , indicating that Cr 6+ poses a low carcinogenic risk. The TCR s of As and Cd exceeded 1 × 10 −4 at 47 and 37 of the sampling points, respectively. Therefore, these heavy metals pose high or extreme carcinogenic risk at more than 30% of the sampling points.

Figure 3 showed the spatial distribution of carcinogenic risks of Cd, As and Cr 6+ . It is clear that the carcinogenic risk of Cr 6+ is low to moderate across most of the study area, the carcinogenic risk of As is high to extreme, and the carcinogenic risk of Cd is moderate to high. The overall level of carcinogenic risk of As is high, indicating that As is the main contributor of carcinogenic risk in the study area. Extreme risk areas of As are mainly concentrated in the north and east of the reservoir. Similarly, high risk areas of Cd and moderate risk areas of Cr 6+ are mainly concentrated in the north and east of the reservoir, indicating that carcinogenic risk in such areas is more serious. Extreme risk areas of Cd are mainly concentrated in the northeast and southwest of the study area.

Spatial distribution map of the human health risks of three heavy metals posed by contaminated soil.

The spatial distribution of the integrated human health risk (carcinogenic), obtained by IDW interpolation of the calculated TCR values, is shown in Figure 4 . The human health risk regionalization results revealed the following: (1) The carcinogenic risk is moderate to high risk across most of the study area; low-risk sites are minimal. (2) Extreme risk areas are mainly concentrated in the northeast of the study area and south of the reservoir. (3) Looking at the land use patterns ( Figure 5 ), high-risk areas are found to be concentrated in the Sanjiaodi industrial area located northeast of the reservoir and in the oilfield industrial zone, with its surroundings (including residential areas). (4) Extreme risk areas are concentrated in the Dagang urban area in the northeast of the study area, the Guangang forest park in the northeast corner, and residential areas affiliated with the oilfield industrial zone. The industrial land uses of these three regions share several common characteristics; complex population composition, frequent living activities, proximity to conservation projects (reservoirs and forest parks), and high vulnerability of receptors. Risk management and pollution prevention in such areas, especially in residential lands, is essential.

Spatial distribution map of the human health risks posed by contaminated soil.

Land use of study area.

3.3. Inherent Risk of Polluting Factories

According to the Census of Pollution Sources conducted by local environmental protection departments and statistical agencies in 2010, 150 polluting factories exist in the study area. The offending industries include electrical power, metallurgy, non-ferrous metals, petrochemicals, manufacturing (brick, paper, and textiles), and a small number of food production industries. The inherent risk of these polluting factories was calculated by the abovementioned risk evaluation method.

The calculated risks of the 150 polluting factories are summarized in Table 4 . As shown in this table, none of the factories incurred no or low risk, indicating that the polluting factories pose a serious threat in the study area. Half (75) of the factories, especially those involved in smelting, forging, production and processing of non-ferrous and ferrous metals, and some petrochemical factories, posed moderate risk. These industries are the leading industries in typical industrial and mining gathering areas. However, some of the large-scale metal production and processing factories achieved low scores in environmental management system and risk prevention measures. This indicates that after 30 years of development and construction, some leading industries had gradually developed and improved their risk monitoring and prevention systems. On the other hand, a large proportion of the 150 factories were assessed as high-risk. Among the 69 factories in this category, the vast majority was brick production and thermoelectric industries; the remainder included chemical industrial factories and metal production and processing factories. Therefore, large state-owned factories remain important sources of pollution in the study area. However, note that a significant number of small-scale factories involved in property management, food production, and light industry were also high-risk. These results highlight the importance of reasonable risk control measures and risk prevention awareness. Among the six extreme risk factories, four were large state-owned petrochemical factories; the remaining two were the largest brick factory in the study area and a glass factory. The pollutant emission levels of these six factories were also high.

Overview of polluting factory risk in the study area.

3.4. Integrated Risk

The classification and spatial distribution of integrated risk of soil environment in the study area is presented in Figure 6 . The integrated risk of soil in the study area was moderate to high. Furthermore, in Dagang urban area that located in the northeast of the study area, living areas affiliated to oilfield industrial zone that located in the southeast of the reservoir, wetland and forest park in the northeast of the study area, the integrated risk of soil was high to extreme, indicating that the integrated risks of residential land and conservation land were relative high. In most areas in the west and south of the study area, the integrated risk of soil was generally moderate. In Sanjiaodi industrial area and industrial production area of oilfield industrial zone, the integrated risk of soil was generally low.

Spatial distribution map of the integrated risk.

More than half of the polluting factories were located in the northeast of the study area, and nearly 20 of these were located in residential areas, posing high to extreme threats to human health. Five of the polluting factories assessed as extreme risk were located in the Sanjiaodi industrial area; the remaining one was surrounded by several polluting factories of different sizes in the southwest of the study area. Polluting factories tended to be distributed throughout green spaces or agricultural lands in the western parts. Such occupancy of non-industrial land by factories typifies industrial and mining gathering areas throughout China, indicating unreasonable planning of land layout and a disordered distribution of factories in the initial stage of regional development. Polluting factories posing high and moderate risks were cluttered and many were located in non-industrial areas. In Figure 6 , we can see that most of the polluting factories are centralized in the three industrial zones, but some are scattered outside of these zones, indicating a need for rational planning and management. Finally, although few of the polluting factories posed extreme risk to human health, their presence around densely populated areas and conservation regions such as reservoirs and forest parks presents a high inherent risk.

4. Discussion

The results revealed that the integrated risk status of residential lands and conservation projects were relatively high and should be improved to ensure that residents are not exposed to contaminated soil and to ensure ecological security. The rational planning of industrial and mining factories is the most important measure against soil environmental risk. In addition, polluting factories near residential lands and conservation areas should maintain their inherent risk below moderate levels. To achieve this goal, they require more comprehensive risk management plans and control measures.

The areas situated northeast to the study area, where polluting factories are largely intermingled with the population and the land use is mainly residential, were assessed as high to extreme integrated risk. Due to the high population density, such areas should be treated as the primary targets for soil risk management and control. A buffer zone should be erected between the urban and industrial areas and the emissions of polluting factories should be monitored.

Polluting factories, especially industrial and mining factories with high or extreme levels of inherent risk, should be relocated to the three major industrial areas to centralize their management and control. Industrial and mining factories located in the southwest of the study area should be shut down or relocated; alternatively, a new industrial zone should be established in this region and the residents should migrate to other residential areas. The non-negligible risks posed by other factories, including brick, food production, papermaking, and property management companies, require special attention. Such factories share common features such as cluttered distribution, lack of proper planning, and poor management. However, the production conditions and requirements and product flow of these industries largely differs from those in the heavy industrial and mining category. Therefore, relocating these factories to the three major industrial zones is neither practical nor wise. One possible solution is to establish specialized industrial areas for light industry and food production and processing in suitable locations. Moreover, these factories require guidance and supervision of their production processes, pollution monitoring systems, and environmental management measures.

From the distribution of industrial land and polluting factories in the southwest of the study area, the staggered distribution of industrial and residential lands was a prominent feature of China’s old industrial areas, and may have derived from the expansion of residential areas and functional areas of factory employees. Such intermingling reflects a lack of overall planning and long-term consideration of risk prevention. To protect residential lands, the long-term planning of industrial and mining gathering areas should separate residential and industrial lands as much as possible. Efficient transportation systems and risk isolation measures would ensure the normal operation of industrial and mining factories without posing risks to the nearby inhabitants.

5. Conclusions

This study established a method for assessing the soil environment risk in industrial and mining gathering areas. To this end, the pollutants and their sources were monitored and investigated. Moreover, the soil environmental risks in a typical industrial and mining gathering area were systematically analyzed. The main contributions of the study are summarized below.

(1) To assess the impacts and damage to human health by soil environmental pollution, a human health risk of heavy metal contaminants and organic pollutants was conducted. Similar to previous studies, heavy metals were identified as the most serious contaminants in the study area. High and extreme risk was found mainly in industrial and residential areas.

(2) The inherent risk level of polluting factories, which pose the main risks in industrial and mining gathering areas, was evaluated. The evaluation system was designed to optimize the layout of the regional environmental risk sources while protecting the residential population and the most sensitive conservation targets.

(3) A comprehensive analysis of soil environmental risk was conducted using a matrix overlay. By this method, the integrated risk in a typical industrial and mining gathering area was assessed. The integrated risk includes the risk level of the soil environment and inherent risk level of the polluting factories.

In industrial and mining gathering areas, the theories and methods of risk assessment and management of the regional soil environment remain at the developmental stage. In particular, the spatial and temporal zoning of environmental risk, multi-risk coupling and risk-field superimposition, and the allocation capacities of regional environmental risk are still being explored. Industrial and mining gathering areas have already implemented technologies and management systems to alleviate their integrated risk to the soil environment. However, further tests, optimization, and upgraded and improved application practices are needed, which should be based on the investigation and evaluation of risk sources.

Acknowledgments

The study was supported by “Funding Project of Environmental Nonprofit Industry Research and Special of China” (NO. 201309032) and “National Natural Science Foundation of China” (No. 41301579).

Appendix A. Assessment of Human Health Risk

Soil contaminants enter the human body mainly via the food chain, oral ingestion, skin contact, and breathing [ 12 , 29 ]. The health impacts of soil pollutants predominantly arise by the distribution and migration of soils, crops, and food. Therefore, in industrial and mining gathering areas, where soil environments are largely contaminated by heavy metals, the oral intake of such pollutants poses the most serious threat. Among the pollutants investigated in the current study, three heavy metals (Cr 6+ , Cd, and As) and seven organics (listed in Table A1 ) were identified as carcinogenic by the Integrated Risk Information System (IRIS) of US Environmental Protection Agency (EPA) [ 30 ] and the Technical Guidelines for Risk Assessment of Contaminated Sites (HJ 25.3-2014) [ 31 ] issued by China’s Ministry of Environmental Protection. The non-carcinogenic contaminants are ten heavy metals and eight organics (listed in Table A2 ).

The oral intakes of carcinogenic and non-carcinogenic pollutants in residential land are calculated by Equations (A1) and (A2), respectively:

Similarly, the oral intakes of carcinogenic and non-carcinogenic pollutants in farmland and industrial land are calculated by Equations (A3) and (A4), respectively:

In the above expressions, CDI car and CDI ncr , respectively, denote the oral intakes (mg/kg·d) of carcinogenic and non-carcinogenic pollutants from residential land, and CDI cafi and CDI ncfi are the corresponding intakes from farmland and industrial land, respectively. The subscripts a and c refer to adults and children, respectively. IR c and IR a denote the daily intake from soil (mg/d), and ED c and ED a are the periods of exposure duration (year) (the corresponding frequencies (d/year) are EF c and EF a ). BW c and BW a denote average body weights (kg), ABS o is the oral intake absorption efficiency factor, and AT ca and AT nc are the average times of the carcinogenic and non-carcinogenic effects, respectively (d). CS is the pollutant content in the soil (mg/kg), and CF is a conversion factor.

Total carcinogenic and non-carcinogenic risks of various soil pollutants are calculated by formulas (A5) and (A6), respectively:

In these formulas, TCR and THI denote the total carcinogenic and non-carcinogenic risks, respectively, CDI cai and CDI nci are the oral intakes (mg/kg·d) of a single pollutant and non-carcinogenic pollutant, respectively, i denotes an individual pollutant, SF oi is the carcinogenic slope factor of the i -th pollutant (kg·d/mg) (listed in Table A1 ), and RfD oi is the reference oral intake dose of the i -th pollutant (mg/kg·d) (listed in Table A2 ).

The parameter values in the above formulas are listed in Table A1 , Table A2 and Table A3 , the values were extracted from the Technical Guidelines for Risk Assessment of Contaminated Sites (HJ 25.3-2014) [ 31 ], issued by China’s Ministry of Environmental Protection.

Carcinogenic slope factor (SF o ) of carcinogenic pollutants (From the Technical Guidelines for Risk Assessment of Contaminated Sites HJ 25.3-2014 ).

Reference oral intake dose (RfD o ) of non-carcinogenic pollutants (From the Technical Guidelines for Risk Assessment of Contaminated Sites HJ 25.3-2014 ).

Exposure assessment parameters (From the Technical Guidelines for Risk Assessment of Contaminated Sites HJ 25.3-2014 ).

In general, acceptable values of the total carcinogenic risk range from 1 × 10 −6 to 1 × 10 −4 , and the total non-carcinogenic risk value ( THI ) should not exceed 1. Therefore, on the basis of the calculation of the total carcinogenic risk, 1 × 10 −5 was selected as an acceptable level of carcinogenic risk in the present study. The adjusted valuation criteria for the TCR are as follows: TCR ≤ 1 × 10 −5 (low risk), 1 × 10 −5 < TCR ≤ 1 × 10 −4 (moderate risk), 1 × 10 −4 < TCR ≤ 5 × 10 −4 (high risk), and TCR > 5 × 10 −4 (extreme risk). The THI was divided into two risk levels: THI < 1 (no risk) and THI ≥ 1 (non-carcinogenic risk).

Appendix B. Evaluation Index System and Weight Distribution

The polluting enterprises in industrial and mining gathering areas significantly differ in type, pollutant emission characteristics, and risk supervision level. Therefore, the environmental risks also differ among enterprises. To comprehensively assess the inherent risk level of polluting enterprises, this study applies an evaluation index system comprising sudden risk, cumulative risk, and risk supervision. Sudden risk refers to the risks on soil environment and human health posed by unexpected environmental accidents, natural disasters, and other transient factors. The sudden risk might reflect the degree to which various factors of polluting enterprises affect the environment. These factors include design layout, technology, management skills, and quality of personnel. Cumulative risk refers to the potential damage to human health and ecological environments by long-term, non-accidental discharge of pollutants during human production and development activities. The cumulative risk could reflect the risk imposed on soils during normal operation of polluting enterprises. Risk supervision focuses on the production safety, elimination of hazards, and treatment of contaminants. Risk supervision reflects the ability of the polluting enterprise to effectively control and prevent risks, and indirectly reflects the enterprises’ handling of production processing, equipment, management, and safety hazards.

To comprehensively assess the risk of polluting enterprises, each index must be weighted to reflect its contribution to the evaluated risk. Indices are commonly weighted by their entropy, Delphi method, the analytic hierarchy process (AHP), or principal component analysis (PCA). The present study adopts a mix of qualitative and quantitative methods. Advice was sought from 20 experts (including five environmental regulators, five engineers specializing in environmental impact assessment, and ten professors engaged in soil environmental protection). The indicator weights of the three evaluations in the criteria layer were then distributed by the AHP and Delphi methods on the basis of the characteristics of the soil environment and the indicators’ individual contributions to the soil quality in the area. The judgment matrix was constructed as S = ( u ij ) m×n (A7):

where u ij denotes the importance of indicator i to indicator j .

The consistency of the judgment matrix can be tested through calculation. The consistency is acceptable when the consistency proportion < 0.1. The feature vector A = (w 1 ,w 2 ,…,w n ) indicates that the largest eigenvalue of the judgment matrix corresponds to the weight distribution. The weight distribution results are presented in Table A3 .

Appendix C. Standardized Indicators and Calculation Method

Grading standards of the assessment indicators.

In Table A4 , the inventory level Q of hazardous substances is determined by Equation (A8):

where Q i and qi denote the inventory and critical value, respectively, of hazardous substance i , and n is the number of different hazardous substances. The critical values of hazardous substances are listed in Enterprise Environmental Risk Level Assessment Method [ 32 ]. Industrial output value refers to the total value of the products of the industrial enterprises within a certain period. Industrial policy requirements are accessible through the Industrial Restructuring Catalog issued by the National Development and Reform Commission [ 33 ]. Soot denotes the smoke and dust generated by industrial production process that performed in industrial furnaces using gas, liquid and solid fuels. According to the National Industrial Soot Emission Standards (GB T9078-1996), soot is an important indicator to measure the pollutant emission level of industrial factories in China. Indicator values were acquired mainly from pollution census, completion acceptance reports of enterprises, statistical yearbooks, and departmental rules and regulations from China’s State Council.

The comprehensive risk levels of polluting enterprises in mining and industrial gathering areas were characterized by comprehensive risk indexing ( CRI ) on the basis of the weight distributions and classification standards. The CRI is calculated by Equation (A9):

where CRI represents the inherent risk level of an enterprise, and w i and I i represent the weight and score of indicator i , respectively. The CRI lies within [0, 3] and is divided into four ranks; a score of [0, 1] indicates a healthy state of all indicators and a low risk level, (1, 2] denotes an alert state with moderate risk level, (2, 2.5] indicates a poor state with high risk level, and (2.5, 3] suggests extreme risk to human and environmental health.

Author Contributions

Yang Guan, Chaofeng Shao, Qingbao Gu, Meiting Ju and Qian Zhang work together. Specifically, Chaofeng Shao brings the idea, provides insight for literature guidance and choice of evaluation model. Yang Guan conducts all simulations and interprets the results. Qingbao Gu, Meiting Ju and Qian Zhang involve in the thesis structure, and provides a large number of basic data that are necessary for doing an assessment.

Conflicts of Interest

The authors declare no conflict of interest.

Business Management Ideas

Project Report on Soil Pollution

ADVERTISEMENTS:

A project report on soil pollution. This project report will help you to learn about: 1. Introduction to Soil Pollution 2. Definition of Soil Pollution 3. Kinds 4. Causes 5. Sources 6. Effects 7. Diseases 8. Controls.

• Project Report on the Controls of Soil Pollution

Project Report # 1. Introduction to Soil Pollution:

Industry being a voracious consumer of natural resources brought in pollution of air, water and soil environment. Soil pollution usually originates from the development of industry, intensive agriculture associated with modern systems of cultivation like use of high analysis chemical fertilizers, use of pesticides, use of sewage, sludge, city composts and other industrial wastes etc.

The contaminants produced by industrial processes ‘reach the soil by three ways:

(a) Through the air—gaseous and particulate contaminants are emitted from chimneys and exhausts or are deflated from spoil heaps and are blown away by the wind ultimately to sink or be washed by rain to the soil environment.

(b) Through drainage system industrial effluents and water drainage from spoil and rubbish heaps either washes direct on to nearby fields or enters the local streams and rivers and ultimately enter into the soil, and

(c) Through direct mechanical or gravitational effects on the soil-solid materials like refuses from mines, animal slurries or sewage-sludge’s etc. can be dumped directly on to the agricultural-land and thereby creates soil pollution.

Once pollutants enter and are incorporated into the soil, their concentrations in soil are continuously increasing and accumulating as toxic to all forms of life like plants, micro­organisms, human beings etc.

Project Report # 2. Definition of Soil Pollution :

The word “soil” have been derived from lain word “solum” meaning upper crust of the earth. Soil is actually formed as a result of long term process of complex interaction leading to the production of a mineral matrix in close association with interstitial organic matter — living as well as dead. Soil is one of the most important ecological fac­tors. Soil is thus usually defined as “any part of earth’s crust in which plant root”.

Soil pollution is the reduction in the productivity of soil due to the presence of soil pollutants. Soil pollutants have an adverse effect on the physical, chemical and biological properties of the soil and reduce its productivity.

Project Report # 3. Kinds of Soil Pollution :

Soil gets polluted by a number of ways. The major kinds of soil pollution are:

i. Acidification:

The causes of acidifi­cation are both natural and anthropogenic as shown below.

Project Report # 5. Sources of Soil Pollution :

i. Industrial Wastes.

ii. Urban Wastes.

iii. Radioactive Pollutants.

iv. Agricultural Practices.

v. Chemical and Metallic Pollutants.

vi. Biological Agents.

i. Soil Pollution by Industrial Wastes:

Disposal of industrial waste is the major problem responsible for soil pollution. These industrial pollutants are mainly discharged from pulp and paper mills, chemical industries, oil refineries, sugar factories, tanneries, textiles, steel, distilleries, fertilizers, pesticide industries, coal and mineral mining indus­tries, metal processing industries, drugs, glass, cement, petroleum and engineering indus­tries etc.

It has been estimated that about 50% of the raw materials ultimately become waste products in industry and about 20% of these wastes are extremely deleterious. In United Kingdom, it has been reported that about 20 million tonnes of substances are disposed off in the soil as industrial waste.

With the advent of technology, newer types of industrial wastes are produced and de­posited in the land. These waste products are also tipped on soil, enhancing the extent of soil pollution. Thermal, atomic and electric power plants are also the villain to add pollut­ants to soil.

The furnaces of such industries generate ‘fly ash’ i.e. un-burnt brownish black substance which severely pollute air, water and soil. Many industrial effluents are either discharged into streams or dumped into the surrounding land.

Industrial wastes mainly consist of organic compounds along with inorganic complexes and non-biodegradable ma­terials. These pollutants affect and alter the chemical and biological properties of soil. As a result hazardous chemicals can enter into human food chain, the soil or water, disturb the biochemical process and finally lead to serious effects on living organisms.

Industrial Sludge’s :

Industrial sludge’s are even more dangerous than industrial solid wastes to dispose of tidily. The composition of industrial sludge’s vary enormously, the common boiler scale, for example, consists of calcium carbonate and flue gas sludge.

This flue gas desulphurization sludge (FGDS) is generated when calcium hydroxide or lime stone slurries are used to trap sulphur dioxide from escaping gases in coal fired power plants. These wastes also consist of calcium salts and several toxic volatile elements such as arsenic, selenium, mercury, lead and cadmium, which pose detrimental effects on the environment.

ii. Soil Pollution by Urban Wastes:

Urban wastes comprises both commercial and domestic wastes consisting of dried sludge of sewage. All the urban solid wastes are com­monly referred to as ‘refuse’. Soil wastes and refuse, particularly in urban areas contribute to soil pollution.

This refuse contains garbage and rubbish materials like plastics, glasses, metallic cans, fibres, papers, rubbers, street sweepings, fuel residues, leaves, containers, abandoned vehicles and other discarded manufactured products.

Recent reports indicate that in United Kingdom nearly 15 million tonnes of domestic sewage are disposed-off into the land. In United States also, each sunset sees a new moun­tain to be precise, 4,10,000 tonnes of solid wastes. New York itself throws out 25,000 tonnes of solid stink when none of the city’s fourteen landfills, in use of more than 20 years, can take any more. Across the Atlantic, the situation of refuse is also critical.

It is estimated that in India along, about 115 million of urban population produces nearly 15 million tonnes of solid wastes causing chronic pollution of land and water, In critically polluted cities like Mumbai, Kolkata, Kanpur and Chennai, about 750 boogies are used.

Delhi, which is the third most polluted city amongst 41 critically polluted cities, collects about 3000 tonnes of garbage from its streets every day, to be thrown into its five landfills, there by polluting the land areas.

Urban domestic wastes though disposed-off separately from the industrial wastes, can still be dangerous. This is so because they cannot be easily degraded. Over population and increasing consumption have totally changed the very complexion of domestic wastes into a complex mixture of food-remains, papers, plastic and many notorious chemicals.

Other items like paints and varnishes which we use to add colour and gloss to everyday life also add poison to be urban wastes posing soil pollution problems.

The leachates from dumping sites and disposal tanks of sewage mixed with industrial effluents and wastes are extremely harmful and toxic. Actually, the leachates that oozes out of the polluted soil, contain poison­ous gases along with the partly decomposed organic material especially food remnants, vegetables, toxic hydrocarbons and pathogenic microbes, many of which can be disease causing.

Pollution concentration in urban areas and unplanned industrial progress in and around these urban areas, have to a greater extent contributed to soil pollution problems in India.

About 12 crore population of India lives in cities while its six times more population lives in villages which dump their waste products into the soil, posing terrestrial and aquatic pollu­tion hazard. It is estimated that Rs.20 crore have been spent simply for burning, compositing and thermally decomposing the refuse of Mumbai and Kolkata only.

iii. Soil Pollution by Radioactive Pollutants:

Radioactive resulting from explo­sions of nuclear devices, atmospheric fall out from nuclear dust and radioactive wastes (pro­duced by nuclear testing laboratories and industries) penetrate the soil and accumulate there creating land pollution.

Radio nuclides of radium, thorium, uranium, isotopes of potassium (K-40) and carbon (C – 14) are very common in soil, rock, water and air. Explosion of Hydrogen weapons and cosmic radiations induce neutron-proton reactions by which nitro­gen (N-15) produces C-14. This CI 4 participates in the carbon metabolism of plants which is then introduced into animals and man.

Radioactive waste contain several radio nuclides such as strontium-90, iodine-129, cae­sium-137 and isotopes of iron which are most injurious. Sr-90 gets deposited in bones and tissues instead of Calcium.

Nuclear reactor produces waste containing Ruthenium-106, Iodine-131, Barium-140 and Lanthanium-140, Caesium-144 with Promethiem-144 along with the primary nuclides Sr-90 and Cs-137. These are also produced from nuclear fission.

Cs-137 has a half-life of 30 years while Sr-90 has half-life 28 years. Rain water carry Sr-90 and Cs-137 to be deposited on the soil where they are held firmly with the soil particles by electrostatic forces.

Soil erosion and heavy rains carry away the deposited Cs-137 and Sr-90 with the silt and clay. All these radionuclides deposited on the soil emit gamma radiations. Recently, it has been indicated that some plants such as lichen and mushroom can accu­mulate Cs-137 and other radio nuclides which concentrates in grazing animals.

iv. Soil Pollution by Agricultural Practices:

Modern agricultural practices pollute the soil to a large extent. Today with the advancing agro-technology, huge quantities of fertilizers, pesticides, herbicides, weedicides and soil conditioning agents are employed to increase the crop yield.

Many agricultural lands have now excessive amounts of plants and animals wastes which are posing soil pollution problems. Apart from these farm wastes, manure slurry, debris, soil erosion containing mostly inorganic chemicals are reported to cause soil pollution. USA alone produces about 18 million tonnes of agricultural wastes every year.

Some of the agents responsible for this pollution are as follows:

(a) Fertilizers:

Now a days agricultural practices rely heavily on artificial fertilizers, which generally contain one or more of the plant nutrients i.e., nitrogen, phosphorous and potassium. Critical pollution problems arise mainly from their excessive application rates. Although the fertilizers are used to fortify the soil, yet they also contaminate the soil with their impurities.

When the fertilizers are contaminated with other synthetic organic pollut­ants, the water present in the soil may also get polluted. Generally fertilizers are retained by the soil and crop efficiently but there are some possibilities for the nitrates to be washed out due to negligence appliances in applying fertilizers to arable lands particularly in a wet spring.

These nitrates causes several undesirable effects on the water quality of low land lakes or rivers creating numerous health hazards. Reports indicate that if phosphate and nitrate concentration exceeds one part and thirty parts per hundred million parts of water respectively, it results in eutrophication, chocking the whole stretch of aquatic ecosystem.

India utilizes 16 kg per hectare of fertilizers whereas the world average is 55 kg/ha. Recently, NCA have estimated the increased use of fertilizer from 2.8 million tonnes in 1976 to 6 million tonnes in 1984 and 9.7 million tonnes in 1995.

Phosphatic fertilizer consumption rose from 2.5 mt. in 1985 to 3.4 mt in 1990. However, it is not only the increasing utilization of fertilizers but also escalated production which creates soil pollution hazards.

The need of increased food production due to growing population density was emphasized since long, which consequently led to manipulations of land re­sources. Different kinds of pesticides used to control pests are causing a stress in the natural environment.

However scattered information’s all over India on paddy fields alone have indicated that there was 108% increase in the yield of TR-8′ and 195% in ‘TN-I’ varieties when grown under plant protection umbrella using some biocides.

But from 1950 onwards the production of numerous synthesized organic pesticides have completely changed the basis and strategy of pest control. With the increasing use of pesticides as part of the newly developing agro-technology it was realized that a single pesticide did not completely elimi­nate all the species of target pest.

As a result, the number of commercial pesticides has increased during the past 20 years and now their number might be over 1000, which in­cludes herbicides, fungicides, insecticides, rodenticides, nematicides, molluscicides and pesticide.

Among pesticides the most important are the chlorinated hydrocarbons e.g., D.D.T., B.H.C aldrin, endrin, dieldrin, ethion, lenthion, trithion dursban, dimethoae phosdrin and metasystox etc. The remanants of these pesticides may get absorbed by soil particles which may contaminate root crops grown in soils.

Unfortunately these pesticide residues coexist within biological system with other forms of life. The elimination of pests in the soil must inevitably produce change and disrupt the balanced natural cycles and food chains within natural ecosystems.

Residual herbicides which are applied to the soil at the time of seeding remain active for several weeks and prevent the growth of weeds in competition with the emerging germinating crop.

(c) Soil Conditioners:

Fumigants and Other Chemical Agents – In addition to the fertilizers, pesticides and biocides, soil conditioners and fumigants are also employed to the land system to increase and protect the soil fertility as well as to kill the hazardous insects. These chemical agents are reported to cause alterations in both agricultural soil areas.

They contain several toxic metals like lead, arsenic, cadmium, mercury and cobalt etc. which when applied to a land will accumulate on the soil permanently there by introducing these chemical components into growing crops.

However, researches are being carried out to synthesize pesticides of short lived degradable residues so that the persistence of the pesti­cide residues and their degraded product on soil, food and large crops may be reduced considerably.

(d) Farm Wastes:

Increasing population of cows, catties, pigs and poultries have resulted in considerable soil pollution. Buildings in which grazing animals are housed can be cleaned using water but the manure is also washed out and pollutes it. When these farm wastes are dumped into heaps, they may become a good breeding ground for insects and several nuisances may arise.

It has been reported that a cow produces as much organic waste as twenty people. While a pig as much as three people. Their faecal matter mainly consists of phosphates which in conjunction with nitrates cause numerous undesirable effects in the soil texture. Animal wastes contain several pathogenic bacteria and viruses which enter into plant metabolism and ultimate to man.

Even the cow dung burning is hazardous to health due to the presence of benzo-pyrene in smoke which is a cancer promoting chemical. By burning cow dung-nitrogen rich manure, which is so essential to our cultivation, is reduced to ashes.

This raw sewage has a high value of B.O.D. (200-ppm), C.O.D. (400 ppm) and nitrogen (40 ppm) while the feedlot runoff have 1000 ppm of B.O.D. 8000 ppm of C.O.D. and 700 ppm, of nitrogen.

So the animal wastes are difficult to treat and the methods applied to treat municipal wastes cannot be used for farm wastes. However, a few biological degradation methods are reported to be satisfactory in their control. Simpler methods have to be re­searched in future to eliminate the difficulties encountered in biological methods.

v. Soil Pollution by Chemical and Metallic Pollutants:

Number of industries including textiles, pesticides, paints, dyes, soap and synthetic detergents, tanneries, drug, batteries, cement, asbestos, rubber, petroleum, paper and pulp, sugar, steel, glass, electro­plating and metal industries pour their hazardous effluents in soil and water creating disas­trous effects on living organisms.

Some of the major industries and their pollutants in water:

(i) Synthetic chemical and fertilizers are a source of trace metals which are added to the soil either deliberately or as impurity. For example, As, Pb, and Cd are the common trace metals in rock phosphate, also occur in super phosphate fertilizers.

Toxic selenium species get less readily oxidised so it cannot be removed easily from soil by weathering. The selenite ion in iron rich soils forms insoluble basic ferric selenite or sorbs strongly on the iron oxides. In acidic soils of Hawaii, selenium content occurs as 20 ppm. But a very little is available to plants; while the rest Se is often found associated with sulphur.

(ii) Today various tract elements such as Fe, Co, Ni, Cu, Zn, Ba, Pb, V, Mn, Ni, As, Hg, Mo and silicon are being added to the soil in one or the other form. Mn and Fe oxides have a tendency to concentrate trace metals by isomorphous replacement of ions.

(iii) In many soils 50 to 100% of soil carbon is found complexes with clay containing organic and inor­ganic components which affect the soil texture, its fertility and stabilization of soil organic matter.

(iv) Presence of high levels of Na, Mg and K causes calcium deficiency in soil. Magnesium deficiency in soil have attributed to high concentration of Ca, Na and K which are added as artificial fertilizers,

(v) Excess of sulphur in soil may be absorbed in and may be involved in photosynthesis but if present at high levels pose lethal effects on crop produc­tion.

(vi) Metallic contaminates in soil are considered to be highly dangerous since they affect the production of atmospheric oxygen as well as living beings.

vi. Soil Pollution by Biological Agents:

Soil gets large quantities of human, animals and birds excreta which constitute the major source of land pollution by biological agents. Digested sewage sludge as well as heavy application of manures to soils without periodic leaching could cause chronic salt hazard to plants within a few years.

In addition to these excreta faulty sanitation, municipal garbage, waste water and wrong methods of agricultural practices also induce heavy soil pollution. Sludge’s do have faults as they contain enough live viruses and viable intestinal worms. In developing western countries, intestinal para­sites constitute the most serious soil pollution problems.

The pathogenic organisms that pollute the soil may be classified into three categories as follows:

(i) Pathogenic Organisms Occurring Naturally in Contaminated Soil :

Soil has its own distinctive flora and fauna i.e. it is inhibited by bacteria, fungi, algae, protozoans, actinomycetes, nematodes, rotifers, earthworms, fungi, molluscs and arthropods etc. These organisms are important agents in increasing or decreasing the soil fertility, in altering the physical texture of the soil and attacking roots of plants.

(ii) Pathogenic Organisms Excreted by Man:

Human excreta includes pathogens such a enteric bacteria and parasitic worms. These organisms are transmitted to the man by the consumption of vegetables or fruits which are grown in the contaminated soil or by direct contact with the contaminated soil.

Beside this, insanitary habits of a large number of people have resulted in the repetition of the cycle of infection with soil transmitted pathogens from man to soil and from soil to man.

(iii) Pathogenic Organism Excreted by Animals:

This category includes pathogenic bacteria and parasitic worms excreted by animals. The animals like earthworms, millipedes, isopodes, dipterous larvae, sloug, snails including higher animals carry fungal and bacterial spores.

The disease producing organisms are transmitted from animals to soil and then from soil to man. Thus biological agents are highly responsible for heavy of soils and crops by pathogens. Now numerous methods have been developed to control pathogens. However, specific treatments are necessary for the effective removal of pathogens sewage effluents required for irrigation purposes.

Soil Pollution by Soluble Salts :

Salt accumulation has been a perpetual problem of civilization in arid and semiarid regions. Today a number of industries discharge their particular pollutants in the form of calcium sulphate, calcium carbonate, FAO states that half of the irrigated farms in the world are damaged by soluble salts deposited in soil.

Even the scientists in their attempts to deter­mine water effluents (less water per unit of crop yield) have sometimes increased salt prob­lems when leaching is too little. All natural water systems contain dissolved mineral substances commonly referred as soluble salts. Some rain waters, far from coastal salt sprays, may be very low in the salt content. As water flow over and through soils, it picks up salt loads.

If water rapidly evaporates as it flows on the surfaces, it results in increasing the concentration of salt. It actually happened in Colorado River in Western United States.

The erosion of salts and return flow water with salts in them add to the increased load of salt. Deicer salts, salty wastes dumped in lakes, rivers or streams and sea sprays are all the chief sources of soluble salts in soil. Actually salts washed from one field ends up in ground water or river to be used by someone else, there by spreading pollution nuisance.

How to Solve Soluble Salt Problem :

Till now, no simple solution to environmental accumulation of soluble salts is known. The only uncontested respites are oceans which are already salty and a few salt basins. One such area is the originally dry Salt on Sea of California, now used to collect drainage.

Today a number of states already have some regulation on salt problems. The main worry is, if dumping is restricted, how is the leaching to be regulated and the sources of contaminating salts to be identified. More careful use of soil as a receptor of salt will be most urgent.

However, careful irrigation to avoid excess water application will allow precipitation of Ca, Mg, SiO 2 and bicarbonates. Some sulphates, carbonates and silica precipitate during drying cycles. These carbonates and silica do not re-dissolve in soils. In some waters having low sodium content but high percentage of calcium and magnesium as much as 60 to 80%, the soluble salts may get precipitated, There is a catch to this seeming light at the end of dark tunnel.

Both Na and K salts cause extensive soil dispersion and are highly corrosive to metals. Chloride is also toxic to plants in high concentration. Thus periodically, even these soils will need leaching and reclamation by removal of exchangeable sodium in soil.

Most dissolved inorganic chemicals in water are observed in soil solutions. These high concentrations of salts are extremely undesirable because they reduce or hinder plant growth, enhance corrosion of metals, make drinking water unpalatable and chronically interfere in several uses of water.

Salt pollution from agricultural run-off water is largely non-point pollution, that is pollution does not always derive from one source or point but from a combination of sources. An example is cited by lateral sewage flow and salt carried from fields in irrigation in waste water.

Salt Stress in Soil :

Saline soil containing excess of soluble salts (about 0.1 %) results in poor plants growth and productivity.

Sources of salts in soil include:

(ii) Rock, and

(iii) Human activi­ties such as industrial wastes, improper agro-techniques, municipal wastes and use of saline water for irrigation under high evaporative conditions etc.

Increasing salinity of soil threatens the civilization with ever reducing areas of normal soil for utilization of the crops. A soil is generally said to be saline if the electrical conduc­tivity (EC) of the saturated extract of soil is greater than 4m mhos/cm at 25°C.

The predomi­nant ions contributing to salinity are sodium and chloride, although irons such as calcium, magnesium, potassium, sulphate, borate and bicarbonate are significantly high in certain regions. This salinity status can be determined by measuring the EC of the soil solution using conductivity bridge. It is estimated that more than 20 million hectares of cultivable land is saline in India.

In saline soil, because of low osmotic potential in the root medium, plants have to develop and maintain lower osmotic potential in cells for water absorption. If the rate of water is too slow for entering in plant cells, it may cause growth reduction. Moreover, intracellular water deficit may occur due to high concentration of salts getting accumulated in apoplast.

Soil Pollution by Sugar Cane Trash in Field :

Agricultural land containing sugar cane trash is the nuclei of several pathogens, bacte­ria, viruses and other micro-organisms. Now restrictions on its burning, to reduce the extent of air pollution, seriously concern sugarcane growers who customarily burn tonnes of cane leaves in fields before harvest.

Project Report # 6. Effects of Soil Pollution :

Sewage and industrial effluents which pollute the soil ultimately affect human health. Various types of chemicals like acids, alkalis, pesticides, insecticides, weedicides, fungicides, heavy metals etc. in the industrial discharges affect soil fertility by causing changes in physical, chemical and biological properties.

Some of the persistent toxic chemicals inhibit the non-target organisms, soil flora and fauna and reduce soil productivity. These chemicals accumulate in food chain and ultimately affect human health. Indiscriminate use of pesticides specially is a matter of concern.

Sewage sludge has many types of pathogenic bacteria, viruses and intestinal worms which may cause various types of diseases. Decomposing organic matter in soil also produces toxic vapours.

Radioactive fallout on vegetation is the source of radio-isotopes which enter the food chain in the grazing animals. Some of these radio isotopes replace essential elements in the body and cause abnormalities e.g. strontium-90 instead of calcium gets deposited in the bones and tissues. The bones become brittle and prone to fracture. Radioisotopes which attach with the clay become a source of radiations in the environment.

Soil damage and environment degradation during surface mining is inevitable as veg­etation has to be removed and huge quantities of top soil and waste rocks are to be shifted to a new location. Mining leads to loss to grazing and fertile land, soil erosion from waste dumps, sedimentation or siltation, danger to aquatic life, damage to flora and fauna as well as water and soil pollution.

A recent estimate showed that in India about 20, 000 hectares of land has been de­graded from mining and another 55,000 hectares of fertile land was degraded to meet out requirement of bricks. Even open-cast coal mining also affects seriously 2,00,000 hectares of land area. It is reported that 73% of the blocks identified for exploration by CIL and Singe Reni coal-fields involve drilling in forest areas.

Mining have also resulted in displacing a large section of people from their resources base. Since the mines are mostly in forest area, they severely affect the symbiotic relation­ship existing between tribals and forests.

Mining activates cause ecological damage and affect natural bio-diversity leading to erosion of environmental richness. Mining would result in high evolution of carbon dioxide, enhancing greenhouse effect, acid rain, global warming and over all climatic changes.

Adopting New Techniques – Modified techniques from dig dump mining to continuous system has been adopted recently by western countries along with sequential technique. Promotion of acceptable substitutes and recycling of all metallic wastes will reduce the potential hazard and will help to achieve sustainability in the long run.

The methods are not only environmentally efficient, but cost effective also. The India Bureau of Mines, Forest Research Institute, CMPDI and universities are actively engaged to disseminate new appropriate and eco-friendly indigenous mining technologies.

Rehabilitation :

The rehabilitation strategy needs to be broad based and made interdisciplinary. Appropriate cost effective measure include: storage of top soil, selection of ecologically and socio-economically suitable species, loss of fertile land due to erosion, loss of water retentivity, improvement of hydrological regime, accelerating natural regeneration enlisting people’s participation support in afforestation, fuel wood conservation and social fencing etc.

The development plant should be blended with action under the Nation wasteland Development Programme. The plan should consider climate, rainfall pattern, soil texture, demand and supply of fodder, timber, population and other bio-mass needs.

Project Report # 7. Diseases Caused by Soil Pollution :

(i) Pathogenic soil bacteria are chronic disease carrier.

(ii) Soil proves the best medium for the growth of eggs, larvae and flies etc.

Project Report # 8. Control of Soil Pollution :

(i) Effluents should be properly treated before discharging them on the soil.

(ii) Solid wastes should be properly collected and disposed off by appropriate method.

(iii) From the wastes, recovery of useful products should be done.

(iv) Biodegradable organic waste should be used for generation of biogas.

(v) Cattle dung should be used for methane generation. Night-soil (human faeces) can also be used in the biogas plant to produce inflammable methane gas.

(vi) Microbial degradation of biodegradable substances is also one of the scientific approaches for reducing soil pollution.

(vii) Collection of waste.

(viii) Disposal of waste.

(ix) Recovery of resources: converting waste into biogas. Sanitation meeting the heat energy demand.

Releted Articles:

• Land Degradation: Causes and Control |Soil Pollution

Project Report , Environment , Pollutions , Soil Pollution , Project Report on Soil Pollution

Comments are closed.

Privacy overview.

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

• View all journals
• Explore content
• About the journal
• Publish with us
• Sign up for alerts
• Open access
• Published: 17 July 2024

Innovative graph neural network approach for predicting soil heavy metal pollution in the Pearl River Basin, China

• Yannan Zha 1 &
• Yao Yang 2 , 3  

Scientific Reports volume  14 , Article number:  16505 ( 2024 ) Cite this article

Metrics details

• Environmental sciences
• Natural hazards
• Risk factors
• Solid Earth sciences

Predicting soil heavy metal (HM) content is crucial for monitoring soil quality and ensuring ecological health. However, existing methods often neglect the spatial dependency of data. To address this gap, our study introduces a novel graph neural network (GNN) model, Multi-Scale Attention-based Graph Neural Network for Heavy Metal Prediction (MSA-GNN-HMP). The model integrates multi-scale graph convolutional network (MS-GCN) and attention-based GNN (AGNN) to capture spatial relationships. Using surface soil samples from the Pearl River Basin, we evaluate the MSA-GNN-HMP model against four other models. The experimental results show that the MSA-GNN-HMP model has the best predictive performance for Cd and Pb, with a coefficient of determination (R 2 ) of 0.841 for Cd and 0.886 for Pb, and the lowest mean absolute error (MAE) of 0.403 mg kg −1 for Cd and 0.670 mg kg −1 for Pb, as well as the lowest root mean square error (RMSE) of 0.563 mg kg −1 for Cd and 0.898 mg kg −1 for Pb. In feature importance analysis, latitude and longitude emerged as key factors influencing the heavy metal content. The spatial distribution prediction trend of heavy metal elements by different prediction methods is basically consistent, with the high-value areas of Cd and Pb respectively distributed in the northwest and northeast of the basin center. However, the MSA-GNN-HMP model demonstrates superior detail representation in spatial prediction. MSA-GNN-HMP model has excellent spatial information representation capabilities and can more accurately predict heavy metal content and spatial distribution, providing a new theoretical basis for monitoring, assessing, and managing soil pollution.

Introduction

Soil serves as the cradle of life, providing the fundamental sustenance and development platform for all living organisms. However, with the influence of human activities and natural processes, the issue of heavy metal contamination in soils has become increasingly critical. Heavy metals, non-biodegradable elements, pose potential long-term negative impacts on the environment and human health due to their accumulation in soils. Therefore, it is of significant importance to predict and analyze the distribution of heavy metals in soils, as well as the factors influencing it, for both environmental protection and human health.

The original content of heavy metals in soil is primarily influenced by natural factors such as parent material, climate, biological activity, and topography 1 . However, with the rapid development of industrialization and urbanization, human activities (including industrial and agricultural pollution, transportation pollution, and household waste pollution) have become significant factors affecting the content of heavy metals in soil 2 . This has led to the complex spatial heterogeneity of soil heavy metals, making the prediction of their spatial distribution more difficult 3 . Classic statistical methods can only describe the overall variability of soil heavy metals content and fail to illustrate their spatial distribution characteristics 4 . Geostatistical models, by studying the spatial distribution and variability of variables with certain randomness and structure through variogram, have been widely used in predicting the spatial distribution of soil heavy metals 5 . However, these models have smoothing effects and struggle to identify local outliers of soil heavy metals; they also fail to describe the impact of environmental variables on the spatial differentiation of soil heavy metals 6 . With the development of 3S(Sensing, System and Service) technology 7 , environmental variables such as topography, climate, remote sensing, land use, and socio-economic factors have become more accessible, and methods like multiple linear regression, regression Kriging, and geographically weighted regression have been extensively used in regional soil heavy metal spatial distribution predictions 8 , 9 , 10 . However, such methods struggle to reveal the complex nonlinear relationships between soil heavy metals and environmental variables 11 .

Machine learning models such as artificial neural networks 12 , 13 , 14 , 15 , support vector machines 16 and random forests 17 , 18 have gradually been used for soil heavy metal spatial prediction. Folorunso et al. reviewed research predicting soil quality based on machine learning and analyzed the composition and quality of the soil, the prediction of soil parameters, existing soil datasets, soil maps, the influence of soil nutrients on crop growth, and the status of soil information system research 19 . Yang et al. proposed a new optimal sampling algorithm capable of realistically enhancing insufficient soil properties using machine learning uncertainty prediction 20 . Sun et al. developed a coupled retrieval approach to quantify nickel (Ni) concentration in agricultural soil using spaceborne hyperspectral imagery. Incorporating machine learning, multi-scale discrete wavelet transform, and other techniques, their work serves as a robust reference for agricultural applications worldwide 21 . Li et al. introduced an adaptive weighted normalization coupled with a linear weighted network framework for detecting chromium in soils 22 . Yang et al. developed machine learning models to study the adsorption of six heavy metals in soil, addressing the inefficiencies of traditional methods 23 . Pyo et al. focused on leveraging visible and near-infrared spectroscopy (VNIRS) combined with deep learning to estimate heavy metal concentrations in soil 24 . A convolutional neural network (CNN) was utilized to estimate arsenic, copper, and lead concentrations, showing higher accuracy compared to artificial neural network (ANN) and random forest regression (RFR) models.

Despite existing machine learning models having certain advantages in dealing with non-linear issues, they face challenges when dealing with high-dimensional data and complex network structures 25 . The spatial heterogeneity of soil heavy metal distribution is manifested in two aspects: (1) spatial dependence, which suggests that nearby locations might have similar soil heavy metal content due to possibly similar environmental conditions and human activities 26 , and (2) spatial self-organization, indicating that soil heavy metal distribution might demonstrate specific spatial patterns mainly due to self-organizing mechanisms in soil formation and development 27 . Current research employing machine learning models often assumes data are independent and identically distributed 28 , overlooking the spatial dependence and self-organization of spatial data. Moreover, there is a notable gap in studies focused on regional pollutant risk predictions 29 , 30 , 31 . Graph Neural Networks (GNN) 32 , 33 are capable of directly handling network-structured data and capturing and utilizing information about spatial dependence and self-organization by learning the features of nodes and edges. Furthermore, GNNs can handle high-dimensional data and potentially reveal complex spatial patterns of soil heavy metal distribution through automatically learning data's inherent structure and patterns 34 . Therefore, GNNs offer a new possibility for handling the spatial heterogeneity of soil heavy metal distribution and complex geographical network structures.

The Pearl River Basin (PRB), a vital economic hub in southern China, faces significant environmental challenges due to rapid urbanization and diverse industrial activities. This region is marked by notable heavy metal contamination, a consequence of complex pollution pathways and varying parent material types. The PRB’s diverse industrial landscape and complex geology make it an ideal case study for understanding the intricate relationships influencing heavy metal distribution in soil. Accurate prediction of heavy metal distribution is crucial for environmental protection and human health, yet the spatial heterogeneity and complex interplay of factors have posed a considerable challenge. This study will utilize GNN to predict and analyze the distribution of heavy metals in the soil of the Pearl River Basin. By considering a variety of influencing factors, including geographical location, parent material type, pollution pathways, and types, we aim to accurately predict the distribution of heavy metals in the soil and analyze the main factors affecting it. The results of this study can contribute to a better understanding of the distribution patterns of heavy metals in soil and the main influencing factors, which hold significant importance for environmental protection and human health. At the same time, this study is the first to apply Graph Neural Networks to the prediction of heavy metal distribution, providing new insights and methods for further research on such issues.

The main contributions of this study are as follows.

Innovative Model Structure: We introduce an advanced MSA-GNN-HMP model that seamlessly integrates two pivotal components: MS-GCN and AGNN. Specifically, the MS-GCN component adeptly identifies multi-scale structural information within the graph, while the AGNN dynamically allocates weights to the neighboring nodes of each node. Experimental results validate the superior performance of our proposed model.

Deep Representation of Geographical Information: Our model offers a profound representation of geographical data, such as latitude and longitude, which is especially vital in predicting environmental issues like heavy metal concentrations.

In-depth Interpretation of Factors Affecting Soil Pb and Cd Concentrations: This research employs a data-driven approach to thoroughly analyze the various factors influencing Pb and Cd concentrations, providing mechanistic explanations that lay a solid theoretical foundation for soil heavy metal prevention.

Overview of the research area

This study collected 142 surface soil samples from the Pearl River Basin, a region strategically selected due to its significance as a mining zone, and the sampling locations are depicted in Fig.  1 . Additionally, the southern regions of China, particularly the Pearl River Basin, have been historically impacted by serious heavy metal contamination 35 making it a critical area for such research 36 . These soil samples are spread across five different types of parent material developed soils, including limestone, sandy shale, alluvial deposits, diluvial deposits, and granite 37 . These soil samples come from different degrees of pollution and geological background regions. For this study, the primary contaminants of interest are cadmium (Cd) and lead (Pb). The choice to focus on these metals is based on their well-documented toxicity, which is among the highest for heavy metals 38 . Cadmium and lead can pose serious health risks to both humans and the environment, and their presence in the soil has the potential to be bioaccumulated in crops, thus entering the human food chain 39 .

Schematic of pollution source and surrounding parent material and sample point distribution. Map generated using ArcGIS software (version 10.2). Geographic data was sourced from our field data collection (coordinates provided in the supplementary data file) and elevation information obtained from the National Geospatial Information website.

The quantity of the samples collected is determined based on the size of the farmland area in different survey regions, which includes 18 samples of limestone developed soil, 36 samples of sandy shale developed soil, 42 samples of alluvial developed soil, 22 samples of diluvial developed soil, and 24 samples of granite developed soil. Each region selected multiple subregions including geological background areas, polluted areas, and mildly polluted areas. There are multiple sampling units and sampling points within each subregion. Two soil samples were collected from each sampling point, and each sample is a mixed sample of five soil samples. In addition, 6 unpolluted soil samples were collected from the same type of parent material developed soil region more than 10 km away from the pollution source. All samples were immediately transported to the laboratory after collection, a portion of which was used for microbial biomass carbon and nitrogen analysis, and the other part was used to determine basic soil physicochemical properties and total heavy metals content.

Determination of basic soil properties and description of predictive variables

Soil pH was determined using a 2.5:1 soil water extract and pH meter; soil CEC was measured by the ammonium acetate exchange method 40 ; soil texture was determined by the siphon method. Soil total organic matter (SOC) content was determined using the wet oxidation-potassium dichromate oxidation method; soil amorphous iron (FeTamm), amorphous manganese (MnTamm), crystalline iron (FeDCB), and crystalline manganese (MnDCB) content was determined by the o-phenanthroline colorimetric method; Dissolved organic matter (DOM) in the soil was extracted with water (soil to water ratio = 5:1) and oscillated for 1 h. The DOC content was determined using a TOC analyzer (vario TOC, Elementar, Germany). The DOM content of the soil was measured using a UV–visible spectrophotometer (UV-2450, Shimadzu, Japan). The absorbance values at 250, 254, 365, 400, 436, 465, 600, and 665 nm were measured. The specific UV–visible photometer value at 254 nm (SUVA254, L·g-1·cm-1) and ΔlogK were calculated according to Eqs. (1) and (2) 41 .

\(A_{i}\) is the absorbance value of DOM at i nm, and DOC is the soluble organic carbon content in the soil solution (mg L −1 ). SUVA254 and ΔlogK usually indicate the degree of humification of aromaticity and DOM 42 . E2/E3, E2/E4, and E4/E6 represent the absorbance ratios at 250 nm and 365 nm, 254 nm and 436 nm, and 465 nm and 665 nm, respectively. The first two ratios reflect DOM, the third ratio reflects the degree of humification, and shows the aromatic DOM 43 . Total heavy metals in the soil (Cd, Pb) were digested using the HCl-HNO3-HClO4-HF method 44 . The digest was measured using a flame atomic absorption spectrometer (Hitachi Z-2300, Hitachi, Japan). Each soil sample was set up in triplicate, and a blank test was also conducted at the same time.

According to the data sampling and measurement conditions, this paper selects a total of 15 variables(features) such as latitude and longitude as shown in the Table 1 . Different geographical locations may have different climatic conditions, soil types, and human activities, and geographical location may affect the distribution of heavy metals; different types of pollution are the direct causes of heavy metal distribution and content; crops grown may affect the distribution of heavy metals in the soil, because different crops may have different absorption and accumulation capabilities for heavy metals in the soil; soil chemical properties such as organic matter content, pH value, iron and manganese content may affect the form and mobility of heavy metals in the soil, thereby affecting their distribution in the soil; the physical properties of the soil will affect the adsorption capacity for heavy metals, thereby affecting their distribution in the soil 45 . Based on the above 15 variables, this paper will predict the content of the two heavy metals Cd and Pb.

The exploration of the spatiotemporal correlations at various monitoring points plays a critical role in enhancing the accuracy of heavy metal distribution predictions. This spatiotemporal correlation stems mainly from the distribution of geographic locations and the influences of factors like parent materials, pollution pathways, and types on each monitoring point. Graph structured data is the optimal solution for modeling the correlation between the spatial distribution of each monitoring point and the distribution of heavy metals. Nodes in the graph are used to represent monitoring points, and edges represent the correlation between the points. On the foundation of constructing a graph structure for heavy metal distribution, this article uses a graph neural network to realize heavy metal distribution prediction.

Heavy metal distribution graph structure modeling

Under the influence of soil physicochemical properties and factors such as parent materials and pollution types, there exists a certain spatiotemporal correlation between monitoring points. This article models these monitoring points as graph-structured data, using nodes to represent each monitoring point and establishing connections between nodes to represent the potential correlation between the monitoring points.

Establishing connections between monitoring points is an essential part of constructing graph-structured data for heavy metal distribution. Unlike graph data such as social networks and transportation networks, which have clear node connection relationships, the construction of connection relationships between nodes within heavy metal distribution monitoring points relies on certain rules. Accurate node connection relationships are vital for effectively extracting features of graph signals. This section proposes different heavy metal distribution graph structure modeling methods under two scenarios: considering parent material distribution and not considering parent material distribution. These methods are used for subsequent prediction models.

Heavy metal distribution graph structure modeling considering geographic spatial distribution

Since the node connection relationships of most graph convolutional neural networks do not change with time, this article first outlines a method for modeling the graph structure of heavy metal distribution without considering parent material distribution.

When the spatial distribution of parent materials is not taken into account, the geographic location of a monitoring point is the primary determinant of its similarity in heavy metal distribution. Factors such as soil type, environmental temperature, and atmospheric conditions may differ significantly at different locations at the same time, thereby affecting the correlation of heavy metal distribution between monitoring points. Monitoring points that are closer together have a stronger correlation in heavy metal distribution due to similar environmental conditions and minor latitude and longitude differences; conversely, monitoring points farther apart have weaker correlations.

Therefore, this paper proposes a fixed graph structure modeling method that takes into account the distance between monitoring points. For any monitoring point, the number of its neighbor points is determined based on the distance from other points to this monitoring point. By establishing connection relationships between each monitoring point and its neighboring points, a graph structure for heavy metal distribution is ultimately formed.

For any two monitoring points v i and v j within the study area, let d ij represent the distance between nodes v i and v j , and M nb,i represents the number of neighboring points for v i . Firstly, the distances from all the monitoring points in the study area to v i are calculated, as follows:

The number of neighboring points is a key parameter that determines the density of the adjacency matrix 32 . To consider the correlation between different monitoring points' heavy metal concentration under the premise of ensuring the sparsity of the adjacency matrix, this paper calculates the distance of related monitoring points through the Haversine formula based on longitude and latitude and further defines the number of neighboring nodes M nb,i as follows:

That is, for each node v i , the number of neighboring nodes is the number of monitoring points within a distance d kilometers from it.

After obtaining the number of neighboring nodes M nb,i , the closest M nb,i nodes to node v i are chosen to form the neighboring node set \(N(v_{i} )\) . The connection relationship is established between each node v i and its neighboring point \(v_{j} ,v_{j} \in N\left( {v_{i} } \right)\) , i.e., \(e_{ij} \in E\) . If A is the preliminary adjacency matrix, then:

After completing the above steps, several connected components of the cluster graph structure are obtained. This paper uses the reachable matrix of the undirected graph to judge the graph's connectivity and realize the connection of connected components. If A is the adjacency matrix of graph G , one of the methods to calculate the reachable matrix \(M_{G}\) is as follows:

Let \(A_{i} = {\text{sgn}}_{M} \left( {\left( {A + In} \right)^{i} } \right)\) , calculate A 1 , A 2 ,… in sequence. If \(A_{k - 2} \ne A_{k - 1} = A_{k}\) , then \(M_{G} = A_{k}\) . The function \({\text{sgn}} M(X)\) applies the sign function to all elements of matrix X and outputs a homomorphic matrix Y , i.e.

The rank of the reachable matrix rank ( M G ) is the number of connected components. Each vector in the largest linearly independent group of this matrix corresponds to the nodes contained in each non-connected subgraph.

Taking the graph structure shown in Fig.  2 as an example, this graph contains 3 connected components. The calculated result of its reachable matrix is as follows:

Example of a non-connected graph.

In this matrix, rank ( M G ) = 3, and the three vectors \(x1 = [1,0,0,1,0,0,0,1]^{T}\) , \(x2 = [0,1,0,0,1,0,0,0]^{T}\) , and \(x3 = [0,0,1,0,0,1,1,0]^{T}[0,0,1,0,0,1,1,0]\) \(^{T}\) make up the maximum linearly independent group of this matrix, \(M_{G}\) , and correspond to the three connected components, with node sets \(V_{1} = \{ v_{1} ,v_{4} ,v_{8} \}\) , \(V_{2} = \{ v_{2} ,v_{5} \}\) , \(V_{3} = \{ v_{3} ,v_{6} ,v_{7} \}\) .

After identifying the number of connected components and the nodes they contain using the reachable matrix, we connect the nodes that are the shortest distance apart between non-connected components to ensure the connectivity of the entire regional cluster graph structure, and update the edge set E. Based on this, the weighted adjacency matrix of the fixed cluster graph structure, \(A_{w}\) , can be obtained. Its calculation method is as follows:

In this paper, we set θ  = 5. Figure  3 shows the modeling process of the fixed cluster graph structure based on the measurement point distance proposed in this paper.

Modeling process of fixed graph structure for measurement point clustering.

Parent material clustering analysis and sampling site environmental type classification

Soil parent material is the foundation for soil formation. Different soil parent materials can lead to different heavy metal contents and distributions within the soil, thereby leading to some correlations in heavy metal distribution across sampling sites. For these heavy metal distribution sampling sites, constructing an appropriate graph structure based on parent material distribution and geographic location of the sites holds significant value for accurately extracting temporal and spatial correlations of heavy metal distribution. Given the vast geographical range and complex parent material distribution of heavy metal distribution sampling sites in the Pearl River Basin, the characteristics of heavy metal distribution under different parent materials can vary. Therefore, this paper first uses a clustering algorithm to classify the parent material, then determines the environmental type by assessing the coverage of parent material at each site. This allows us to divide the entire heavy metal distribution into several subclusters, simplifying the process of graph structure modeling.

The clustering algorithm is one of the commonly used algorithms in data classification. Given the complexity of the parent material distribution, distance-based clustering algorithms, such as K-means, KNN, etc., may not be suitable for parent material area division. This paper chooses the DBSCAN clustering algorithm for parent material area division. This method does not require specifying the number of clusters beforehand, it has resistance to noise, it is insensitive to outliers, and can handle clusters of any shape and size.

In graph convolutional models, the connection relationships between nodes in the graph represent the transmission of information, and the features of connected nodes often tend to be consistent. For heavy metal distribution sampling sites, the characteristics of heavy metal distribution under different environmental conditions are varied. To differentiate these characteristics, we classify the environmental conditions at each site into three types based on the distribution of parent material around the site: organic-rich, mineral-rich, and mixed type.

For any given sampling site v i , we consider certain features of the site, such as location, latitude, longitude, and parent material. By analyzing these features, we classify the environmental conditions of the site. In this classification, we define an "Environmental Type Index (ETI)" that can be calculated based on the content of organic matter and mineral matter at the site. For example, if the organic content at a site is higher than a certain threshold, we classify it as "organic-rich"; if the mineral content is higher than a certain threshold, we classify it as "mineral-rich"; if both are not high, but both exist, we classify it as "mixed".

After determining the environmental status of each site within the cluster, we define the environmental type matrix M eit for use in subsequent weighted adjacency matrix calculations. The matrix is defined as follows:

Dynamic graph structure modeling of heavy metal distribution sampling sites considering parent material type

In the graph structure modeling of heavy metal distribution sampling sites considering parent material type, we first calculate the set of neighboring nodes for each sampling site based on geographic location in “ Heavy metal distribution graph structure modeling considering geographic spatial distribution ”, and form a preliminary adjacency matrix A . We then modify A based on the parent material type matrix from “ Parent material clustering analysis and sampling site environmental type classification ”, as follows:

where A' is the preliminary adjacency matrix considering parent material type, A is the preliminary adjacency matrix considering the geographic location of the sampling sites, M eit is the parent material type matrix defined in “ Parent material clustering analysis and sampling site environmental type classification ”, and ⊙ represents the Hadamard product. The specific meaning of this modification method is: for any node v i within the cluster, eliminate the nodes in the set of neighboring nodes N(v i ) that have different parent material types from v i , and establish the connection relationship between vi and the remaining nodes to get A' .

Let the set of edges corresponding to matrix A' be E' . After calculating the reachable matrix using A' , we connect the nodes that are closest to each other between different connected components after judging the connectivity of the graph, update E' , and finally form the graph structure of heavy metal distribution sampling sites considering the parent material type. Let A w be the corresponding weighted adjacency matrix, and its calculation method is as follows:

where \(W \in R^{N \times N}\) is the distance weight coefficient of the sampling points. The spatial position of the sampling point does not change, but W may change because the connection relationship E' may change:

Ee take θ  = 5. Figure  4 shows the modeling process of the graph structure of the sampling site cluster considering the parent material type. By analyzing the parent material type of each sampling site, we can establish a corresponding graph structure and calculate the weighted adjacency matrix to generate the weighted adjacency matrix.

Modeling process of sampling site cluster graph structure considering parent material type.

For future graph structure data, we first use the sampling site's parent material relative position prediction method based on the parametric equation to get the predicted position of each sampling site. Assume that the predicted position of sampling site v i is ( x i , y i ), analyze the parent material type of sampling site iv based on the distribution of parent material around the sampling site, thereby obtaining the parent material type matrix M eit , and further calculate the weighted adjacency matrix A w .

Following these steps, using the predicted position of each sampling site, according to the partition results of parent material, a dynamically changing sequence of adjacency matrices A 1 , A 2 , …, A n can be generated, achieving prediction of the graph structure.

Heavy metal distribution prediction based on attention graph convolution model

Graph Convolutional Neural Networks can effectively process graph structured information. However, traditional GCNs usually assume that all neighbor nodes have the same impact on the target node when processing node features. Such an assumption might not be applicable in many real-world scenarios. For example, in the task of predicting soil heavy metal distribution, different neighbor nodes (i.e., different geographical locations) might have different impacts on the target node (i.e., target geographical location), which depends on various factors such as their geographical distance, soil type, etc.

In contrast, the Attention-based Graph Neural Network (AGNN) can dynamically allocate weights for each neighbor node of a target node, reflecting the importance of the neighbor nodes to the target node. This mechanism can better capture complex relationships between nodes and therefore improve the predictive performance of the model. In predicting soil heavy metal distribution, the AGNN can dynamically assign weights according to the degree of influence different neighbor nodes (i.e., different geographical locations) have on the target node (i.e., target geographical location), and thus more accurately predict the heavy metal distribution at the target location.

This paper proposes the heavy metal distribution prediction model as shown in Fig.  5 . It should be noted that the graph depicted here is schematic, and the final figure can be found in Fig. S1 . The model combines the Multi-Scale Graph Convolutional Network (MS-GCN) and Attention-based Graph Neural Network (AGNN). We refer to it as MSA-GNN-HMP (Multi-Scale Attention-based Graph Neural Network for Heavy Metal Prediction).

Architecture of the MSA-GNN-HMP model.

Multi-scale graph convolutional network (MS-GCN) module.

In the Multi-Scale Graph Convolutional Network (MS-GCN), we use graph convolution operations to handle graph data. The graph convolution operation can be considered as a convolution operation on the structure of the graph, which can capture the local structural information of nodes in the graph. In MS-GCN, we have designed three parallel graph convolution layers with small, medium, and large receptive fields, respectively, to capture graph structural features at different scales.

Small receptive field graph convolution layer It is mainly used to capture the local structural information in the graph. We can set the size of the convolution kernel to 1, i.e., only consider the features of the node itself and its direct neighbors. The computation formula for the graph convolution operation is:

where A is the adjacency matrix of the graph, D is the degree matrix, H (0) is the input node feature matrix, W (1) is the weight matrix of the first graph convolution layer, and σ is the activation function.

Medium receptive field graph convolution layer This layer is mainly used to capture medium-scale structural information in the graph. We can set the size of the convolution kernel to 2, i.e., consider the features of the node itself and its two-hop neighbors. The computation formula for the graph convolution operation is:

where A 2 is the square of the adjacency matrix of the graph, representing the connection relationship between two-hop neighbors, and the meanings of the other symbols are the same as above.

Large receptive field graph convolution layer This layer is mainly used to capture global structural information in the graph. We can set the size of the convolution kernel to 3, i.e., consider the features of the node itself and its three-hop neighbors. The computation formula for the graph convolution operation is:

where A 3 is the cube of the adjacency matrix of the graph, representing the connection relationship between three-hop neighbors, and the meanings of the other symbols are the same as above.

The outputs of these three graph convolution layers will be merged into one feature map, then passed through an activation function (like ReLU) for a non-linear transformation. In this way, the MS-GCN can capture structural information from the graph at different scales.

Attention-based graph neural network (AGNN).

After MS-GCN, we designed an Attention-based Graph Neural Network (AGNN). In AGNN, we use a graph attention mechanism to assign weights to each neighbor node of a node, reflecting the importance of neighbor nodes to the target node. Specifically, AGNN can be implemented using the following steps:

Compute attention scores For each node in the graph, first calculate the attention scores with its neighbor nodes. The attention scores are computed based on the features of nodes and can be calculated using the following formula:

where hi and hj are the features of node i and node j , respectively, W is the weight matrix, a is the parameter of the attention mechanism, || represents concatenation, and LeakyReLU is the activation function.

Normalize attention scores In order to make the attention scores fall between 0 and 1, use the softmax function to normalize the attention scores:

where N ( i ) represents the set of neighbor nodes of node i .

Update node features Finally, use the normalized attention scores to assign weights to each neighbor node of a node, and then average the features of neighbor nodes by weight to obtain new node features:

AGNN can dynamically assign weights to each neighbor node of a node, reflecting the importance of neighbor nodes to the target node. This mechanism can enhance the graph neural network model's ability to represent geographical information (longitude and latitude).

Evaluation metrics

In order to provide a rigorous evaluation of the predictive performance for the distribution of heavy metals in the soil, this study applies a nested cross-validation (CV) strategy alongside independent training/testing splits to ensure a robust and unbiased model assessment. Specifically, a k-fold cross-validation strategy, commonly with k = 10, is embedded within another layer of cross-validation to optimize model parameters while preventing data leakage between the training and test sets during the assessment phase.

Moreover, to further enhance the reliability of our model evaluation, we conduct 100 independent training/testing splits. For each split, the dataset is randomly partitioned into training and test sets, and the model is trained and validated accordingly. This iterative process assists in safeguarding against potential biases or anomalies that may emerge from a single random data split, thereby offering a more stable and reliable performance estimate.

Within each training/testing split or fold in the CV, the model is trained, and subsequently, the predicted values are compared with the actual measured values, with prediction accuracy being assessed by calculating the Mean Absolute Error (MAE), Root Mean Squared Error (RMSE), and Coefficient of Determination ( R 2 ) across all folds:

where P i and O i are the predicted and actual values of the i -th sample, respectively; \(\overline{O}_{i}\) is the average of the actual values; and n is the number of samples in the test set.

Experiments and analysis

Experimental setup.

The experiments were conducted on a custom-built workstation equipped with an Intel(R) Core(TM) i9-10900K CPU (10 cores) running at 3.70 GHz, an NVIDIA GeForce RTX 3080 GPU. The software environment was anchored on Ubuntu 20.04 LTS, utilizing Python 3.8.5. For data manipulation and numerical operations, The MSA-GNN-HMP model specifically being implemented in the Deep Graph Library (DGL). During the training process, Model training was initiated with parameters set at a batch size of 8, 100 epochs, and an Adam optimizer with a learning rate of 0.001. Figure S2 shows the training curve of the model proposed in this paper, and it can be seen that the model converges around epoch 80.

Characteristics of heavy metal content in the soil of the Pearl River Basin

The descriptive statistics of the two soil heavy metals, Cd and Pb, in the study area are shown in Table S2 . The average content of Cd and Pb are 1.58 mg kg −1 and 105.83 mg kg −1 , respectively. The coefficient of variation can reflect the extent of the impact of environmental changes and human activities on the accumulation of heavy metals in the soil. In this data set, the coefficients of variation for Cd and Pb are 1.37% and 1.67%, respectively. This could indicate that the distribution of these two heavy metals in the soil of the Pearl River Basin is not uniform, and may be influenced by human activities to a certain extent. The skewness and kurtosis of the data can reflect the shape of the distribution. Here, the skewness of Cd and Pb are 4.16 and 2.92, respectively, both positively skewed, indicating that the content of both heavy metals is lower in most samples, but there are some samples with higher content, which may be due to severe pollution in some areas. At the same time, the kurtosis of Cd and Pb are 24.73 and 8.53, respectively, higher than the kurtosis of a normal distribution, indicating that the distribution of these two heavy metals is more concentrated around the average value, and there are more outliers.

Figure  6 is the histogram and box plot of the two heavy metals. In the histogram and box plot for "Cd", we can see that the data for "Cd" concentration is right-skewed, with some larger values. At the same time, the box plot shows many outliers, which are all values greater than 1.5 times the interquartile range from the upper quartile. In the histogram and box plot for "Pb", we can also see that the data for "Pb" concentration is right-skewed, with some larger values. Similarly, the box plot shows many outliers, which are all values greater than 1.5 times the interquartile range from the upper quartile. Overall, the distribution of the two heavy metals in the study area is extremely uneven.

Descriptive statistics of Cd and Pb.

Model prediction accuracy analysis

We compared the MSA-GNN-HMP with four other models, namely Support Vector Regression, Random Forest, Fully Connected Neural Network, Convolutional Neural Network, and Spatial Random Forests 45 , on the test set. During the training process, the training hyperparameters for the aforementioned comparison models can be found in Table S1 . The MAE, RMSE, and R 2 values of each model are shown in Fig.  7 .

Model evaluation metrics heatmap for predicting Cd ( a ) and Pb ( b ).

We observe a distinct advantage of the MSA-GNN-HMP model over other models for soil heavy metal (Cd and Pb) prediction. A comprehensive examination of the metrics Mean Absolute Error (MAE), Root Mean Square Error (RMSE), and Coefficient of Determination (R 2 ) further elucidates the superiority of the MSA-GNN-HMP model.

Concerning the MAE, a standard metric used to evaluate the prediction error, the MSA-GNN-HMP model demonstrated the smallest error in both Cd and Pb prediction. Specifically, the MSA-GNN-HMP model meticulously captures the spatial dependencies between different geographical locations through its attention-based Graph Neural Network (AGNN), enabling it to discern intricate patterns in the distribution of Cd and Pb. This nuanced understanding minimizes the prediction error, particularly when compared to models like Support Vector Regression, which might struggle with the spatial heterogeneity and non-linearity in the data.

Likewise, an analysis of the RMSE scores, another common gauge for prediction error, reveals that the MSA-GNN-HMP model surpasses other models in terms of accuracy for both Cd and Pb. It’s noteworthy that models like Random Forest and Fully Connected Neural Network, while being potent in various predictive tasks, might fall short in accurately mapping the spatial distribution due to their inability to inherently consider spatial relationships among data points, which is a crucial aspect in predicting heavy metal concentrations in soil.

Further reinforcing the model's supremacy, the R 2 values, indicative of the accuracy of prediction, were highest for the MSA-GNN-HMP model for both Cd and Pb. This suggests that the model not only offers precision but also provides an exceptional fit to the observed data.

Spatial Random Forests (SRF) is a variant of the traditional Random Forests model, which incorporates spatial autocorrelation into the model by considering the spatial relationship between observations. Spatial Random Forests exhibited commendable predictive accuracy, particularly with an MAE and RMSE of 0.460, 0.710 for Cd, and 0.720, 0.960 for Pb, respectively. While the model does leverage spatial dependencies, which is a critical factor in predicting soil heavy metal distribution. However SRF performed worse compared to the MSA-GNN-HMP model.

The MSA-GNN-HMP model, leveraging Multi-Scale Graph Convolutional Networks (MS-GCN) and attention-based Graph Neural Networks (AGNN), effectively captures geographical information and the influence of neighboring nodes on the target node. This renders superior predictive results for soil heavy metal distribution, underscoring the importance of harnessing geographical information and considering spatial relationships when addressing such issues. Furthermore, this validates the potential of Graph Neural Networks for handling graph-structured geographical information.

Ablation experiment

To validate the roles of the MS-GCN and AGNN parts in the MSA-GNN-HMP model, we conducted ablation experiments. We removed the MS-GCN and AGNN respectively and observed the changes in model performance. The experimental results are shown in Fig.  8 .

Ablation experiment results.

Firstly, we ablated the MS-GCN part, i.e., we only retained the AGNN part. In this case, the model only considered the importance of each node to its neighboring nodes, without capturing the different scale structure information in the graph. We found that, compared to the complete MSA-GNN-HMP model, removing MS-GCN resulted in an increase in MAE to 1.13, an increase in RMSE to 1.61, and a decrease in R 2 to 0.62. These results demonstrate that the MS-GCN part plays a crucial role in the model. It effectively captures different scale structure information in the graph, enhancing the predictive performance of the model.

Next, we ablated the AGNN part, i.e., we only retained the MS-GCN part. In this case, the model only considered the different scale structure information in the graph without weighting the importance of each node to its neighboring nodes. We found that, compared to the complete MSA-GNN-HMP model, removing AGNN increased the MAE to 0.94, RMSE to 1.83, and decreased R 2 to 0.59. These results suggest that the AGNN part also plays a significant role in the model. It dynamically assigns weights to the neighboring nodes of each node, enhancing the model's representation ability for geographical information (latitude and longitude).

Through the ablation experiments, we can see that both the MS-GCN and AGNN parts play crucial roles in the MSA-GNN-HMP model, jointly improving the predictive performance of the model. This validates our original intention of designing the MSA-GNN-HMP model: to combine MS-GCN and AGNN, capturing different scale structure information in the graph, and dynamically assigning weights to the neighboring nodes of each node, to enhance the predictive accuracy of the model.

Factor importance analysis

We examined the significance of features in predicting heavy metal concentrations of Cd and Pb, as shown in Figs.  9 and 10 . The Fig. S3 a,b illustrate the correlation and fitting between the top six significant features and the content of Cd and Pb, respectively.Notably, for both Cd and Pb, "Type of Pollution" emerged as the most critical feature with weights of 0.3904 and 0.4126, respectively. This implies that the source of pollution plays a pivotal role in determining the levels of Cd and Pb in the soil. Latitude ranks second in feature importance for both metals, with weights of 0.2192 and 0.2472, indicating that geographic location also significantly influences the soil concentrations of Cd and Pb.

Feature importance analysis. ( a ) Feature Importance for Predicting Cd. ( b ) Feature Importance for Predicting Pb.

Comparison of spatial prediction of soil heavy metals in the Pearl River Basin by MSA-GNN-HMP and CNN.

Latitude ranks second in feature importance for both metals, with weights of 0.2192 and 0.2472, indicating that geographic location also significantly influences the soil concentrations of Cd and Pb. In our study, latitude and longitude play a key role in reflecting the multiple factors influencing the distribution of heavy metals in soils. They serve as proxies for climatic conditions, soil properties, and human activities. For instance, climatic conditions affect soil pH and thus heavy metal availability, with these changes being reflected by different latitudes and longitudes 46 , 47 . The latitude and longitude in our graph neural network also indicate the distance from the Pearl River Basin, where heavy metal content varies with the migration and change of surface runoff. Heavy metals migrate laterally with surface runoff, expanding the contamination range and potentially reducing their content due to deposition 48 . In our study area, rivers flow from west to east and north to south, representing a "lateral migration process" from upstream to downstream, mirrored by changes in latitude and longitude. Incorporating latitude and longitude into our model comprehensively reflects influencing factors and accurately predicts heavy metal distribution, offering a new perspective on the complex mechanisms of soil heavy metal distribution.

Other features, such as Organic Matter, pH, and CEC, vary in their importance between the two metals but still exert noticeable impacts. Some features, like Sand Percentage, hold relatively lower importance for Cd (0.0016), yet they remain noteworthy.

Analysis of heavy metal distribution prediction

As shown in Fig.  10 , the spatial distribution trends of the same soil heavy metal elements Pb and Cd predicted by the two different prediction methods, MSA-GNN-HMP and CNN, are largely consistent. In the predictions made by the MSA-GNN-HMP model, the high value areas of Cd are mainly distributed in the northwestern part of the sample area, while the high value areas of Pb are mainly distributed in the northeastern part. The MSA-GNN-HMP model is superior to the CNN model in predicting the low-value areas, reflecting more clearly the local variation of soil Pb and Cd content. The MSA-GNN-HMP model is better able to capture the microscopic distribution features of both heavy metals (Pb and Cd), and the prediction bias for high and low values is significantly reduced.

This study provides a detailed analysis of the distribution characteristics of the two heavy metals, Cd and Pb, in the soils of the Pearl River Basin, highlighting their uneven distribution and potential anthropogenic impacts. Statistical analysis shows that these two heavy metals have relatively low content in most samples, but there are some samples with high content, suggesting potential varying levels of soil contamination. Due to the harmfulness of these two heavy metals in the environment, such research is significant for understanding the degree of soil contamination and formulating effective remediation strategies.

In the comparison of prediction models, the MSA-GNN-HMP model demonstrated superior performance, outperforming other models in predictive accuracy and fitting results for soil heavy metal content. This is attributed to the roles of the MS-GCN and AGNN parts in the MSA-GNN-HMP model, with the MS-GCN part effectively capturing different scale structural information in the graph, and the AGNN part enhancing the model's representation ability for geographical information (latitude and longitude) through dynamic weight allocation. These results affirm the original intention and effectiveness of the MSA-GNN-HMP model design. Further ablation experiments show that both the MS-GCN and AGNN parts play important roles in the MSA-GNN-HMP model. Removal of either part leads to a decline in model performance, further affirming their importance within the model and validating the necessity of this model design.

Although our study provides an innovative model to predict soil heavy metal distribution, we acknowledge that there are certain uncertainties, particularly at finer spatial scales. The MSA-GNN-HMP model, while effective at a broader scale, may not fully capture the local variations in areas with intricate spatial features. Future research should integrate higher resolution data and additional spatial analysis methods to improve the model's precision and reliability 49 , 50 .

Our findings indicate that latitude and longitude significantly influence soil heavy metal distribution. These effects are indirect, as latitude and longitude serve as proxies for a complex set of environmental variables, including climatic conditions, soil properties, and human activities. For instance, climatic conditions affect soil pH and thus the availability of heavy metals, with these changes being reflected by different latitudes and longitudes 46 . The latitude and longitude in our graph neural network also indicate the distance from the Pearl River Basin, where heavy metal content varies with the migration and change of surface runoff. Heavy metals migrate laterally with surface runoff, expanding the contamination range and potentially reducing their content due to deposition 48 . In our study area, rivers flow from west to east and north to south, representing a "lateral migration process" from upstream to downstream, mirrored by changes in latitude and longitude. Incorporating latitude and longitude into our model comprehensively reflects influencing factors and accurately predicts heavy metal distribution, offering a new perspective on the complex mechanisms of soil heavy metal distribution.

In our study, we closely examined the key factors affecting the levels of Cd and Pb in the soil. For Cd, the type of pollution stands out as the most influential factor. This suggests that some pollution sources, like specific industrial activities, are direct contributors of Cd to the environment. Latitude also plays an important role, influencing how organic materials in the soil break down. These organic materials can combine with Cd, potentially affecting how it behaves in the soil and its availability to plants. Longitude gives us insights into how different areas, with varied land uses and industrial activities, might have differing amounts of Cd in the soil. Organic matter, known for its ability to bind with metals, can interact with Cd, impacting its movement and availability in the soil. The behavior of Pb in soil shows some differences compared to Cd. The type of pollution remains a dominant factor for Pb levels, highlighting the role of external sources. Latitude's impact might be linked to specific soil characteristics and plant types in an area, which can influence the form and availability of Pb in the soil. The relationship between soil pH and Pb is well-known: a more acidic soil can increase Pb's solubility, making it easier for plants to absorb or for it to seep into groundwater. Finally, just like with Cd, longitude and organic matter also play roles in determining Pb levels in the soil.

In the spatial distribution map of heavy metal distribution prediction, we can see that the results predicted by the MSA-GNN-HMP model are fairly consistent with the actual distribution. Particularly in areas with low heavy metal content, the MSA-GNN-HMP model can more clearly reflect their local variations. This indicates that the MSA-GNN-HMP model is not only superior in prediction accuracy compared to other models but also more accurate in capturing the microscopic distribution features of soil heavy metals. Figure S4 demonstrates the application of our MSA-GNN-HMP model in conducting a probabilistic risk assessment for heavy metal contamination in the Pearl River Basin. This Monte Carlo simulation provides a risk-based perspective essential for guiding environmental restoration and management actions. It showcases the model’s capability in offering a spectrum of risk scenarios, enabling targeted interventions and strategic planning for environmental protection.

In summary, this study demonstrates that the MSA-GNN-HMP model is a highly effective tool for predicting soil heavy metal content, aiding us in better understanding the distribution of heavy metals in the soils of the Pearl River Basin.

This study has delineated the distribution patterns of heavy metals, specifically cadmium (Cd) and lead (Pb), within the soil samples of the Pearl River Basin. The analysis revealed notable disparities in the distribution of these heavy metals, with average concentrations of 1.58 mg kg −1 for Cd and 105.83 mg kg −1 for Pb, suggesting an uneven distribution that could be attributed to human activities and accumulation.

Model prediction accuracy analysis showed that the MSA-GNN-HMP model performed well in predicting the spatial distribution of Cd and Pb in soil. The model had the lowest mean absolute error (MAE) and root mean square error (RMSE) and the highest coefficient of determination (R 2 ), indicating its excellent ability to predict the distribution of these heavy metals.

In addition, the study also demonstrated that the MSA-GNN-HMP model integrates MS-GCN and AGNN components, which can effectively capture the spatial heterogeneity of heavy metal distributions.The MS-GCN module efficiently extracts the multi-scale structural information from the graph, which improves the prediction performance of the model, while the AGNN component dynamically assigns weights to neighbouring nodes, which enhances the model's ability to represent geographic information, such as latitude and longitude. capability. These findings confirm the design principle of the MSA-GNN-HMP model, which combines MS-GCN and AGNN to improve prediction accuracy.

This study has fundamentally revealed the distribution characteristics and potential risks of heavy metals in soils near heavy industrial areas in the Pearl River Basin. The results of the study indicate the urgency of taking preventive measures and provide a strong scientific basis for local governments to formulate targeted risk control and pollution management strategies.

Data availability

All data generated or analyzed during this study are included in this manuscript.

Wang, Y., Duan, X. & Wang, L. Spatial distribution and source analysis of heavy metals in soils influenced by industrial enterprise distribution: Case study in Jiangsu Province. Sci. Total Environ. 710 , 134953 (2020).

Article   CAS   PubMed   Google Scholar  

Yuan, X., Xue, N. & Han, Z. A meta-analysis of heavy metals pollution in farmland and urban soils in China over the past 20 years. J. Environ. Sci. 101 , 217–226 (2021).

Article   CAS   Google Scholar  

Li, X., et al . (2021). Quantifying the influencing factors and multi-factor interactions affecting cadmium accumulation in limestone-derived agricultural soil using random forest (RF) approach. Ecotoxicol. Environ. Safety . 209 .

Li, J. et al. Analysis and assessment on heavy metal sources in the coastal soils developed from alluvial deposits using multivariate statistical methods. J. Hazard. Mater. 2 , 164 (2009).

Google Scholar  

Sun, C. Y. et al. Multivariate and geostatistical analyses of the spatial distribution and sources of heavy metals in agricultural soil in Dehui, Northeast China. Chemosphere. 92 (5), 517–523 (2013).

Article   ADS   CAS   PubMed   Google Scholar  

Chen, D., et al . (2023). Delineating and identifying risk zones of soil heavy metal pollution in an industrialized region using machine learning. Environ. Pollution . 318 .

Cao, S. et al. Modeling and mapping of cadmium in soils based on qualitative and quantitative auxiliary variables in a cadmium contaminated area. Sci. Total Environ. 580 , 430–439 (2017).

Agyeman, P. C. et al. Trend analysis of global usage of digital soil mapping models in the prediction of potentially toxic elements in soil/sediments: A bibliometric review. Environ. Geochem. Health. 43 , 1715–1739 (2021).

Yang, L., et al. (2022). Elucidating the spatial determinants of heavy metals pollution in different agricultural soils using geographically weighted regression. Sci. Total Environ . 853 .

Li, X. et al. Accumulation and source apportionment of heavy metal (loid) s in agricultural soils based on GIS, SOM and PMF: A case study in superposition areas of geochemical anomalies and zinc smelting, Southwest China. Process Safety Environ. Protection. 159 , 964–977 (2022).

Article   ADS   CAS   Google Scholar  

Jia, X., et al . (2021). VIRS based detection in combination with machine learning for mapping soil pollution. Environ. Pollution . 268 .

Wang, Z.-J., et al . (2020). BNNmix: A new approach for predicting the mixture toxicity of multiple components based on the back-propagation neural network. Sci. Total Environ . 738 .

Ucun Ozel, H. et al. Application of artificial neural networks to predict the heavy metal contamination in the Bartin River. Environ. Sci. Pollution Res. 27 , 42495–42512 (2020).

Mi, Y., et al . (2023). Machine learning method for predicting cadmium concentrations in rice near an active copper smelter based on chemical mass balance. Chemosphere . 319 .

Guo, X., & J. Wang. Projecting the sorption capacity of heavy metal ions onto microplastics in global aquatic environments using artificial neural networks. J. Hazardous Mater . 402 , 123709.

Bhagat, S.K., T.M. Tung, Z.M. Yaseen. Heavy metal contamination prediction using ensemble model: Case study of Bay sedimentation, Australia. J. Hazardous Mater . 403 , 123492.

Tan, K. et al. Random forest-based estimation of heavy metal concentration in agricultural soils with hyperspectral sensor data. Environ. Monitoring Assessment. 191 , 1–14 (2019).

Article   Google Scholar  

Jia, X., et al . (2020). Identification of the potential risk areas for soil heavy metal pollution based on the source-sink theory. J. Hazardous Mater . 393 .

Folorunso, O. et al. Exploring machine learning models for soil nutrient properties prediction: A systematic review. Big Data Cognitive Computing. 7 (2), 113 (2023).

Yang, H. et al. Simple optimal sampling algorithm to strengthen digital soil mapping using the spatial distribution of machine learning predictive uncertainty: A case study for field capacity prediction. Land 11 , 11 (2022).

Sun, Y., et al. (2023). Coupled retrieval of heavy metal nickel concentration in agricultural soil from spaceborne hyperspectral imagery. J. Hazardous Mater . 446 .

Li, X., et al . (2023). Chromium in soil detection using adaptive weighted normalization and linear weighted network framework for LIBS matrix effect reduction. J. Hazardous Mater . 448 .

Yang, H. et al. Predicting heavy metal adsorption on soil with machine learning and mapping global distribution of soil adsorption capacities. Environ. Sci. Technol. 55 , 20 (2021).

Pyo, J., et al . (2020). Estimation of heavy metals using deep neural network with visible and infrared spectroscopy of soil. Sci. Total Environ . 741 .

Lu, W., et al . (2022). Detection of heavy metals in vegetable soil based on THz spectroscopy. Comput. Electron. Agric . 197 .

Zhang, J., Zhang, Z. & Huang, X. Spatial heterogeneity of pH and heavy metal Cd in the soils of tea gardens in the plateau mountain regions, PR China. Environ. Monitoring Assessment. 193 , 1–12 (2021).

CAS   Google Scholar  

Gui, H., et al . (2023). Spatial distribution, contamination characteristics and ecological-health risk assessment of toxic heavy metals in soils near a smelting area. Environ. Res . 222 .

Wang, Qi. et al. Development of a new framework to estimate the environmental risk of heavy metal (loid) s focusing on the spatial heterogeneity of the industrial layout. Environ. Int. 147 , 106315 (2021).

Luo, H., et al . (2022). Heavy metal pollution levels, source apportionment and risk assessment in dust storms in key cities in Northwest China. J. Hazardous Mater . 422 .

Wang, J., et al . (2019). Source apportionment of heavy metal and their health risks in soil-dustfall-plant system nearby a typical non-ferrous metal mining area of Tongling, Eastern China. Environ. Pollut . 254 .

Li, Z. et al. A review of soil heavy metal pollution from mines in China: Pollution and health risk assessment. Sci. Total Environ. 468 , 843–853 (2014).

Article   ADS   PubMed   Google Scholar  

Meyer, H. & Pebesma, E. Machine learning-based global maps of ecological variables and the challenge of assessing them. Nat. Commun. 13 (1), 2208 (2022).

Article   ADS   CAS   PubMed   PubMed Central   Google Scholar  

Scarselli, F. et al. The graph neural network model. IEEE Trans. Neural Netw. 20 , 1 (2008).

Li, P., et al . (2022). A field study to estimate heavy metal concentrations in a soil-rice system: Application of graph neural networks. Sci. Total Environ . 832 .

Miranda, L.S., et al . (2021). Water-sediment interactions and mobility of heavy metals in aquatic environments. Water Res . 202 .

Wang, J., et al . (2023). Spatiotemporal characterization of heavy metal and antibiotics in the Pearl River Basin and pollutants removal assessment using invasive species-derived biochars. J. Hazardous Mater . 454 .

Lu, Q., et al . (2020). Effects of coastal reclamation history on heavy metals in different types of wetland soils in the Pearl River Delta: Levels, sources and ecological risks. J. Cleaner Production . 272 .

Chen, S.-Y., et al . (2021). Small molecular fluorescent probes for the detection of lead, cadmium and mercury ions. Coordination Chem. Rev . 429 .

Gan, Y. et al. Source quantification and potential risk of mercury, cadmium, arsenic, lead, and chromium in farmland soils of Yellow River Delta. J. Cleaner Production. 221 , 98–107 (2019).

Kong, X., et al . (2021). Comparison of highly-weathered acid soil CEC determined by NH4OAc (pH = 7.0) exchange method and BaCl2-MgSO4 forced-exchange method. Agric. Sci . 12 (9), 917–927.

Kumada, K. Chemistry of soil organic matter. Environ. Soil Chem. 22 , 53–80 (1987).

Ni, X. et al. Forest gaps slow the sequestration of soil organic matter: A humification experiment with six foliar litters in an alpine forest. Sci. Rep. 6 , 19744 (2016).

Helms, J.R. Spectral shape as an indicator of molecular weight in chromophoric dissolved organic matter. Old Dominion University, 2006.

Li, Y. T. et al. Ion activity and distribution of heavy metals in acid mine drainage polluted subtropical soils. Environ. Pollution. 157 , 1249–1257 (2009).

Fang, H., et al . (2021). Machine learning approaches for predicting the performance of stormwater biofilters in heavy metal removal and risk mitigation. Water Res . 200 .

Zhong, X. et al. Factors influencing heavy metal availability and risk assessment of soils at typical metal mines in Eastern China. J. Hazardous Mater. 400 , 123289. https://doi.org/10.1016/j.jhazmat.2020.123289 (2020).

Pan, Y., Chen, M., Wang, X., & Chen, Y. (2024). Ecological risk, source apportionment, and influencing factors of heavy metals in soil in a typical lead-zinc mining watershed, Guangxi, China. J. Environ. Chem. Eng .

Qiao, P., Wang, S., Li, J., Zhao, Q., Wei, Y., Lei, M., Yang, J., & Zhang, Z. (2022). Process, influencing factors, and simulation of the lateral transport of heavy metals in surface runoff in a mining area driven by rainfall: A review. Sci. Total Environ . 159119.

Yan, Y. & Yang, Y. Uncertainty assessment of spatiotemporal distribution and variation in regional soil heavy metals based on spatiotemporal sequential Gaussian simulation. Environ. Pollution (Barking, Essex: 1987). 322 , 121243. https://doi.org/10.1016/j.envpol.2023.121243 (2023).

Chen, M., Chen, Y. & Zhang, Q. Assessing global carbon sequestration and bioenergy potential from microalgae cultivation on marginal lands. https://ssrn.com/abstract=4789619 . https://doi.org/10.2139/ssrn.4789619

Download references

Acknowledgements

This research was financially supported by the Guangdong University VR/AR application integration engineering technology development center (2023GCZX013). Soil data should be acknowledged with thanks to the Key Laboratory of Arable Land Conservation (South China). All authors consent to publish this manuscript.

Author information

Authors and affiliations.

Guangzhou Institute of Technology, Guangzhou, Computer Simulation Research and Development Center, 465 Huanshi East Road, Guangzhou, 510075, China

Guangdong Provincial Key Laboratory of Agricultural & Rural Pollution Abatement and Environmental Safety, College of Natural Resources and Environment, Joint Institute for Environment & Education, South China Agricultural University, 483 Wushan St., Guangzhou, 510642, China

Key Laboratory of Arable Land Conservation (South China), Ministry of Agriculture, Guangzhou, 510642, China

You can also search for this author in PubMed   Google Scholar

Contributions

Yannan Zha: Conceptualization, Review & Editing, Funding acquisition, Software, Visualization, Review & Editing, Supervision, Project administration, Original Draft. Yao Yang: Methodology, Formal analysis, Investigation, Data Curation.

Corresponding author

Correspondence to Yannan Zha .

Ethics declarations

Competing interests.

The authors declare no competing interests.

Additional information

Publisher's note.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Supplementary information., rights and permissions.

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ .

Reprints and permissions

About this article

Cite this article.

Zha, Y., Yang, Y. Innovative graph neural network approach for predicting soil heavy metal pollution in the Pearl River Basin, China. Sci Rep 14 , 16505 (2024). https://doi.org/10.1038/s41598-024-67175-7

Download citation

Received : 10 May 2024

Accepted : 09 July 2024

Published : 17 July 2024

DOI : https://doi.org/10.1038/s41598-024-67175-7

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Heavy metal distribution prediction
• Graph neural network
• Soil pollution
• Geographic information

By submitting a comment you agree to abide by our Terms and Community Guidelines . If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.

Quick links

• Explore articles by subject
• Guide to authors
• Editorial policies

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Information

• Author Services

Initiatives

You are accessing a machine-readable page. In order to be human-readable, please install an RSS reader.

All articles published by MDPI are made immediately available worldwide under an open access license. No special permission is required to reuse all or part of the article published by MDPI, including figures and tables. For articles published under an open access Creative Common CC BY license, any part of the article may be reused without permission provided that the original article is clearly cited. For more information, please refer to https://www.mdpi.com/openaccess .

Feature papers represent the most advanced research with significant potential for high impact in the field. A Feature Paper should be a substantial original Article that involves several techniques or approaches, provides an outlook for future research directions and describes possible research applications.

Feature papers are submitted upon individual invitation or recommendation by the scientific editors and must receive positive feedback from the reviewers.

Editor’s Choice articles are based on recommendations by the scientific editors of MDPI journals from around the world. Editors select a small number of articles recently published in the journal that they believe will be particularly interesting to readers, or important in the respective research area. The aim is to provide a snapshot of some of the most exciting work published in the various research areas of the journal.

Original Submission Date Received: .

• Active Journals
• Find a Journal
• Proceedings Series
• For Authors
• For Reviewers
• For Editors
• For Librarians
• For Publishers
• For Societies
• For Conference Organizers
• Open Access Policy
• Institutional Open Access Program
• Special Issues Guidelines
• Editorial Process
• Research and Publication Ethics
• Article Processing Charges
• Testimonials
• Preprints.org
• SciProfiles
• Encyclopedia

Article Menu

• Subscribe SciFeed
• Recommended Articles
• Author Biographies
• Google Scholar
• on Google Scholar
• Table of Contents

Find support for a specific problem in the support section of our website.

Please let us know what you think of our products and services.

Visit our dedicated information section to learn more about MDPI.

JSmol Viewer

Advances in studies on heavy metals in urban soil: a bibliometric analysis.

1. Introduction

2. data sources and methodology, 2.1. research method, 2.2. data sources and search strategy, 3. results and discussion, 3.1. publication statistics, 3.2. core authors, 3.3. high-yielding journals, 3.4. high output countries and institutes, 3.5. highly cited document network analysis, 3.5.1. co-citation analysis of document, 3.5.2. document co-citation cluster analysis, 3.6. profiling of keywords in the paper, 3.6.1. urban soil heavy metal pollution source apportionment, 3.6.2. pollution risk assessment, 3.6.3. applications of environmental magnetism, 3.7. keyword bursting detection, 3.8. limitations of this study, 4. conclusions and outlook, author contributions, data availability statement, conflicts of interest.

• Kuźniar, A.; Banach, A.; Stępniewska, Z.; Frąc, M.; Oszust, K.; Gryta, A.; Kłos, M.; Wolińska, A. Community-level physiological profiles of microorganisms inhabiting soil contaminated with heavy metals. Int. Agrophys. 2018 , 32 , 101–109. [ Google Scholar ] [ CrossRef ]
• Xie, Y.; Fan, J.; Zhu, W.; Amombo, E.; Lou, Y.; Chen, L.; Fu, J. Effect of heavy metals pollution on soil microbial diversity and bermudagrass genetic variation. Front. Plant Sci. 2016 , 7 , 755. [ Google Scholar ] [ CrossRef ]
• Chen, J.; He, F.; Zhang, X.; Sun, X.; Zheng, J.; Zheng, J. Heavy metal pollution decreases microbial abundance, diversity and activity within particle-size fractions of a paddy soil. FEMS Microbiol. Ecol. 2014 , 87 , 164–181. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Oves, M.; Saghir, K.M. Heavy metals: Biological importance and detoxification strategies. J. Bioremediat. Biodegrad. 2016 , 7 , 1–15. [ Google Scholar ]
• Zaltauskaite, J.; Sliumpaite, I. Single and combined toxicity of copper and cadmium to H. vulgare growth and heavy metal bioaccumulation. E3S Web Conf. 2013 , 1 , 15013. [ Google Scholar ] [ CrossRef ]
• Li, Z.; Ma, Z.; van der Kuijp, T.J.; Yuan, Z.; Huang, L. A review of soil heavy metal pollution from mines in China: Pollution and health risk assessment. Sci. Total Environ. 2014 , 468 , 843–853. [ Google Scholar ] [ CrossRef ]
• Yadav, P.; Singh, B.; Garg, V.K.; Mor, S.; Pulhani, V. Bioaccumulation and health risks of heavy metals associated with consumption of rice grains from croplands in Northern India. Hum. Ecol. Risk Assess. 2017 , 23 , 14–27. [ Google Scholar ] [ CrossRef ]
• Yousaf, B.; Liu, G.; Wang, R.; Zia-ur-Rehman, M.; Rizwan, M.S.; Imtiaz, M.; Murtaza, G.; Shakoor, A. Investigating the potential influence of biochar and traditional organic amendments on the bioavailability and transfer of Cd in the soil-plant system. Environ. Earth Sci. 2016 , 75 , 374. [ Google Scholar ] [ CrossRef ]
• Liang, Y.; Yi, X.; Dang, Z.; Wang, Q.; Luo, H.; Tang, J. Heavy metal contamination and health risk assessment in the vicinity of a tailing pond in Guangdong, China. Int. J. Environ. Res. Public Health 2017 , 14 , 1557. [ Google Scholar ] [ CrossRef ]
• Silva, A.L.O.; Barrocas, P.R.G.; Jacob, S.C.; Moreira, J.C. Dietary intake and health effects of selected toxic elements. Braz. J. Plant Physiol. 2005 , 17 , 79–93. [ Google Scholar ] [ CrossRef ]
• Liu, X.; Song, Q.; Tang, Y.; Li, W.; Xu, J.; Wu, J.; Wang, F.; Brookes, P.C. Human health risk assessment of heavy metals in soil–vegetable system: A multi-medium analysis. Sci. Total Environ. 2013 , 463 , 530–540. [ Google Scholar ] [ CrossRef ]
• Tiller, K.G. Urban soil contamination in Australia. Aust. J. Soil. Res. 1992 , 30 , 937–957. [ Google Scholar ] [ CrossRef ]
• Napoletano, P.; Colombo, C.; Di Iorio, E.; Memoli, V.; Panico, S.C.; Ruggiero, A.G.; Santorufo, L.; Maisto, G.; De Marco, A. Integrated approach for quality assessment of technosols in experimental mesocosms. Sustainability 2021 , 13 , 9101. [ Google Scholar ] [ CrossRef ]
• Mielke, H.W.; Gonzales, C.R.; Smith, M.K.; Mielke, P.W. The urban environment and children’s health: Soils as an integrator of lead, zinc, and cadmium in New Orleans, Louisiana, USA. Environ. Res. 1999 , 81 , 117–129. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Kelly, J.; Thornton, I.; Simpson, P.R. Urban geochemistry: A study of the influence of anthropogenic activity on the heavy metal content of soils in traditionally industrial and non-industrial areas of Britain. Appl. Geochem. 1996 , 11 , 363–370. [ Google Scholar ] [ CrossRef ]
• Bi, X.; Zhang, M.; Wu, Y.; Fu, Z.; Sun, G.; Shang, L.; Li, Z.; Wang, P. Distribution patterns and sources of heavy metals in soils from an industry undeveloped city in Southern China. Ecotoxicol. Environ. Saf. 2020 , 205 , 111115. [ Google Scholar ] [ CrossRef ]
• Huang, S.; Wang, L.; Zhao, Y. Ecological risk assessment from the perspective of soil heavy metal accumulations in Xiamen city, China. Int. J. Sust. Dev. World 2018 , 25 , 411–419. [ Google Scholar ] [ CrossRef ]
• Pan, L.B.; Ma, J.; Wang, X.L.; Hou, H. Heavy metals in soils from a typical county in Shanxi Province, China: Levels, sources and spatial distribution. Chemosphere 2016 , 148 , 248–254. [ Google Scholar ] [ CrossRef ]
• Cao, S.; Duan, X.; Zhao, X.; Chen, Y.; Wang, B.; Sun, C.; Zheng, B.; Wei, F. Health risks of children’s cumulative and aggregative exposure to metals and metalloids in a typical urban environment in China. Chemosphere 2016 , 147 , 404–411. [ Google Scholar ] [ CrossRef ]
• Haefliger, P.; Mathieu-Nolf, M.; Lociciro, S.; Ndiaye, C.; Coly, M.; Diouf, A.; Faye, A.L.; Sow, A.; Tempowski, J.; Pronczuk, J.; et al. Mass lead intoxication from informal used lead-acid battery recycling in Dakar, Senegal. Environ. Health Perspect. 2009 , 117 , 1535–1540. [ Google Scholar ] [ CrossRef ]
• Roels, H.A.; Buchet, J.P.; Lauwerys, R.R.; Bruaux, P.; Claeysthoreau, F.; Lafontaine, A.; Verduyn, G. Exposure to lead by the oral and the pulmonary routes of children living in the vicinity of a primary lead smelter. Environ. Res. 1980 , 22 , 81–94. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Chen, T.B.; Zheng, Y.M.; Lei, M.; Huang, Z.C.; Wu, H.T.; Chen, H.; Fan, K.K.; Yu, K.; Wu, X.; Tian, Q.Z. Assessment of heavy metal pollution in surface soils of urban parks in Beijing, China. Chemosphere 2005 , 60 , 542–551. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Madrid, L.; Díaz-Barrientos, E.; Madrid, F. Distribution of heavy metal contents of urban soils in parks of Seville. Chemosphere 2002 , 49 , 1301–1308. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Chen, W.; Cai, Y.; Zhu, K.; Wei, J.; Lu, Y. Spatial heterogeneity analysis and source identification of heavy metals in soil: A case study of Chongqing, Southwest China. Chem. Biol. Technol. Agric. 2022 , 9 , 50. [ Google Scholar ] [ CrossRef ]
• Navarrete, I.A.; Gabiana, C.C.; Dumo, J.R.E.; Salmo, S.G.; Guzman, M.A.L.G.; Valera, N.S.; Espiritu, E.Q. Heavy metal concentrations in soils and vegetation in urban areas of Quezon City, Philippines. Environ. Monit. Assess. 2017 , 189 , 145. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Lu, S.; Wang, H.; Bai, S. Heavy metal contents and magnetic susceptibility of soils along an urban-rural gradient in rapidly growing city of eastern China. Environ. Monit. Assess. 2009 , 155 , 91–101. [ Google Scholar ] [ CrossRef ]
• Li, F.J.; Yang, H.W.; Ayyamperumal, R.; Liu, Y. Pollution, sources, and human health risk assessment of heavy metals in urban areas around industrialization and urbanization-Northwest China. Chemosphere 2022 , 308 , 136396. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Hani, A.; Pazira, E. Heavy metals assessment and identification of their sources in agricultural soils of Southern Tehran, Iran. Environ. Monit. Assess. 2011 , 176 , 677–691. [ Google Scholar ] [ CrossRef ]
• Guney, M.; Onay, T.T.; Copty, N.K. Impact of overland traffic on heavy metal levels in highway dust and soils of Istanbul, Turkey. Environ. Monit. Assess. 2010 , 164 , 101–110. [ Google Scholar ] [ CrossRef ]
• Du, H.; Lu, X. Spatial distribution and source apportionment of heavy metal(loid)s in urban topsoil in Mianyang, Southwest China. Sci. Rep. 2022 , 12 , 10407. [ Google Scholar ] [ CrossRef ]
• Vorobyov, S.; Kozlov, D.; Poturaeva, T.; Chernyaeva, I. Distribution of heavy metals in soil of urban ecosystems by the example of the Oryol City. IOP Conf. Ser. Earth Environ. Sci. 2017 , 66 , 012002. [ Google Scholar ] [ CrossRef ]
• Zhao, Y.; Shi, X.; Huang, B.; Yu, D.; Wang, H.; Sun, W.; Oboern, I.; Blomback, K. Spatial distribution of heavy metals in agricultural soils of an industry-based peri-urban area in Wuxi, China. Pedosphere 2007 , 17 , 44–51. [ Google Scholar ] [ CrossRef ]
• Zhou, H.; Ouyang, T.; Guo, Y.; Peng, S.; He, C.; Zhu, Z. Assessment of soil heavy metal pollution and its ecological risk for city parks, vicinity of a landfill, and an industrial area within Guangzhou, South China. Appl. Sci. 2022 , 12 , 9345. [ Google Scholar ] [ CrossRef ]
• Moghtaderi, T.; Alamdar, R.; Rodríguez-Seijo, A.; Naghibi, S.J.; Kumar, V. Ecological risk assessment and source apportionment of heavy metal contamination in urban soils in Shiraz, Southwest Iran. Arab. J. Geosci. 2020 , 13 , 797. [ Google Scholar ] [ CrossRef ]
• Praveena, S.M.; Ismail, S.N.S.; Aris, A.Z. Health risk assessment of heavy metal exposure in urban soil from Seri Kembangan (Malaysia). Arab. J. Geosci. 2015 , 8 , 9753–9761. [ Google Scholar ] [ CrossRef ]
• Kong, S.; Tang, J.; Ouyang, F.; Chen, M. Research on the treatment of heavy metal pollution in urban soil based on biochar technology. Environ. Technol. Innov. 2021 , 23 , 101670. [ Google Scholar ] [ CrossRef ]
• Baghdadi, M.E.I.; Barakat, A.; Sajieddine, M.; Nadem, S. Heavy metal pollution and soil magnetic susceptibility in urban soil of Beni Mellal City (Morocco). Environ. Earth Sci. 2012 , 66 , 141–155. [ Google Scholar ] [ CrossRef ]
• Morton-Bermea, O.; Hernandez, E.; Martinez-Pichardo, E.; Soler-Arechalde, A.M.; Santa-Cruz, R.L.; Gonzalez-Hernandez, G.; Beramendi-Orosco, L.; Urrutia-Fucugauchi, J. Mexico City topsoils: Heavy metals vs. magnetic susceptibility. Geoderma 2009 , 151 , 121–125. [ Google Scholar ] [ CrossRef ]
• Tong, S.; Li, H.; Wang, L.; Tudi, M.; Yang, L. Concentration, Spatial Distribution, Contamination Degree and Human Health Risk Assessment of Heavy Metals in Urban Soils across China between 2003 and 2019—A Systematic Review. Int. J. Environ. Res. Public Health 2020 , 17 , 3099. [ Google Scholar ] [ CrossRef ]
• Wei, B.; Yang, L. A review of heavy metal contaminations in urban soils, urban road dusts and agricultural soils from China. Microchem. J. 2010 , 94 , 99–107. [ Google Scholar ] [ CrossRef ]
• Barsova, N.; Yakimenko, O.; Tolpeshta, I.; Motuzova, G. Current state and dynamics of heavy metal soil pollution in Russian Federation—A review. Environ. Pollut. 2019 , 249 , 200–207. [ Google Scholar ] [ CrossRef ]
• Sugimoto, C.R.; Ahn, Y.Y.; Smith, E.; Macaluso, B.; Lariviere, V. Factors affecting sex-related reporting in medical research: A cross-disciplinary bibliometric analysis. Lancet 2019 , 393 , 550–559. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Ali, F.; Park, E.O.; Kwon, J.; Chae, B.K. 30 years of contemporary hospitality management. Int. J. Contemp. Hosp. Manag. 2019 , 31 , 2641–2665. [ Google Scholar ] [ CrossRef ]
• Zupic, I.; Cater, T. Bibliometric methods in management and organization. Organ. Res. Methods 2015 , 18 , 429–472. [ Google Scholar ] [ CrossRef ]
• Zhang, P.; Yan, F.; Du, C. A comprehensive analysis of energy management strategies for hybrid electric vehicles based on bibliometrics. Renew. Sustain. Energy Rev. 2015 , 48 , 88–104. [ Google Scholar ] [ CrossRef ]
• Pauna, V.H.; Buonocore, E.; Renzi, M.; Russo, G.F.; Franzese, P.P. The issue of microplastics in marine ecosystems: A bibliometric network analysis. Mar. Pollut. Bull. 2019 , 149 , 110612. [ Google Scholar ] [ CrossRef ]
• Zhang, Y.; Yao, X.; Qin, B. A critical review of the development, current hotspots, and future directions of Lake Taihu research from the bibliometrics perspective. Environ. Sci. Pollut. Res. 2016 , 23 , 12811–12821. [ Google Scholar ] [ CrossRef ]
• Chen, C. Searching for intellectual turning points: Progressive knowledge domain visualization. Proc. Natl. Acad. Sci. USA 2004 , 101 (Suppl. S1), 5303–5310. [ Google Scholar ] [ CrossRef ]
• Chen, C. CiteSpace II: Detecting and visualizing emerging trends and transient patterns in scientific literature. J. Am. Soc. Inf. Sci. Technol. 2006 , 57 , 359–377. [ Google Scholar ] [ CrossRef ]
• Eck, N.J.V.; Waltman, L.; Dekker, R.; Berg, J.V.D. A comparison of two techniques for bibliometric mapping: Multidimensional scaling and VOS. J. Am. Soc. Inf. Sci. Technol. 2010 , 61 , 2405–2416. [ Google Scholar ]
• Tong, C.H. Let interdisciplinary research begin in undergraduate years. Nature 2010 , 463 , 157. [ Google Scholar ] [ CrossRef ]
• Eck, N.J.V.; Waltman, L. Software survey: VOSviewer, a computer program for bibliometric mapping. Scientometrics 2010 , 84 , 523–538. [ Google Scholar ] [ PubMed ]
• Waltman, L.; Eck, N.J.V.; Noyons, E.C.M. A unified approach to mapping and clustering of bibliometric networks. J. Informetr. 2010 , 4 , 629–635. [ Google Scholar ] [ CrossRef ]
• Price, D.J. Networks of scientific papers. Science 1965 , 149 , 510–515. [ Google Scholar ] [ CrossRef ]
• Milojevic, S. Practical method to reclassify Web of Science articles into unique subject categories and broad disciplines. Quant. Sci. Stud. 2020 , 1 , 183–206. [ Google Scholar ] [ CrossRef ]
• Birkle, C.; Pendlebury, D.A.; Schnell, J.; Adams, J. Web of Science as a data source for research on scientific and scholarly activity. Quant. Sci. Stud. 2020 , 1 , 363–376. [ Google Scholar ] [ CrossRef ]
• Liu, W.; Tang, L.; Hu, G. Funding information in Web of Science: An updated overview. Scientometrics 2020 , 122 , 1509–1524. [ Google Scholar ] [ CrossRef ]
• Liu, F. Retrieval strategy and possible explanations for the abnormal growth of research publications: Re-evaluating a bibliometric analysis of climate change. Scientometrics 2023 , 128 , 853–859. [ Google Scholar ] [ CrossRef ]
• Akhtar, S.; Luqman, M.; Farooq Awan, M.U.; Saba, I.; Khan, Z.I.; Ahmad, K.; Muneeb, A.; Nadeem, M.; Batool, A.I.; Shahzadi, M.; et al. Health risk implications of iron in wastewater soil-food crops grown in the vicinity of peri urban areas of the District Sargodha. PLoS ONE 2022 , 17 , e0275497. [ Google Scholar ] [ CrossRef ]
• Khan, Z.I.; Ahmad, K.; Ashraf, M.; Parveen, R.; Arshad, F.; Hussain, A.; Bibi, Z.; Akram, N.A.; Noorka, I.R.; Mustafa, I. Risk assessment of heavy metal toxicity through contaminated vegetable from sewage water: Implications for populace health. Hum. Ecol. Risk Assess. 2016 , 22 , 302–311. [ Google Scholar ] [ CrossRef ]
• Ahmad, K.; Ashfaq, A.; Khan, Z.I.; Ashraf, M.; Akram, N.A.; Yasmin, S.; Batool, A.I.; Sher, M.; Shad, H.A.; Khan, A.; et al. Health risk assessment of heavy metals and metalloids via dietary intake of a potential vegetable ( Coriandrum sativum L.) grown in contaminated water irrigated agricultural sites of Sargodha, Pakistan. Hum. Ecol. Risk Assess. 2016 , 22 , 597–610. [ Google Scholar ] [ CrossRef ]
• Yang, X.; Yang, Y.; Wan, Y.; Wu, R.; Feng, D.; Li, K. Source identification and comprehensive apportionment of the accumulation of soil heavy metals by integrating pollution landscapes, pathways, and receptors. Sci. Total Environ. 2021 , 786 , 147436. [ Google Scholar ] [ CrossRef ]
• Yang, Y.; Christakos, G. Uncertainty assessment of heavy metal soil contamination mapping using spatiotemporal sequential indicator simulation with multi-temporal sampling points. Environ. Monit. Assess. 2015 , 187 , 571. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Wang, S.; Zhang, Y.; Cheng, J.; Li, Y.; Li, F.; Li, Y.; Shi, Z. Pollution assessment and source apportionment of soil heavy metals in a coastal industrial city, Zhejiang, Southeastern China. Int. J. Environ. Res. Public Health 2022 , 19 , 3335. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Xiang, M.; Ma, J.; Cheng, J.; Lei, K.; Li, F.; Shi, Z.; Li, Y. Collaborative evaluation of heavy metal pollution of soil-crop system in the southeast of Yangtze River Delta, China. Ecol. Indic. 2022 , 143 , 109412. [ Google Scholar ] [ CrossRef ]
• Hu, B.; Shao, S.; Fu, Z.; Li, Y.; Ni, H.; Chen, S.; Zhou, Y.; Jin, B.; Shi, Z. Identifying heavy metal pollution hot spots in soil-rice systems: A case study in South of Yangtze River Delta, China. Sci. Total Environ. 2019 , 658 , 614–625. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Hu, B.; Shao, S.; Ni, H.; Fu, Z.; Hu, L.; Zhou, Y.; Min, X.; She, S.; Chen, S.; Huang, M.; et al. Current status, spatial features, health risks, and potential driving factors of soil heavy metal pollution in China at province level. Environ. Pollut. 2020 , 266 , 114961. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Dzikowski, P. A bibliometric analysis of born global firms. J. Bus. Res. 2018 , 85 , 281–294. [ Google Scholar ] [ CrossRef ]
• Yang, Q.; Li, Z.; Lu, X.; Duan, Q.; Huang, L.; Bi, J. A review of soil heavy metal pollution from industrial and agricultural regions in China: Pollution and risk assessment. Sci. Total Environ. 2018 , 642 , 690–700. [ Google Scholar ] [ CrossRef ]
• Jiang, Y.; Chao, S.; Liu, J.; Yang, Y.; Chen, Y.; Zhang, A.; Cao, H. Source apportionment and health risk assessment of heavy metals in soil for a township in Jiangsu Province, China. Chemosphere 2017 , 168 , 1658–1668. [ Google Scholar ] [ CrossRef ]
• Morton-Bermea, O.; Hernández-Álvarez, E.; González-Hernández, G.; Romero, F.; Lozano, R.; Beramendi-Orosco, L.E. Assessment of heavy metal pollution in urban topsoils from the metropolitan area of Mexico City. J. Geochem. Explor. 2009 , 101 , 218–224. [ Google Scholar ] [ CrossRef ]
• Wang, X.; Qin, Y. Magnetic properties of urban topsoils and correlation with heavy metals: A case study from the city of Xuzhou, China. Environ. Geol. 2006 , 49 , 897–904. [ Google Scholar ] [ CrossRef ]
• Zhong, B.; Liang, T.; Wang, L.; Li, K. Applications of stochastic models and geostatistical analyses to study sources and spatial patterns of soil heavy metals in a metalliferous industrial district of China. Sci. Total Environ. 2014 , 490 , 422–434. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Krami, L.K.; Amiri, F.; Sefiyanian, A.; Shariff, A.R.B.M.; Tabatabaie, T.; Pradhan, B. Spatial patterns of heavy metals in soil under different geological structures and land uses for assessing metal enrichments. Environ. Monit. Assess. 2013 , 185 , 9871–9888. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Guagliardi, I.; Cicchella, D.; De Rosa, R. A geostatistical approach to assess concentration and spatial distribution of heavy metals in urban soils. Water Air Soil. Pollut. 2012 , 223 , 5983–5998. [ Google Scholar ] [ CrossRef ]
• Ahmed, F.; Ishiga, H. Trace metal concentrations in street dusts of Dhaka city, Bangladesh. Atmos. Environ. 2006 , 40 , 3835–3844. [ Google Scholar ] [ CrossRef ]
• Ferreira-Baptista, L.; Miguel, E.D. Geochemistry and risk assessment of street dust in Luanda, Angola: A tropical urban environment. Atmos. Environ. 2005 , 39 , 4501–4512. [ Google Scholar ] [ CrossRef ]
• Arnich, N.; Sirot, V.; Rivière, G.; Jean, J.; Noël, L.; Guérin, T.; Leblanc, J. Dietary exposure to trace elements and health risk assessment in the 2nd French Total Diet Study. Food Chem. Toxicol. 2012 , 50 , 2432–2449. [ Google Scholar ] [ CrossRef ]
• Sarwar, N.; Imran, M.; Shaheen, M.R.; Ishaque, W.; Kamran, M.A.; Matloob, A.; Rehim, A.; Hussain, S. Phytoremediation strategies for soils contaminated with heavy metals: Modifications and future perspectives. Chemosphere 2017 , 171 , 710–721. [ Google Scholar ] [ CrossRef ]
• Mahar, A.; Wang, P.; Ali, A.; Awasthi, M.K.; Lahori, A.H.; Wang, Q.; Li, R.; Zhang, Z. Challenges and opportunities in the phytoremediation of heavy metals contaminated soils: A review. Ecotoxicol. Environ. Saf. 2016 , 126 , 111–121. [ Google Scholar ] [ CrossRef ]
• Whiting, S.N.; de Souza, M.P.; Terry, N. Rhizosphere bacteria mobilize Zn for hyperaccumulation by thlaspi caerulescens. Environ. Sci. Technol. 2001 , 35 , 3144–3150. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Sobariu, D.L.; Fertu, D.I.T.; Diaconu, M.; Pavel, L.V.; Hlihor, R.; Drăgoi, E.N.; Curteanu, S.; Lenz, M.; Corvini, P.F.; Gavrilescu, M. Rhizobacteria and plant symbiosis in heavy metal uptake and its implications for soil bioremediation. New Biotechnol. 2017 , 39 , 125–134. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Liang, J.; Feng, C.; Zeng, G.; Gao, X.; Zhong, M.; Li, X.; Li, X.; He, X.; Fang, Y. Spatial distribution and source identification of heavy metals in surface soils in a typical coal mine city, Lianyuan, China. Environ. Pollut. 2017 , 225 , 681–690. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Solgi, E.; Oshvandi, Z. Spatial patterns, hotspot, and risk assessment of heavy metals in different land uses of urban soils (case study: Malayer city). Hum. Ecol. Risk Assess. 2018 , 24 , 256–270. [ Google Scholar ] [ CrossRef ]
• Shi, X.M.; Liu, S.; Song, L.; Wu, C.S.; Yang, B.; Lu, H.Z.; Wang, X.; Zakari, S. Contamination and source-specific risk analysis of soil heavy metals in a typical coal industrial city, central China. Sci. Total Environ. 2022 , 836 , 155694. [ Google Scholar ] [ CrossRef ]
• Huang, J.; Wu, Y.; Sun, J.; Li, X.; Geng, X.; Zhao, M.; Sun, T.; Fan, Z. Health risk assessment of heavy metal(loid)s in park soils of the largest megacity in China by using monte carlo simulation coupled with positive matrix factorization model. J. Hazard. Mater. 2021 , 415 , 125629. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Vince, T.; Szabó, G.; Csoma, Z.; Sándor, G.; Szabó, S. The spatial distribution pattern of heavy metal concentrations in urban soils-a study of anthropogenic effects in Berehove, Ukraine. Open Geosci. 2014 , 6 , 330–343. [ Google Scholar ] [ CrossRef ]
• Sun, Y.; Zhou, Q.; Xie, X.; Liu, R. Spatial, sources and risk assessment of heavy metal contamination of urban soils in typical regions of Shenyang, China. J. Hazard. Mater. 2010 , 174 , 455–462. [ Google Scholar ] [ CrossRef ]
• Huang, Y.; Deng, M.; Wu, S.; Japenga, J.; Li, T.; Yang, X.; He, Z. A modified receptor model for source apportionment of heavy metal pollution in soil. J. Hazard. Mater. 2018 , 354 , 161–169. [ Google Scholar ] [ CrossRef ]
• Qu, M.; Wang, Y.; Huang, B.; Zhao, Y. Source apportionment of soil heavy metals using robust absolute principal component scores-robust geographically weighted regression (RAPCS-RGWR) receptor model. Sci. Total Environ. 2018 , 626 , 203–210. [ Google Scholar ] [ CrossRef ]
• Liu, J.; Kang, H.; Tao, W.; Li, H.; He, D.; Ma, L.; Tang, H.; Wu, S.; Yang, K.; Li, X. A spatial distribution-principal component analysis (SD-PCA) model to assess pollution of heavy metals in soil. Sci. Total Environ. 2023 , 859 , 160112. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Cortés, J.L.; Bautista, F.; Delgado, C.; Quintana, P.; Aguilar, D.; García, A.; Figueroa, C.; Gogichaishvili, A. Spatial distribution of heavy metals in urban dust from Ensenada, Baja California, Mexico. Rev. Chapingo Ser. Cienc. For. Ambiente 2016 , 23 , 47–60. [ Google Scholar ] [ CrossRef ]
• Ruhling, A. A European survey of atmospheric heavy metal deposition in 2000–2001. Environ. Pollut. 2002 , 120 , 23–25. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Muya, F.N.; Sunday, C.E.; Baker, P.; Iwuoha, E. Environmental remediation of heavy metal ions from aqueous solution through hydrogel adsorption: A critical review. Water Sci. Technol. 2016 , 73 , 983–992. [ Google Scholar ] [ CrossRef ]
• Akhtar, S.; Khan, Z.I.; Ahmad, K.; Nadeem, M.; Ejaz, A.; Hussain, M.I.; Ashraf, M.A. Assessment of lead toxicity in diverse irrigation regimes and potential health implications of agriculturally grown crops in Pakistan. Agric. Water Manag. 2022 , 271 , 107743. [ Google Scholar ] [ CrossRef ]
• Yang, L.; Ren, Q.; Zheng, K.; Jiao, Z.; Ruan, X.; Wang, Y. Migration of heavy metals in the soil-grape system and potential health risk assessment. Sci. Total Environ. 2022 , 806 , 150646. [ Google Scholar ] [ CrossRef ]
• Mehmood, K.; Ahmad, H.R.; Abbas, R.; Saifullah; Murtaza, G. Heavy metals in urban and peri-urban soils of a heavily-populated and industrialized city: Assessment of ecological risks and human health repercussions. Hum. Ecol. Risk Assess. 2019 , 26 , 1705–1722. [ Google Scholar ] [ CrossRef ]
• Adimalla, N.; Chen, J.; Qian, H. Spatial characteristics of heavy metal contamination and potential human health risk assessment of urban soils: A case study from an urban region of South India. Ecotoxicol. Environ. Saf. 2020 , 194 , 110406. [ Google Scholar ] [ CrossRef ]
• Massas, I.; Ehaliotis, C.; Kalivas, D.; Panagopoulou, G. Concentrations and availability indicators of soil heavy metals; the case of children’s playgrounds in the city of Athens (Greece). Water Air Soil Pollut. 2010 , 212 , 51–63. [ Google Scholar ] [ CrossRef ]
• Song, D.; Zhuang, D.; Jiang, D.; Fu, J.; Wang, Q. Integrated health risk assessment of heavy metals in Suxian County, South China. Int. J. Environ. Res. Public Health 2015 , 12 , 7100–7117. [ Google Scholar ] [ CrossRef ]
• Khan, S.; Munir, S.; Sajjad, M.; Li, G. Urban park soil contamination by potentially harmful elements and human health risk in Peshawar City, Khyber Pakhtunkhwa, Pakistan. J. Geochem. Explor. 2016 , 165 , 102–110. [ Google Scholar ] [ CrossRef ]
• Teixeira, E.C.; Agudelo-Castañeda, D.M.; Mattiuzi, C.D.P. Contribution of polycyclic aromatic hydrocarbon (PAH) sources to the urban environment: A comparison of receptor models. Sci. Total Environ. 2015 , 538 , 212–219. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Tang, L.; Tang, X.Y.; Zhu, Y.G.; Zheng, M.H.; Miao, Q.L. Contamination of polycyclic aromatic hydrocarbons (PAHs) in urban soils in Beijing, China. Environ. Int. 2005 , 31 , 822–828. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Wu, S.; Zhou, S.; Bao, H.; Chen, D.; Wang, C.; Li, B.; Tong, G.; Yuan, Y.; Xu, B. Improving risk management by using the spatial interaction relationship of heavy metals and PAHs in urban soil. J. Hazard. Mater. 2019 , 364 , 108–116. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Gao, L.; Shi, Y.; Li, W.; Liu, J.; Cai, Y. Occurrence and distribution of antibiotics in urban soil in Beijing and Shanghai, China. Environ. Sci. Pollut. Res. 2015 , 22 , 11360–11371. [ Google Scholar ] [ CrossRef ]
• Weissmannova, H.D.; Pavlovsky, J. Indices of soil contamination by heavy metals-methodology of calculation for pollution assessment (minireview). Environ. Monit. Assess. 2017 , 189 , 616–625. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Karim, Z.; Qureshi, B.A.; Mumtaz, M. Geochemical baseline determination and pollution assessment of heavy metals in urban soils of Karachi, Pakistan. Ecol. Indic. 2015 , 48 , 358–364. [ Google Scholar ] [ CrossRef ]
• Liu, J.; Liu, C.; Han, W. Efficiently evaluating heavy metal urban soil pollution using an improved entropy-method-based topsis model. Arch. Environ. Contam. Toxicol. 2016 , 71 , 377–382. [ Google Scholar ] [ CrossRef ]
• Zhang, J.; Lin, Q.; Liu, B.; Guan, Y.; Wang, Y.; Li, D.; Zhou, X.; Kang, X. Magnetic response of heavy metal pollutionin soil of urban street greenbelts. Pol. J. Environ. Stud. 2022 , 31 , 1923–1933. [ Google Scholar ] [ CrossRef ]
• Hu, X.F.; Su, Y.; Ye, R.; Li, X.Q.; Zhang, G.L. Magnetic properties of the urban soils in Shanghai and their environmental implications. CATENA 2007 , 70 , 428–436. [ Google Scholar ] [ CrossRef ]
• Zhang, C.; Appel, E.; Qiao, Q. Heavy metal pollution in farmland irrigated with river water near a steel plant-magnetic and geochemical signature. Geophys. J. Int. 2013 , 192 , 963–974. [ Google Scholar ] [ CrossRef ]
• Hu, X.F.; Li, M.; He, Z.C.; Cui, L.; Liu, R.; Wang, X.D.; Wang, Z.H. Magnetic responses to heavy metal pollution of the industrial soils in Shanghai: Implying the influences of anthropogenic magnetic dustfall on urban environment. J. Appl. Geophys. 2022 , 197 , 104544. [ Google Scholar ] [ CrossRef ]
• Liu, D.; Ma, J.; Sun, Y.; Li, Y. Spatial distribution of soil magnetic susceptibility and correlation with heavy metal pollution in Kaifeng City, China. CATENA 2016 , 139 , 53–60. [ Google Scholar ] [ CrossRef ]
• Magiera, T.; Parzentny, H.; Róg, L.; Chybiorz, R.; Wawer, M. Spatial variation of soil magnetic susceptibility in relation to different emission sources in southern Poland. Geoderma 2015 , 255 , 94–103. [ Google Scholar ] [ CrossRef ]
• Wang, B.; Xia, D.; Yu, Y.; Jia, J.; Xu, S. Detection and differentiation of pollution in urban surface soils using magnetic properties in arid and semi-arid regions of northwestern China. Environ. Pollut. 2014 , 184 , 335–346. [ Google Scholar ] [ CrossRef ] [ PubMed ]
• Rachwał, M.; Magiera, T.; Wawer, M. Coke industry and steel metallurgy as the source of soil contamination by technogenic magnetic particles, heavy metals and polycyclic aromatic hydrocarbons. Chemosphere 2015 , 138 , 863–873. [ Google Scholar ] [ CrossRef ]
• Wang, X.S. Assessment of heavy metal pollution in Xuzhou urban topsoils by magnetic susceptibility measurements. J. Appl. Geophys. 2013 , 92 , 76–83. [ Google Scholar ] [ CrossRef ]
• Li, X.D.; Poon, C.S.; Liu, P.S. Heavy metal contamination of urban soils and street dusts in Hong Kong. Appl. Geochem. 2001 , 16 , 1361–1368. [ Google Scholar ] [ CrossRef ]
• Pranckutė, R. Web of Science (WoS) and Scopus: The Titans of Bibliographic Information in Today’s Academic World. Publications 2021 , 9 , 12. [ Google Scholar ] [ CrossRef ]
• Franceschini, F.; Maisano, D.; Mastrogiacomo, L. Empirical analysis and classification of database errors in Scopus and Web of Science. J. Informetr. 2016 , 10 , 933–953. [ Google Scholar ] [ CrossRef ]

Click here to enlarge figure

Share and Cite

Tang, S.; Wang, C.; Song, J.; Ihenetu, S.C.; Li, G. Advances in Studies on Heavy Metals in Urban Soil: A Bibliometric Analysis. Sustainability 2024 , 16 , 860. https://doi.org/10.3390/su16020860

Tang S, Wang C, Song J, Ihenetu SC, Li G. Advances in Studies on Heavy Metals in Urban Soil: A Bibliometric Analysis. Sustainability . 2024; 16(2):860. https://doi.org/10.3390/su16020860

Tang, Shuya, Chunhui Wang, Jing Song, Stanley Chukwuemeka Ihenetu, and Gang Li. 2024. "Advances in Studies on Heavy Metals in Urban Soil: A Bibliometric Analysis" Sustainability 16, no. 2: 860. https://doi.org/10.3390/su16020860

Article Metrics

Article access statistics, further information, mdpi initiatives, follow mdpi.

Subscribe to receive issue release notifications and newsletters from MDPI journals