留言板

尊敬的读者、作者、审稿人, 关于本刊的投稿、审稿、编辑和出版的任何问题, 您可以本页添加留言。我们将尽快给您答复。谢谢您的支持!

姓名
邮箱
手机号码
标题
留言内容
验证码

The Massive Expansion and Spatial Transformation of Potentially Contaminated Land Across China in 1990–2020 Observed from Remote Sensing and Big-data

Yinyin DOU Changqing GUO Wenhui KUANG Wenfeng CHI Mei LEI

DOU Yinyin, GUO Changqing, KUANG Wenhui, CHI Wenfeng, LEI Mei, 2022. The Massive Expansion and Spatial Transformation of Potentially Contaminated Land Across China in 1990–2020 Observed from Remote Sensing and Big-data. Chinese Geographical Science, 32(5): 776−791 doi:  10.1007/s11769-022-1300-6
Citation: DOU Yinyin, GUO Changqing, KUANG Wenhui, CHI Wenfeng, LEI Mei, 2022. The Massive Expansion and Spatial Transformation of Potentially Contaminated Land Across China in 1990–2020 Observed from Remote Sensing and Big-data. Chinese Geographical Science, 32(5): 776−791 doi:  10.1007/s11769-022-1300-6

doi: 10.1007/s11769-022-1300-6

The Massive Expansion and Spatial Transformation of Potentially Contaminated Land Across China in 1990–2020 Observed from Remote Sensing and Big-data

Funds: Under the auspices of the National Key Research and Development Program (No. 2018YFC1800103, 2018YFC1800106)
More Information
    • 关键词:
    •  / 
    •  / 
    •  / 
    •  / 
    •  
  • Figure  1.  Study areas and six types of potentially contaminated land in China. EE, energy exploitation; EP, energy processing; CRME, chemical raw material exploitation; MSP, metal smelting & processing; MOMD, metal ore mining & dressing; CP, chemical processing

    Figure  2.  Flow chart of the study

    Figure  3.  Spatial distribution of sampling points in China. Not including Taiwan Province of China due to data acquisition limitations

    Figure  4.  Spatial patterns of different types of PCL across China in 2020. Not including Taiwan Province of China due to data acquisition limitations. CP, chemical processing; MSP, metal smelting and processing; EP, energy processing; MOMD, metal ore mining and dressing; EE, energy exploitation; CRME, chemical raw material exploitation

    Figure  5.  Intensity of change of China’s potentially contaminated land (PCL) in different regions during 1990–2020. Not including Taiwan Province of China due to data acquisition limitations. CRME, chemical raw material exploitation; CP, chemical processing; EE, energy exploitation; EP, energy processing; MOMD, metal ore mining and dressing; MSP, metal smelting and processing

    Figure  6.  Evolution of different types of potentially contaminated land (PCL) across China during 1990–2020. CP, chemical processing; MSP, metal ore smelting and processing; EP, energy processing; CRME, chemical raw material exploitation; EE, energy exploitation; MOMD, metal ore mining and dressing

    Figure  7.  Area and distribution of China’s potentially contaminated land in different zones during 1990–2020. SW, Southwest; NW, Northwest; QTP, Qinghai-Tibet Plateau; N, North; SE, Southeast; NE, Northeast

    Figure  8.  Hotspot distributions of potentially contaminated land (PCL) in China during 1990–2020. Not including Taiwan Province of China due to data acquisition limitations. MOMD, metal ore mining and dressing; CP, chemical processing; EE, energy exploitation; MSP, metal smelting and processing; CRME, chemical raw material exploitation; EP, energy processing

    Table  1.   Classification of potentially contaminated land (PCL)

    Level-I typeLevel-II typeMain content
    Mining Metal ore mining & dressing Areas for minerals predominated by ferrous metals, nonferrous metals and other metal mineral resources, including the mining area, tailings reservoir, stacking yard, dumping site, and their ancillary facilities
    Energy exploitation Areas for minerals predominated by coal, coalbed methane, bone coal, oil shale, petroleum, natural gas, and other fossil fuel resources, including the mining area, tailings reservoir, stacking yard, dumping site, and their ancillary facilities
    Chemical raw material exploitation Areas for minerals predominated by chemical, fertilizer minerals, and other nonmetal chemical raw material mineral resources, including the mining area, tailings reservoir, stacking yard, dumping site, and their ancillary facilities
    Processing Metal smelting & processing Contiguous regions where the areas dominated by the smelting and processing industry of ferrous, nonferrous, and other metals are centralized
    Energy processing Contiguous regions where the areas dominated by the processing industry of coal, crude oil, and other fossil fuels are centralized
    Chemical processing Contiguous regions where the areas dominated by the manufacturing industry of inorganic chemicals, organic chemicals, fertilizers, pesticides, and other chemical raw materials and products are centralized
    下载: 导出CSV

    Table  2.   Accuracy of potentially contaminated land (PCL) in different geographical zones in different years

    Zone in ChinaIndex1990200020102020Overall accuracya
    North Total number of samples 643 831 1030 918 93.87
    Correct number of samples 594 774 967 882
    Incorrect number of samples 49 57 63 36
    Producer’s accuracy / % 92.38 93.14 93.88 96.08
    Northeast Total number of samples 294 380 400 334 95.08
    Correct number of samples 277 357 383 322
    Incorrect number of samples 17 23 17 12
    Producer’s accuracy / % 94.22 93.95 95.75 96.41
    Northwest Total number of samples 266 336 436 428 92.11
    Correct number of samples 246 308 397 399
    Incorrect number of samples 20 28 39 29
    Producer’s accuracy / % 92.48 91.67 91.06 93.22
    Southeast Total number of samples 541 603 973 789 91.06
    Correct number of samples 485 549 887 729
    Incorrect number of samples 56 54 86 60
    Producer’s accuracy / % 89.65 91.04 91.16 92.40
    Southwest Total number of samples 432 530 807 599 94.43
    Correct number of samples 413 493 758 570
    Incorrect number of samples 19 37 49 29
    Producer’s accuracy / % 95.60 93.02 93.93 95.16
    Qinghai-Tibet Plateau Total number of samples 44 51 64 53 93.54
    Correct number of samples 41 48 58 51
    Incorrect number of samples 3 3 6 2
    Producer’s accuracy / % 93.18 94.12 90.63 96.23
    Nationwide Total number of samples 2220 2731 3710 3121 93.21
    Correct number of samples 2056 2529 3450 2953
    Incorrect number of samples 83 114 96 62
    Producer’s accuracy / % 92.61 92.60 92.99 94.62
    Note: a Overall accuracy is the average of data accuracies of 1990, 2000, 2010, and 2020
    下载: 导出CSV

    Table  3.   Variations in area of China’s potentially contaminated land (PCL) during 1990–2020

    TypeScale of areaChanges in size of area / km2Changes in size of area per year / (km2/yr)
    20201990–20002000–20102010–20201990–20201990–2020
    Metal ore mining & dressing Large 241.84 36.67 34.91 77.28 148.86 4.96
    Medium 548.32 207.40 137.49 −276.60 68.29 2.28
    Small 1089.88 366.11 727.22 −963.05 130.28 4.34
    Energy exploitation Large 433.11 122.45 92.73 87.02 302.20 10.07
    Medium 603.87 90.90 234.50 −57.50 267.90 8.93
    Small 649.31 218.03 356.10 −450.22 123.91 4.13
    Chemical raw material exploitation Large 527.58 1.64 224.47 131.88 357.99 11.93
    Medium 73.43 23.56 7.21 −13.96 16.81 0.56
    Small 152.56 59.34 113.49 −147.80 25.03 0.83
    Energy processing Large 534.24 107.53 226.53 142.69 476.75 15.89
    Medium 66.61 11.51 44.17 4.76 60.44 2.01
    Small 3.22 1.77 1.02 0.43 3.22 0.11
    Metal smelting & processing Large 3043.14 251.74 1225.74 623.55 2101.03 70.03
    Medium 1010.85 223.18 350.63 295.52 869.33 28.98
    Small 26.01 13.96 1.34 8.28 23.58 0.79
    Chemical processing Large 3550.49 318.48 1233.89 1057.47 2609.84 86.99
    Medium 1263.78 277.38 453.41 368.55 1099.34 36.64
    Small 37.84 23.95 2.32 4.31 30.58 1.02
    All Nationwide 13,856.08 2355.60 5467.17 892.61 8715.38 290.51
    下载: 导出CSV

    Table  4.   Accuracy of the hotspot of China’s potentially contaminated land (PCL) in different geographical zones in different years

    Zone in ChinaTotal number of samplesProducer’s accuracy / %
    19902000201020201990200020102020
    North 469 583 950 610 86.43 86.79 87.30 89.30
    Northeast 137 190 297 181 87.10 86.32 86.98 88.79
    Northwest 55 82 165 103 87.88 85.37 86.87 89.03
    Southeast 649 830 1481 958 86.44 86.87 87.31 88.94
    Southwest 373 516 903 455 86.86 87.02 87.38 89.45
    Qinghai-Tibet Plateau 8 10 10 10 75.00 80.00 80.00 90.00
    Total 1691 2211 3806 2317 86.52 86.75 87.23 89.17
    下载: 导出CSV
  • [1] Ahmed A I, Bryant R G, Edwards D P, 2021. Where are mines located in sub Saharan Africa and how have they expanded overtime? Land Degradation & Development, 32(1): 112–122. doi:  10.1002/ldr.3706
    [2] Bernhardt E S, Lutz B D, King R S et al., 2012. How many mountains can we mine? assessing the regional degradation of Central Appalachian rivers by surface coal mining. Environmental Science & Technology, 46(15): 8115–8122. doi:  10.1021/es301144q
    [3] Boldi P, Codenotti B, Santini M et al., 2004. Ubicrawler: a scalable fully distributed web crawler. Software:Practice and Experience, 34(8): 711–726. doi:  10.1002/spe.587
    [4] Chakrabarti S, van den Berg M, Dom B, 1999. Focused crawling: a new approach to topic-specific Web resource discovery. Computer Networks, 31(11-16): 1623–1640. doi:  10.1016/S1389-1286(99)00052-3
    [5] Chen J, Chen J, Liao A P et al., 2015. Global land cover mapping at 30 m resolution: a POK-based operational approach. ISPRS Journal of Photogrammetry and Remote Sensing, 103: 7–27. doi:  10.1016/j.isprsjprs.2014.09.002
    [6] Cribari V, Strager M P, Maxwell A E et al., 2021. Landscape changes in the southern coalfields of west Virginia: multi-level intensity analysis and surface mining transitions in the headwaters of the coal river from 1976 to 2016. Land, 10(7): 748. doi:  10.3390/land10070748
    [7] Deng M, Liu Q L, Cheng T et al., 2011. An adaptive spatial clustering algorithm based on delaunay triangulation. Computers, Environment and Urban Systems, 35(4): 320–332. doi:  10.1016/j.compenvurbsys.2011.02.003
    [8] Endl A, Tost M, Hitch M et al., 2021. Europe’s mining innovation trends and their contribution to the sustainable development goals: blind spots and strong points. Resources Policy, 74: 101440. doi:  10.1016/j.resourpol.2019.101440
    [9] Gallwey J, Robiati C, Coggan J et al., 2020. A Sentinel-2 based multispectral convolutional neural network for detecting artisanal small-scale mining in Ghana: applying deep learning to shallow mining. Remote Sensing of Environment, 248: 111970. doi:  10.1016/j.rse.2020.111970
    [10] Gong P, Liu H, Zhang M N et al., 2019. Stable classification with limited sample: transferring a 30-m resolution sample set collected in 2015 to mapping 10-m resolution global land cover in 2017. Science Bulletin, 64(6): 370–373. doi:  10.1016/j.scib.2019.03.002
    [11] Gong P, Li X C, Wang J et al., 2020. Annual maps of global artificial impervious area (GAIA) between 1985 and 2018. Remote Sensing of Environment, 236: 111510. doi:  10.1016/j.rse.2019.111510
    [12] Guo Changqing, Chi Wenfeng, Kuang Wenhui et al., 2022. Mapping and spatio-temporal changes analysis of energy mining and producing sites in China using multi-source data from 1990 to 2020. Journal of Geo-Information Science, 24(1): 127–140. (in Chinese)
    [13] He C Y, Liu Z F, Gou S Y, 2019. Detecting global urban expansion over the last three decades using a full convolutional network. Environmental Research Letters, 14(3): 034008. doi:  10.1088/1748-9326/aaf936
    [14] Jiang Y F, Wang H L, Lei M, 2021. An integrated assessment methodology for management of potentially contaminated sites based on public data. Science of the Total Environment, 783: 146913. doi:  10.1016/j.scitotenv.2021.146913
    [15] Kausar A, Dhaka V S, Singh S K, 2013. Web crawler: a review. International Journal of Computer Applications, 63(2): 31–36. doi:  10.5120/10440-5125
    [16] Kim K, Park O, Barr J et al., 2019. Tourists’ shifting perceptions of UNESCO heritage sites: lessons from Jeju Island-South Korea. Tourism Review, 74(1): 20–29. doi:  10.1108/TR-09-2017-0140
    [17] Kimijima S, Sakakibara M, Nagai M et al., 2021. Time-seres assessment of camp-type artisanal and small-scale gold mining sectors with large influxes of miners using Landsat imagery. International Journal of Environmental Research and Public Health, 18(18): 9441. doi:  10.3390/ijerph18189441
    [18] Kuang W H, Liu J Y, Dong J W et al., 2016. The rapid and massive urban and industrial land expansions in China between 1990 and 2010: a CLUD-based analysis of their trajectories, patterns, and drivers. Landscape and Urban Planning, 145: 21–33. doi:  10.1016/j.landurbplan.2015.10.001
    [19] Kuang W H, Liu J Y, Tian H Q et al., 2022. Cropland redistribution to marginal lands undermines environmental sustainability. National Science Review, 9(1): nwab091. doi:  10.1093/nsr/nwab091
    [20] Li H K, Xu F, Li Q, 2020. Remote sensing monitoring of land damage and restoration in rare earth mining areas in 6 counties in southern Jiangxi based on multisource sequential images. Journal of Environmental Management, 267: 110653. doi:  10.1016/j.jenvman.2020.110653
    [21] Li T K, Liu Y, Bjerg P L, 2021. Prioritization of potentially contaminated sites: a comparison between the application of a solute transport model and a risk-screening method in China. Journal of Environmental Management, 281: 111765. doi:  10.1016/j.jenvman.2020.111765
    [22] Li W W, Yang C W, Yang C J, 2010. An active crawler for discovering geospatial web services and their distribution pattern-a case study of OGC Web Map Service. International Journal of Geographical Information Science, 24(8): 1127–1147. doi:  10.1080/13658810903514172
    [23] Liu X P, Hu G H, Chen Y M et al., 2018. High-resolution multi-temporal mapping of global urban land using Landsat images based on the Google Earth Engine platform. Remote Sensing of Environment, 209: 227–239. doi:  10.1016/j.rse.2018.02.055
    [24] Luckeneder S, Giljum S, Schaffartzik A et al., 2021. Surge in global metal mining threatens vulnerable ecosystems. Global Environmental Change, 69: 102303. doi:  10.1016/j.gloenvcha.2021.102303
    [25] Maus V, Giljum S, Gutschlhofer J et al., 2020. A global-scale data set of mining areas. Scientific Data, 7(1): 289. doi:  10.1038/s41597-020-00624-w
    [26] MEE (Ministry of Ecology and Environment of the People’s Republic of China), 2014. Available at:http://www.gov.cn/xinwen/2014-04/17/content_2661765.htm. Cited 20 October 2021.
    [27] NBS (National Bureau of Statistics of the People’s Republic of China), 2021. Available at:http://www.stats.gov.cn/. Cited 20 October 2021.
    [28] Ning J, Liu J Y, Kuang W H et al., 2018. Spatiotemporal patterns and characteristics of land-use change in China during 2010–2015. Journal of Geographical Sciences, 28(5): 547–562. doi:  10.1007/s11442-018-1490-0
    [29] Nortcliff S, 2001. Sampling and pre-treatment-some observations from the United Kingdom. Science of the Total Environment, 264(1-2): 163–168. doi:  10.1016/s0048-9697(00)00619-7
    [30] Okabe A, Satoh T, Sugihara K, 2009. A kernel density estimation method for networks, its computational method and a GIS-based tool. International Journal of Geographical Information Science, 23(1): 7–32. doi:  10.1080/13658810802475491
    [31] Owusu-Nimo F, Mantey J, Nyarko K B et al., 2018. Spatial distribution patterns of illegal artisanal small scale gold mining (Galamsey) operations in Ghana: a focus on the Western Region. Heliyon, 4(2): e00534. doi:  10.1016/j.heliyon.2018.e00534
    [32] Palmer M A, Bernhardt E S, Schlesinger W H et al., 2010. Mountaintop mining consequences. Science, 327(5962): 148–149. doi:  10.1126/science.1180543
    [33] Peng Y, Zhang Z M, He G J et al., 2019. An improved GrabCut method based on a visual attention model for rare-earth ore mining area recognition with high-resolution remote sensing images. Remote Sensing, 11(8): 987. doi:  10.3390/rs11080987
    [34] Pericak A A, Thomas C J, Kroodsma D A et al., 2018. Mapping the yearly extent of surface coal mining in Central Appalachia using Landsat and Google Earth Engine. PLoS One, 13(7): e0197758. doi:  10.1371/journal.pone.0197758
    [35] Rampanelli G B, Braun A B, Visentin C et al., 2021. The process of selecting a method for identifying potentially contaminated sites-a case study in a municipality in Southern Brazil. Water, Air, & Soil Pollution, 232(1): 26. doi:  10.1007/s11270-020-04970-8
    [36] Rojano R E, Manzano C A, Toro R A et al., 2018. Potential local and regional impacts of particulate matter emitted from one of the world’s largest open-pit coal mines. Air Quality, Atmosphere & Health, 11(5): 601–610. doi:  10.1007/s11869-017-0542-4
    [37] Saley A A, Baratoux D, Baratoux L et al., 2021. Evolution of the koma bangou gold panning site (Niger) from 1984 to 2020 using Landsat imagery. Earth and Space Science, 8(11): e2021EA001879. doi:  10.1029/2021EA001879
    [38] SCPRC (State Council of the People’s Republic of China), 2013. Air pollution prevention and control action plan. Available at: http://www.gov.cn/zwgk/2013-09/12/content_2486773.htm. Cited 20 October 2021.
    [39] SCPRC, 2015. Water pollution prevention and control action plan. Available at: http://www.gov.cn/zhengce/content/2015-04/16/content_9613.htm. Cited 20 October 2021.
    [40] SCPRC, 2016. Soil pollution prevention and control action plan. Available at: http://www.gov.cn/zhengce/content/2016-05/31/content_5078377.htm. Cited 20 October 2021.
    [41] Song Z C, Wang C, Ding L et al., 2021. Soil mercury pollution caused by typical anthropogenic sources in China: evidence from stable mercury isotope measurement and receptor model analysis. Journal of Cleaner Production, 288: 125687. doi:  10.1016/j.jclepro.2020.125687
    [42] Sonter L J, Herrera D, Barrett D J et al., 2017. Mining drives extensive deforestation in the Brazilian Amazon. Nature Communications, 8(1): 1013. doi:  10.1038/s41467-017-00557-w
    [43] Sonter L J, Dade M C, Watson J E M et al., 2020. Renewable energy production will exacerbate mining threats to biodiversity. Nature Communications, 11(1): 4174. doi:  10.1038/s41467-020-17928-5
    [44] Tang L, Liu X M, Wang X Q et al., 2020. Statistical analysis of tailings ponds in China. Journal of Geochemical Exploration, 216: 106579. doi:  10.1016/j.gexplo.2020.106579
    [45] Tang L, Werner T T, Xie H P et al., 2021. A global-scale spatial assessment and geodatabase of mine areas. Global and Planetary Change, 204: 103578. doi:  10.1016/j.gloplacha.2021.103578
    [46] Tao S L, Fang J Y, Zhao X et al., 2015. Rapid loss of lakes on the Mongolian Plateau. Proceedings of the National Academy of Sciences of the United States of America, 112(7): 2281–2286. doi:  10.1073/pnas.1411748112
    [47] UN (United Nations), 2015. Transformation our world: the 2030 agenda for sustainable development A/RES/70/1. Available at: http://undocs.org/en/A/RES/70/1.
    [48] Walker L R, 1999. Ecosystems of disturbed ground. In: Ecosystems of the World. Amsterdam: Elsevier, 16.
    [49] Wang Fei, Zhang Manli, Wang Xuejiao et al., 2021. Generation and pollution characteristics of hazardous wastes from smelting of copper, lead and zinc in China. Journal of Environmental Engineering Technology, 11(5): 1012–1019. (in Chinese)
    [50] Wang Z X, Chen J Q, Chai L Y et al., 2011. Environmental impact and site-specific human health risks of chromium in the vicinity of a ferro-alloy manufactory, China. Journal of Hazardous Materials, 190(1-3): 980–985. doi:  10.1016/j.jhazmat.2011.04.039
    [51] Werner T T, Mudd G M, Schipper A M et al., 2020. Global-scale remote sensing of mine areas and analysis of factors explaining their extent. Global Environmental Change, 60: 102007. doi:  10.1016/j.gloenvcha.2019.102007
    [52] Yang H, Huang X J, Thompson J R et al., 2014. China’s soil pollution: urban brownfields. Science, 344(6185): 691–692. doi:  10.1126/science.344.6185.691-b
    [53] Yuan D Z, Hu Z Q, Yang K et al., 2021. Assessment of the ecological impacts of coal mining and restoration in alpine areas: a case study of the Muli coalfield on the Qinghai-Tibet Plateau. IEEE Access, 9: 162919–162934. doi:  10.1109/ACCESS.2021.3133478
    [54] Zeng X J, Liu Z F, He C Y et al., 2017. Detecting surface coal mining areas from remote sensing imagery: an approach based on object-oriented decision trees. Journal of Applied Remote Sensing, 11(1): 015025. doi:  10.1117/1.JRS.11.015025
    [55] Zhang C, Kuang W H, Wu J G et al., 2021. Industrial land expansion in rural China threatens environmental securities. Frontiers of Environmental Science & Engineering, 15(2): 29. doi:  10.1007/s11783-020-1321-2
    [56] Zhang X, Liu L Y, Wu C S et al., 2020. Development of a global 30 m impervious surface map using multisource and multitemporal remote sensing datasets with the Google Earth Engine platform. Earth System Science Data, 12(3): 1625–1648. doi:  10.5194/essd-12-1625-2020
    [57] Zhang Z X, Wang X, Zhao X L et al., 2014. A 2010 update of National Land Use/Cover Database of China at 1: 100000 scale using medium spatial resolution satellite images. Remote Sensing of Environment, 149: 142–154. doi:  10.1016/j.rse.2014.04.004
  • [1] Shiman HUANG, Qingwu HU, Shaohua WANG, Haidong LI.  Ecological Risk Assessment of World Heritage Sites Using RS and GIS: A Case Study of Huangshan Mountain, China . Chinese Geographical Science, 2022, 32(5): 808-823. doi: 10.1007/s11769-022-1302-4
    [2] Li MA, Lei KANG, Huazheng TIAN, Die XU.  Quantitative Evaluation of Ecological Cumulative Effects and Zoning Regulations for Prefectural Mining Units in China . Chinese Geographical Science, 2022, 32(1): 49-63. doi: 10.1007/s11769-022-1258-4
    [3] Xinshuang WANG, Jiancheng CAO, Jiange LIU, Xiangwu LI, Lu WANG, Feihang ZUO, Mu BAI.  Improving the Interpretability and Reliability of Regional Land Cover Classification by U-Net Using Remote Sensing Data . Chinese Geographical Science, 2022, 32(6): 979-994. doi: 10.1007/s11769-022-1315-z
    [4] Baolin LIU, Deming DONG, Xiuyi HUA, Weihua DONG, Ming LI.  Spatial Distribution and Ecological Risk Assessment of Heavy Metals in Surface Sediment of Songhua River, Northeast China . Chinese Geographical Science, 2021, 31(2): 223-233. doi: 10.1007/s11769-021-1186-8
    [5] Xiao WANG, Lei CHE, Liang ZHOU, Jiangang XU.  Spatio-temporal Dynamic Simulation of Land use and Ecological Risk in the Yangtze River Delta Urban Agglomeration, China . Chinese Geographical Science, 2021, 31(5): 829-847. doi: 10.1007/s11769-021-1229-1
    [6] DU Jia, SONG Kaishan, YAN Baohua.  Impact of the Zhalong Wetland on Neighboring Land Surface Temper-ature Based on Remote Sensing and GIS . Chinese Geographical Science, 2019, 20(5): 798-808. doi: 10.1007/s11769-019-1050-2
    [7] WANG Xuecheng, YANG Fei, GAO Xing, WANG Wei, ZHA Xinjie.  Evaluation of Forest Damaged Area and Severity Caused by Ice-snow Frozen Disasters over Southern China with Remote Sensing . Chinese Geographical Science, 2019, 20(3): 405-416. doi: 10.1007/s11769-019-1041-3
    [8] ZHU Xiaohua, ZHAO Yingshi, FENG Xiaoming.  A Methodology for Estimating Leaf Area Index by Assimilating Remote Sensing Data into Crop Model Based on Temporal and Spatial Knowledge . Chinese Geographical Science, 2013, 23(5): 550-561. doi: 10.1007/s11769-013-0621-x
    [9] GONG Huili, JIAO Cuicui, ZHOU Demin, LI Na.  Scale Issues of Wetland Classification and Mapping Using Remote Sensing Images: A Case of Honghe National Nature Reserve in Sanjiang Plain, Northeast China . Chinese Geographical Science, 2011, 21(2): 230-240.
    [10] DU Peijun, YUAN Linshan, XIA Junshi, et al..  Fusion and Classification of Beijing-1 Small Satellite Remote Sensing Image for Land Cover Monitoring in Mining Area . Chinese Geographical Science, 2011, 21(6): 656-665.
    [11] WANG Ning, ZHANG Gang, LIU Te, CAI Qingxiang.  Distribution of Methyl Mercury in Rana chensinensis and Environmental Media in Gold-mining Areas of Upper Reaches of Songhua River, China . Chinese Geographical Science, 2010, 20(4): 330-336. doi: 10.1007/s11769-010-0405-5
    [12] YIN Hai-wei, KONG Fan-hua, LI Xiu-zhen.  RS AND GIS-BASED FOREST FIRE RISK ZONE MAPPING IN DA HINGGAN MOUNTAINS . Chinese Geographical Science, 2004, 14(3): 251-257.
    [13] DU Pei-jun, ZHOU Xing-dong, GUO Da-zhi.  SOME KEY ISSUES ON THE APPLICATION OF SATELLITE REMOTE SENSING TO MINING AREAS . Chinese Geographical Science, 2003, 13(1): 79-83.
    [14] XU Han-qiu.  AN ASSESSMENT OF LAND USE CHANGES IN FUQING COUNTY OF CHINA USING REMOTE SENSING TECHNOLOGY . Chinese Geographical Science, 2002, 12(2): 126-135.
    [15] HU Yuan-man, JIANG Yan, CHANG Yu, BU Ren-cang, LI Yue-hui, XU Chong-gang.  THE DYNAMIC MONITORING OF HORQIN SAND LAND USING REMOTE SENSING . Chinese Geographical Science, 2002, 12(3): 238-243.
    [16] LIU Ming-liang, ZHUANG Da-fang, LIU Ji-yuan.  FARMLAND AND URBAN AREA DYNAMICS MONITORING IN CHINA USING REMOTE SENSING AND SPATIAL STATISTICS METHODOLOGY . Chinese Geographical Science, 2001, 11(1): 42-49.
    [17] 庄大方, 凌扬荣, Yoshio Awaya.  INTEGRATED VEGETATION CLASSIFICATION AND MAPPING USING REMOTE SENSING AND GIS TECHNIQUES . Chinese Geographical Science, 1999, 9(1): 49-56.
    [18] 千怀遂.  REGIONALIZATION FOR LARGE AREA CROP ESTIMATES BY REMOTE SENSING ——A Case Study of Chinese Wheat . Chinese Geographical Science, 1998, 8(3): 204-211.
    [19] 黄铁青, 刘兆礼, 潘瑜春, 张养贞.  LAND COVER SURVEY IN NORTHEAST CHINA USING REMOTE SENSING AND GIS . Chinese Geographical Science, 1998, 8(3): 264-270.
    [20] 张养贞, 常丽萍, 张柏, 张树文, 黄铁青, 刘雅琴.  LAND RESOURCES SURVEY BY REMOTE SENSING AND ANALYSIS OF LAND CARRYING CAPACITY FOR POPULATION IN TUMEN RIVER REGION . Chinese Geographical Science, 1996, 6(4): 342-350.
  • 加载中
图(8) / 表ll (4)
计量
  • 文章访问数:  495
  • HTML全文浏览量:  187
  • PDF下载量:  21
  • 被引次数: 0
出版历程
  • 收稿日期:  2022-05-14
  • 录用日期:  2022-07-15
  • 网络出版日期:  2022-08-30
  • 刊出日期:  2022-09-05

The Massive Expansion and Spatial Transformation of Potentially Contaminated Land Across China in 1990–2020 Observed from Remote Sensing and Big-data

doi: 10.1007/s11769-022-1300-6
    基金项目:  Under the auspices of the National Key Research and Development Program (No. 2018YFC1800103, 2018YFC1800106)
    通讯作者: KUANG Wenhui. E-mail: kuangwh@igsnrr.ac.cn

English Abstract

DOU Yinyin, GUO Changqing, KUANG Wenhui, CHI Wenfeng, LEI Mei, 2022. The Massive Expansion and Spatial Transformation of Potentially Contaminated Land Across China in 1990–2020 Observed from Remote Sensing and Big-data. Chinese Geographical Science, 32(5): 776−791 doi:  10.1007/s11769-022-1300-6
Citation: DOU Yinyin, GUO Changqing, KUANG Wenhui, CHI Wenfeng, LEI Mei, 2022. The Massive Expansion and Spatial Transformation of Potentially Contaminated Land Across China in 1990–2020 Observed from Remote Sensing and Big-data. Chinese Geographical Science, 32(5): 776−791 doi:  10.1007/s11769-022-1300-6
    • A series of social and environmental issues arising from industrial activities has received growing attentions (Sonter et al., 2017; Werner et al., 2020; Luckeneder et al., 2021). Recognizing and understanding the spatiotemporal characteristics and hotspots of industrial and mining areas are key parts of ecological environment construction (Palmer et al., 2010; Sonter et al., 2020), and are essential in attaining the sustainable development goals (SDGs) by 2030 (UN, 2015; Endl et al., 2021). Roughly 1% of global land is affected by industrial and mining activities (Walker, 1999). For example, remote sensing monitoring has found the area of global mine lots to be 57 277 km2 (Maus et al., 2020). Industrial and mining activities could leave a hugely negative impact on the environment by not only severely deteriorating or devastating the natural ecosystem (Sonter et al., 2017; Luckeneder et al., 2021), but also adversely affecting human health with water and air pollutants (Palmer et al., 2010; Rojano et al., 2018). Therefore, it is imperative to recognize and clarify the spatiotemporal patterns of potentially contaminated land (PCL) and identify their hotspots, which will provide an important basis for ecological governance and control.

      With the increasing development of industrialization in China, the number and scale of PCL are growing rapidly, bringing about a series of environmental pollutions and public health risks (Yang et al., 2014; Li et al., 2021; Song et al., 2021). In particular, serious problems include the heavy metal pollution of soil around industrial and mining processes (Wang et al., 2011). According to a general survey of soil pollution in China, 20%–37% of the soil positioned around PCL across all of China suffered from excessive pollution, and approximately 2 × 107 ha farmland soils were contaminated, resulting in an economic loss of up to 2 billion yuan (RMB) during 2005–2013 (MEE, 2014). Therefore, timely recognition of the spatial patterns and hotspots and rapid relocation of China’s PCL are essential to attaining the SDGs.

      Numerous methods have been applied recently in information acquisition from and mapping of PCL (Rampanelli et al., 2021; Sonter et al., 2020; Tang et al., 2021). For example, the traditional field investigation (Nortcliff, 2001), the combination of remote sensing image interpretation and actual measurement investigation (Kimijima et al., 2021), the object-oriented method (Zeng et al., 2017), and deep learning (Gallwey et al., 2020) have provided vital techniques and technologies for timely and accurately acquiring information about PCL. However, due to current technological limitations, most acquisition methods are targeted at acquiring information from a single type of area (Peng et al., 2019; Maus et al., 2020), such as coal mining areas (Pericak et al., 2018) and metal mining areas (Owusu-Nimo et al., 2018). With the burgeoning development in GIS techniques, location-based service technology, and web crawler technology, the development in satellite remote sensing and big-data technologies has provided novel techniques and methods for solving this problem (Boldi et al., 2004; Pericak et al., 2018). These approaches have overcome the disadvantages of heavy workload and time consumption in traditional position sampling while demonstrating the advantages of speed, economy, and fineness in PCL recognition (Pericak et al., 2018). Numerous scholars have used web crawler technology for data crawling and visualized analysis (Li et al., 2010; Kausar et al., 2013; Kim et al., 2019). Therefore, the combination of remote sensing technology with web crawler and/or other technologies can not only acquire more detailed information about the types of areas but also provide favorable technical support for quickly recognizing PCL.

      Some scholars have researched the spatiotemporal characteristics of industrial and mining areas at the global, regional, national, and more diverse scales (Bernhardt et al., 2012; Sonter et al., 2017; Werner et al., 2020; Ahmed et al., 2021;Cribari et al., 2021; Saley et al., 2021; Tang et al., 2021). For example, they have performed analyses and mapping of the distributions of mining areas at global scale (Werner et al., 2020), analyzed the mining land extension in sub Saharan Africa at regional scale (Ahmed et al., 2021), and analyzed the evolution characteristics of the gold mine areas in Niger (Saley et al., 2021). However, recognition of spatiotemporal patterns and hotspots of different types of PCL at different scales in China remains unclear, given that the existing studies are mostly targeted at the evolution of areas at a single scale and of a single type (Li et al., 2020; Tang et al., 2021). For example, some have analyzed the pattern of potentially contaminated sites in China at the national scale (Jiang et al., 2021), some have analyzed spatiotemporal variations in coal mining and restoration and their ecological environment effects in the Qinghai-Tibet Plateau (QTP) at regional scale (Yuan et al., 2021), and some have analyzed the spatiotemporal evolution of the mine lots in six counties of Gansu Province, China (Li et al., 2020). Therefore, it is urgent to grasp the spatiotemporal evolution characteristics and hotspots of different types of PCL across China at multiple scales.

      Our study aimed to analyze the spatiotemporal patterns and hotspots of China’s PCL during 1990–2020 based on acquisition of the information about China’s PCL. First, we acquired the dataset of China’s PCL in 1990, 2000, 2010, and 2020 based on multisource data fusion technology by using high-resolution remote-sensing images, a land-use/cover change database, crawler data from websites, Open Street Map (OSM) data, and field investigation data. Second, we analyzed the spatiotemporal patterns and hotspots of China’s PCL during 1990–2020. Third, we examined the influencing factors of PCL relocation and their influences on the ecological environment. This study provides scientific information for ecological governance.

    • Considering the characteristics and disparities across China’s geographic regions, we divided the whole territory of China into six zones for better depiction of the spatiotemporal evolution and migration patterns of China’s PCL. The six geographical zones include Northeast China (NE), North China (N), Southeast China (SE), Southwest China (SW), Northwest China (NW), and Qinghai-Tibet Plateau (QTP) (Kuang et al., 2022). Taiwan Province of China was excluded due to data acquisition limitations (Fig. 1).

      Figure 1.  Study areas and six types of potentially contaminated land in China. EE, energy exploitation; EP, energy processing; CRME, chemical raw material exploitation; MSP, metal smelting & processing; MOMD, metal ore mining & dressing; CP, chemical processing

      The data sources contain six major types, including a land-use/cover change database, OSM data, data from web crawler, high-resolution remote-sensing images, field investigation data, and other auxiliary data (Fig. 1). Among them, China’s Land Use/cover Dataset (CLUD) was downloaded from the Institute of Geographic Sciences and Natural Resources Research, Chinese Academy of Sciences (http://www.igsnrr.ac.cn/). This dataset contains the CLUD vector data from 1990, 2000, 2010, and 2020, and includes six Level-I types and more than 25 Level-II types, with data accuracy higher than 92% (Kuang et al., 2016; Ning et al., 2018). The OSM data of 2020 was obtained from the freely accessible website of OSM (https://www.openstreetmap.org). The high-resolution remote-sensing images mainly include GaoFen-2 images (GF-2) of 2016 and Google Earth images of 1990, 2000, 2010, and 2020. GF-2 images were obtained from the Institute of Remote Sensing and Digital Earth, Chinese Academy of Sciences (http://www.ceode.cas.cn/), with a spatial resolution of 1 m, whereas the Google Earth image data were downloaded from the website of Google Earth (http://google-earth.en.softonic.com/), with spatial resolutions of 0.5 m, 1.0 m, and 2.0 m. The data from a web crawler were predominantly in the directories of national industrial enterprises for 1990–2020, namely the public information crawled from the website of national pollutant emission permit management information (http://permit.mee.gov.cn/) based on web crawler technology (Chakrabarti et al., 1999). The crawled data form the spotty spatial distribution information about national industries and enterprises based on data cleansing and screening. The field investigation data contain geospatial position, type, and scale of areas from about 2200 sampling points acquired by the Institute of Geographic Sciences and Natural Resources Research, Chinese Academy of Sciences (Zhang et al., 2021). The auxiliary data contain the national administrative division data at a scale of 1 : 1 000 000, from the National Geomatics Center of China (http://www.ngcc.cn/).

    • The industrial ore mining areas in this research included open-pit mines, ore dressing infrastructures, waste ore heaps, and ore tailing storage facilities, which can be delimited in the publicly accessible Google Earth images through visual interpretation (Werner et al., 2020). The industrial ore processing areas are the sites with resource exploitation and processing as the leading industry, including industrial parks and their ancillary facilities. First, the initial boundaries of mining and processing land are fused with the point distribution about industries and enterprises based on high-resolution remote sensing images, namely the Google Earth images (supplemented by GF-2 and Landsat images). The spatial relationship between the initial PCL boundary (mining land and processing land) and the point distributions of industrial enterprises was obtained using overlay spatial analysis method. Second, the industry type with the largest number of industrial enterprises in the patch is classified as the types of PCL. Finally, the PCL are divided into two Level-I types and six Level-II types, referring to relevant studies (Kuang et al., 2016; Ning et al., 2018; Li et al., 2021; Rampanelli et al., 2021) and existing taxonomical systems ‘Land-Use Status Classification (GB/T 21010–2017)’ and ‘National Industries Classification (GB/T 4754–2017)’. Level-I types include two types—mining and processing—whereas Level-II types include six types: metal ore mining and dressing, metal smelting and processing, energy exploitation, energy processing, chemical raw material exploitation, and chemical processing (Table 1, Fig. 1).

      Table 1.  Classification of potentially contaminated land (PCL)

      Level-I typeLevel-II typeMain content
      Mining Metal ore mining & dressing Areas for minerals predominated by ferrous metals, nonferrous metals and other metal mineral resources, including the mining area, tailings reservoir, stacking yard, dumping site, and their ancillary facilities
      Energy exploitation Areas for minerals predominated by coal, coalbed methane, bone coal, oil shale, petroleum, natural gas, and other fossil fuel resources, including the mining area, tailings reservoir, stacking yard, dumping site, and their ancillary facilities
      Chemical raw material exploitation Areas for minerals predominated by chemical, fertilizer minerals, and other nonmetal chemical raw material mineral resources, including the mining area, tailings reservoir, stacking yard, dumping site, and their ancillary facilities
      Processing Metal smelting & processing Contiguous regions where the areas dominated by the smelting and processing industry of ferrous, nonferrous, and other metals are centralized
      Energy processing Contiguous regions where the areas dominated by the processing industry of coal, crude oil, and other fossil fuels are centralized
      Chemical processing Contiguous regions where the areas dominated by the manufacturing industry of inorganic chemicals, organic chemicals, fertilizers, pesticides, and other chemical raw materials and products are centralized
    • To implement scientific spatial measurement of PCL, remote sensing recognition of PCL was carried out by dividing the taxonomical system to acquire the vector datasets about China’s PCL during 1990–2020. In accordance with our previous research (Guo et al., 2022), the acquisition of these data includes two links (mining land and processing land) and three steps (Fig. 2). First, the initial boundaries of mining land were obtained from the rural industrial land (Zhang et al., 2021) extracted by CLUD. The initial boundary information about processing land was sourced from the land use extracted by OSM based on keyword retrievals (industry, park, development zone, etc.) (Guo et al., 2022). The network data were crawled using web crawler technology, and the spotty distribution information about industries and enterprises (mainly including enterprise name, license or registration number, address of production and operation site, geographical location, industry category, establishment date, business term, and main pollutant category) was acquired after data cleansing, screening, and spatialization. Next, the multisource data fusion based on high-resolution remote-sensing images, namely Google Earth images (supplemented by GF-2 and Landsat images) and the spotty distribution information about industries and enterprises, traversed the initial PCL boundaries (mining land and processing land) and delimited the land patch type and integrated attribute information based on the industry type with the largest number of industrial enterprises in the patch. In this process, the PCL boundary was revised through visual interpretations to obtain high-precision site boundary information (Guo et al., 2022). Finally, based on previous research (Zhang et al., 2014, 2021; Guo et al., 2022), the accuracy of the data about China’s PCL was verified by using high-resolution images combined with the field investigation data.

      Figure 2.  Flow chart of the study

    • To better describe the evolution and migration patterns at different scales of PCL, the areas were classified into three types: small-scale areas (< 1 km2), medium-scale areas (1–10 km2), and large-scale areas (>10 km2). The changes in PCL were analyzed in terms of such indicators as the speed of change, proportion of change, and intensity of change.

    • The speed of change of PCL (IMS) refers to the annual average variation in size of PCL within a specific time interval, as given by the formula below:

      $$ {I M S}_{{t}_{1}-{t}_{2}}=\frac{{I M S}_{{t}_{2}}-{I M S}_{{t}_{1}}}{{t}_{2}-{t}_{1}} $$ (1)

      where $ {IMS}_{{t}_{1}-{t}_{2}} $ is the speed of area change (km2/yr) of PCL within the interval from t1 to t2; $ {IMS}_{{t}_{2}} $ and $ {IMS}_{{t}_{1}} $ are the areas (km2) of PCL at points-in-time t2 and t1, respectively.

    • The proportion of change of PCL (IMP) refers to the proportion of the variation in area of PCL within the specific time interval over the area of PCL at the start year, as given by the formula below:

      $$ {I M P}_{{t}_{1}-{t}_{2}}=\frac{{I M S}_{{t}_{2}}-{I M S}_{{t}_{1}}}{{I M S}_{{t}_{1}}}\times 100\% $$ (2)

      where $ {IMP}_{{t}_{1}-{t}_{2}} $ is the proportion of change (%) of PCL within the interval from t1 to t2.

    • The intensity of change of PCL (IMI) refers to the proportion of variation in area of PCL over the 10 km × 10 km grid cells (100 km2) within the specific time interval, as given by the formula below (Guo et al., 2022):

      $$ {I M I}_{{t}_{1}-{t}_{2}}=\frac{{I M S}_{{t}_{2}}-{I M S}_{{t}_{1}}}{100\;}\times 100\% $$ (3)

      where $ {IMI}_{{t}_{1}-{t}_{2}} $ is the intensity of change (%) of PCL within the interval from t1 to t2.

    • The distribution aggregation effect was further detected in the county-level dimension to identify the distribution hotspots of China’s PCL. The distribution heat index data of PCL were acquired for analysis on the hotspot distribution characteristics of China’s PCL using the Adaptive Spatial Clustering Algorithm based on Delaunay Triangulation (ASCDT) method (Deng et al., 2011).

      The kernel density estimation of PCL was performed with optimized ASCDT. Compared with traditional kernel density estimation algorithms (Okabe et al., 2009), ASCDT lays more emphasis on the spatial interrelation and morphological distribution of areas (Deng et al., 2011). Based on optimized ASCDT, we obtained the values of spatial distribution density of point features of PCL across China, and calculated the heat index of PCL. The formula is as follows:

      $$ \;{R}_{i}=\frac{{D}_{i}-{D}_{\mathrm{m}\mathrm{i}\mathrm{n}}}{{D}_{\mathrm{m}\mathrm{a}\mathrm{x}}-{D}_{\mathrm{m}\mathrm{i}\mathrm{n}}} $$ (4)

      where Ri is the value of the spatial distribution heat index of PCL corresponding to the ith grid cell, ranging within 0–1; Di is the value of distribution density of PCL corresponding to the ith grid cell; Dmin and Dmax refer to the minimum and maximum of all values of Di, respectively.

      The verification data of hotspots of China’s PCL was obtained from the field survey data and the data on the concentration area of PCL publicly released by the Ministry of Ecology and Environment of the People’s Republic of China. We compared the verification points with the identification of hotspots and used the ratio of the number of the verification points in hotspot areas to the total number of verification points as the verification result of the hotspots of China’s PCL.

      The overall accuracy of the data about PCL was verified by using random sampling combined with the ground survey method, applying high-resolution images and field investigation data. The accuracy assessment of China’s PCL in 1990, 2000, 2010 and 2020 were selected (the number of samples was about 10% of the total data) 2220, 2731, 3710, and 3121 sampling points for accuracy verification, respectively, and the accuracy assessment was carried out using the correct rate and transfer matrix. The correct rate is the ratio of correctly identified type samples to the total number of samples.

    • The accuracy of the PCL dataset was reassessed in six geographical zones of China based on previous research (Guo et al., 2022). With random selection and field survey data, we generated 2220, 2731, 3710, and 3121 sampling points from among the data products about China’s PCL in 1990, 2000, 2010, and 2020, respectively, and carried out data accuracy evaluation (Fig. 3). The accuracies of the products were 92.61%, 92.60%, 92.99%, and 94.62%, respectively, and the overall accuracy of the data products about China’s PCL reached 93.21% (Table 2).

      Figure 3.  Spatial distribution of sampling points in China. Not including Taiwan Province of China due to data acquisition limitations

      Table 2.  Accuracy of potentially contaminated land (PCL) in different geographical zones in different years

      Zone in ChinaIndex1990200020102020Overall accuracya
      North Total number of samples 643 831 1030 918 93.87
      Correct number of samples 594 774 967 882
      Incorrect number of samples 49 57 63 36
      Producer’s accuracy / % 92.38 93.14 93.88 96.08
      Northeast Total number of samples 294 380 400 334 95.08
      Correct number of samples 277 357 383 322
      Incorrect number of samples 17 23 17 12
      Producer’s accuracy / % 94.22 93.95 95.75 96.41
      Northwest Total number of samples 266 336 436 428 92.11
      Correct number of samples 246 308 397 399
      Incorrect number of samples 20 28 39 29
      Producer’s accuracy / % 92.48 91.67 91.06 93.22
      Southeast Total number of samples 541 603 973 789 91.06
      Correct number of samples 485 549 887 729
      Incorrect number of samples 56 54 86 60
      Producer’s accuracy / % 89.65 91.04 91.16 92.40
      Southwest Total number of samples 432 530 807 599 94.43
      Correct number of samples 413 493 758 570
      Incorrect number of samples 19 37 49 29
      Producer’s accuracy / % 95.60 93.02 93.93 95.16
      Qinghai-Tibet Plateau Total number of samples 44 51 64 53 93.54
      Correct number of samples 41 48 58 51
      Incorrect number of samples 3 3 6 2
      Producer’s accuracy / % 93.18 94.12 90.63 96.23
      Nationwide Total number of samples 2220 2731 3710 3121 93.21
      Correct number of samples 2056 2529 3450 2953
      Incorrect number of samples 83 114 96 62
      Producer’s accuracy / % 92.61 92.60 92.99 94.62
      Note: a Overall accuracy is the average of data accuracies of 1990, 2000, 2010, and 2020

      In the geographical zones, the overall accuracies were 93.87%, 95.08%, 92.11%, 91.06%, 94.43%, and 93.54% for N, NE, NW, SE, SW, and QTP, respectively. Among them, the SE and NW zones had relatively low accuracies, both falling below the national average (Table 2). Two main types of problems may exist with our multimethod. First, in the arid and semi-arid climate zones, it is difficult to distinguish the subsurface of potentially polluted sites from desert and bare soil, which easily leads to the phenomenon of mapping errors and omissions, resulting in a low average accuracy of the NW zone. Second, it is difficult to accurately map the energy processing sites interspersed with urban land and rural settlements in the high-speed economic development areas, yielding a relatively low overall accuracy of the SE zone.

    • The area of China’s PCL is 13 856.08 km2, over which the chemical processing areas account for approximately one-third (Table 3). In 2020, China’s PCL accounted for 0.14% of the country’s total territorial area. Within these areas, chemical processing took up the largest area of 4852.11 km2, accounting for 35.02% and was distributed mostly in the SE and N zones (Fig. 4).

      Table 3.  Variations in area of China’s potentially contaminated land (PCL) during 1990–2020

      TypeScale of areaChanges in size of area / km2Changes in size of area per year / (km2/yr)
      20201990–20002000–20102010–20201990–20201990–2020
      Metal ore mining & dressing Large 241.84 36.67 34.91 77.28 148.86 4.96
      Medium 548.32 207.40 137.49 −276.60 68.29 2.28
      Small 1089.88 366.11 727.22 −963.05 130.28 4.34
      Energy exploitation Large 433.11 122.45 92.73 87.02 302.20 10.07
      Medium 603.87 90.90 234.50 −57.50 267.90 8.93
      Small 649.31 218.03 356.10 −450.22 123.91 4.13
      Chemical raw material exploitation Large 527.58 1.64 224.47 131.88 357.99 11.93
      Medium 73.43 23.56 7.21 −13.96 16.81 0.56
      Small 152.56 59.34 113.49 −147.80 25.03 0.83
      Energy processing Large 534.24 107.53 226.53 142.69 476.75 15.89
      Medium 66.61 11.51 44.17 4.76 60.44 2.01
      Small 3.22 1.77 1.02 0.43 3.22 0.11
      Metal smelting & processing Large 3043.14 251.74 1225.74 623.55 2101.03 70.03
      Medium 1010.85 223.18 350.63 295.52 869.33 28.98
      Small 26.01 13.96 1.34 8.28 23.58 0.79
      Chemical processing Large 3550.49 318.48 1233.89 1057.47 2609.84 86.99
      Medium 1263.78 277.38 453.41 368.55 1099.34 36.64
      Small 37.84 23.95 2.32 4.31 30.58 1.02
      All Nationwide 13,856.08 2355.60 5467.17 892.61 8715.38 290.51

      Figure 4.  Spatial patterns of different types of PCL across China in 2020. Not including Taiwan Province of China due to data acquisition limitations. CP, chemical processing; MSP, metal smelting and processing; EP, energy processing; MOMD, metal ore mining and dressing; EE, energy exploitation; CRME, chemical raw material exploitation

      China’s PCL areas are significantly different in size, with small-scale areas predominately in mining and large-scale areas predominately affected by processing (Fig. 4, Table 3). In 2020, the total of small-scale areas was 1958.82 km2, accounting for 14.14% of the total area of PCL; moreover, metal ore mining and dressing areas and energy exploitation areas preponderated, with areas of 1089.88 km2 and 649.31 km2, respectively. While the metal ore mining and dressing areas were distributed mostly in the SE, N, and NW zones, the energy exploitation areas were centralized in the NW and N zones (Fig. 4). The total of large-scale areas was 8330.40 km2, accounting for 60.12% of the total area of PCL, predominated by chemical processing (3550.49 km2) and metal smelting and processing (3043.14 km2), and centralized in the SE and N zones (Fig. 4, Table 3).

    • China’s PCL has been in overall growth for the last three decades, especially during 2000–2010, although with significant differences between the ten-year intervals (Fig. 5, Table 3). The area of PCL has increased by 8715.38 km2 at a rate of 290.51 km2/yr, and the increased area mainly stems from the rapid development of chemical processing and metal smelting and processing. The decade 2000–2010 saw the highest growth rate in PCL area, with 59.76% of the increased area contributed from chemical processing and metal smelting and processing. The area of China’s PCL increased from 5140.70 km2 to 7496.30 km2 during 1990–2000. After 2000, China’s accession to the WTO and the implementation of China’s Western Development Strategy further boosted the growth rate of PCL. The speed of change was 546.72 km2/yr during 2000–2010, 2.32 times that during 1990–2000. After 2010, China released and implemented a series of policies and suggestions on environmental protection and pollution control, strengthening the governance over industrial ore mining and processing. During 2010–2020, the mining areas across China decreased by an average of 161.30 km2/yr.

      Figure 5.  Intensity of change of China’s potentially contaminated land (PCL) in different regions during 1990–2020. Not including Taiwan Province of China due to data acquisition limitations. CRME, chemical raw material exploitation; CP, chemical processing; EE, energy exploitation; EP, energy processing; MOMD, metal ore mining and dressing; MSP, metal smelting and processing

      During the past 30 yr, China’s PCL has transformed progressively from predominately small- and medium-scale mining to large-scale processing (Fig. 6, Table 3). The proportion of the area of small- and medium-scale areas over the total area of PCL dropped while the proportion of large-scale areas rose during 1990–2020. Among them, the proportion of the size of small- and medium-scale areas dropped from 54.60% in 1990 to 39.88% in 2020, with the greatest contribution from the mining areas, especially from metal ore mining and dressing, whose proportion dropped from 18.67% to 7.78%. On the contrary, the proportion of large-scale areas in the same period rose from 45.40% to 60.12%, with considerable contribution from the processing areas. The proportion of the area of large-scale processing rose from 37.74% to 51.44% during 1990–2020, with the major contribution from the chemical processing and metal smelting and processing areas, whose proportions rose by 6.69 and 4.16 percentage points, respectively (Fig. 6, Table 3).

      Figure 6.  Evolution of different types of potentially contaminated land (PCL) across China during 1990–2020. CP, chemical processing; MSP, metal ore smelting and processing; EP, energy processing; CRME, chemical raw material exploitation; EE, energy exploitation; MOMD, metal ore mining and dressing

      The spatial pattern of different types of PCL presents distinctive regional variation (Fig. 7). During 1990–2020, medium- and large-scale metal smelting and processing areas transformed progressively from being distributed mostly in the N zone to being distributed mostly in the SE zone (Fig. 7b). The proportion of the area of medium- and large-scale metal smelting and processing areas in the N zone over the total area of metal smelting and processing areas across China dropped from 36.34% in 1990 to 34.48% in 2020; in the same period, the proportion in the SE zone rose from 4.87% to 40.05%, and the size of areas increased from 41.74 km2 to 2238.19 km2. Medium- and large-scale energy processing areas expanded from the N and SE zones into the NE and NW zones (Fig. 7d). In 1990, these areas constituted proportions of 1.88% and 2.48%, respectively. In 2020, their proportions had increased by 2.80% and 1.56% in the N and SE zones, respectively, and had appeared and risen by 2.37% and 6.82% in the NE and NW zones, respectively. Large-scale chemical raw material exploitation areas were relocated to the NW zone, while medium- and large-scale chemical processing areas based in the N zone were shifted to the SE zone (Fig. 7e). The large-scale chemical raw material exploitation areas in the NW zone took up a proportion of 14.75% by 2020. The proportion of medium- and large-scale chemical processing areas in the N zone rose from 34.19% in 1990 to 37.04% in 2020, while the proportion in the SE zone rose from 9.57% to 43.84%. Metal ore mining and dressing areas have not presented distinctive variation of pattern since 1990 (Fig. 7a). Small- and medium-scale energy exploitation areas appeared to have been relocated progressively to the NW (Fig. 7c), with their area increasing from 206.27 km2 in 1990 to 295.39 km2 in 2020.

      Figure 7.  Area and distribution of China’s potentially contaminated land in different zones during 1990–2020. SW, Southwest; NW, Northwest; QTP, Qinghai-Tibet Plateau; N, North; SE, Southeast; NE, Northeast

    • With field survey data and the data of PCL publicly released by the environmental department, we generated 1691, 2211, 3806, and 2317 verification points from among the data products about the hotspot identification of China’s PCL in 1990, 2000, 2010, and 2020, and carried out data accuracy evaluation. The accuracies of hotspots of China’s PCL were 86.52%, 86.75%, 87.23%, and 89.17% (Table 4).

      Table 4.  Accuracy of the hotspot of China’s potentially contaminated land (PCL) in different geographical zones in different years

      Zone in ChinaTotal number of samplesProducer’s accuracy / %
      19902000201020201990200020102020
      North 469 583 950 610 86.43 86.79 87.30 89.30
      Northeast 137 190 297 181 87.10 86.32 86.98 88.79
      Northwest 55 82 165 103 87.88 85.37 86.87 89.03
      Southeast 649 830 1481 958 86.44 86.87 87.31 88.94
      Southwest 373 516 903 455 86.86 87.02 87.38 89.45
      Qinghai-Tibet Plateau 8 10 10 10 75.00 80.00 80.00 90.00
      Total 1691 2211 3806 2317 86.52 86.75 87.23 89.17

      China’s PCL featured distinctive spatial agglomeration in 2020, with their hotspots agglomerated mostly in the N, SE, and SW zones (Fig. 8). Meanwhile, the spatial distribution of hotspots had considerable differences between different types of PCL. In 2020, 149 county-level administrative regions were dominated by chemical raw material exploitation, accounting for 5.24% and centralized mainly in Sichuan Province, Yunnan Province, Guizhou Province, Fujian Province, and Inner Mongolia Autonomous Region.

      Figure 8.  Hotspot distributions of potentially contaminated land (PCL) in China during 1990–2020. Not including Taiwan Province of China due to data acquisition limitations. MOMD, metal ore mining and dressing; CP, chemical processing; EE, energy exploitation; MSP, metal smelting and processing; CRME, chemical raw material exploitation; EP, energy processing

      China’s PCL has rendered an overall intensified tendency of agglomeration over a markedly enlarged sphere since 1990, as well as a phenomenon of transferring progressively to the NW, SW, and QTP zones (Fig. 8). During 1990–2020, China’s PCL rendered a further intensified tendency of agglomeration among the city cluster in the middle reach of the Yangtze River, the Guangdong-Hong Kong-Macao Greater Bay Area, and other developed regions. The distribution of PCL showed a tendency of being relocated to Baotou-Ordos-Yulin city cluster and other northwestern regions. Among all PCL, the large-scale energy processing areas transferred to the QTP zone, centralized in Dagze District, Maizhokunggar County, and Ngagzha County of Tibet Autonomous Region, with their area increasing by 6963.83 km2. In the same period, energy exploitation areas had a significant agglomeration effect in the SW zone across 12 counties, such as Nanhua and Yaoan. Meanwhile, the agglomeration sphere of metal smelting and processing areas in the SW zone reached 166 counties, and their area increased by 2993.94 km2.

    • Identifying and obtaining the boundary information about PCL accurately and in a timely manner on a large scale can provide an important data basis for effectively detecting and quantifying the spatiotemporal patterns of PCL. The vector datasets about China’s PCL of 1990, 2000, 2010, and 2020 have been developed with remote sensing monitoring over the PCL in satellite images. The geographical databases that were set up include 100 163 vector diagram spots of area boundaries. Compared with the existing data developed and used by research institutes (Jiang et al., 2021), our developed datasets about China’s PCL can represent the spatiotemporal patterns of PCL with more details. Our datasets contain not only the information about the positions of areas, such as the longitude and latitude, the names and types of areas, but also the spatial range attributes of area vectors. Meanwhile, the overall accuracy of these PCL databases are above 93% (Table 2), satisfying the requirement for area-related spatiotemporal analysis and regional comparative study. On the other hand, based on the existing vector datasets from CLUD, we have adopted web crawler, data fusion, human-computer interaction interpretation, and other technologies and approaches to acquire the area boundary information, which has shortened data acquisition time and significantly reduced workload and difficulty (Chakrabarti et al., 1999; Kuang et al., 2022). Furthermore, most of the currently released land-use/cover change data products in a large-scale scope (Chen et al., 2015; Liu et al., 2018; Gong et al., 2019, 2020; He et al., 2019; Zhang et al., 2020) have included PCL in such types as urban built-up area, impervious surface, and mining area. Due to the failure to acquire more detailed information such as related enterprises and environments, the above-mentioned publicly released data products have not recognized different types of PCL.

    • The spatial pattern and migration of China’s PCL are joint driving effects of policy, mineral resources, economy, and other factors over the past 30 years, among which policy and economy factors have contributed more prominently to the long time-sequence transition of areas. In the 1990s, small-scale PCL were relatively prominent (average area 0.29 km2) (Fig. 6, Fig. 7), which might be ascribed to plateau landforms, illegal mining, or the mining of small-scale scattered metallic ore deposits (Tang et al., 2020). The quantity and nature of keener oversight have increased the mining costs. Under this circumstance, only the larger-area, high-value ores can be mined economically, driving this industry towards operations producing large-scale PCL (Werner et al., 2020). With the implementation of China’s Western Development Strategy, PCL have been relocated to the NW and SW zones, where medium- and large-scale metal ore processing and chemical processing areas have increased by 211.16 km2 and 600.34 km2, respectively (Fig. 7, Table 3); particularly, the western rural land for industrial ores surged at an alarming rate from 3%–10% in the 1990s to 115% during 2005–2015 (Zhang et al., 2021). After entering the 21st century, China joined the WHO. Under the influence of intensified international competition, the less competitive industries and enterprises were relocated progressively from eastern coastal regions to western regions. During that period, PCL exhibited the spatial characteristic of being relocated to the central and western regions (Fig. 5, Fig. 7). China has taken an active part in global energy governance and successively issued a series of policies and opinions related to ecological governance and supervision to cope with global climate change (SCPRC, 2013, 2015, 2016). Relevant policies have contributed to a significant increase of energy exploitation areas in the NW and QTP zones (Fig. 7). Furthermore, as China’s economic growth moves in the green and low-carbon direction, photovoltaic, wind energy, and other renewable energy sources have risen with strong potentials. The power sector is transforming from predominately fossil energy to renewable energy sources, and the area of large-scale energy exploitation areas in the QTP zone has increased dramatically (Fig. 7, Table 3).

    • Although the distribution of China’s industrial ore mining areas obeys the ultimate law of geological enrichment, these areas are centralized in places of population aggregation, lowland regions, and flat regions (Fig. 8) probably due to economic factors. Of the 3.9 million t of wastes generated from nonferrous metal processing in China in 2016, almost 80% came from the northwestern region, Yunnan Province, Inner Mongolia Autonomous Region, Gansu Province, Hunan Province, and Qinghai Province (NBS, 2021). Numerous hazardous wastes are generated amid the metal ore processing and smelting processes (Wang et al., 2021), severely threatening the atmosphere, water environment, other ecological environments, and human health. The coal mining activities in Inner Mongolia Autonomous Region have caused a sharp decrease in lake area: 64.6% of the sharp decrease in lakes resulted from the water consumption for coal mining (Tao et al., 2015). The threat of PCL might extend to a radius of 70 km from the mining locale (Sonter et al., 2017), leading to water and soil degradation along the river basin. The treatment zone of PCL in the upper reaches may cause a series of environmental issues to the lower reaches. The impact on Brahmaputra (or the Yarlung Zangbo Jiang) rivers might stem from the mining areas at extremely high altitudes in the QTP zone (MEE, 2014).

    • The vector datasets about China’s PCL at 1990, 2000, 2010, and 2020 have been developed using data fusion technology based on high-resolution remote sensing images, a land-use/cover change database, data from web crawler, and other multisource data. The spatial distribution and transfer pattern of China’s PCL have been analyzed in national-, regional-, and county-level scales, the hotspot distribution of these areas has been recognized, and the influencing factors for the variation of their pattern have been further discussed.

      The vector data products about China’s PCL developed and acquired based on multisource data fusion technology have high quality and reliability, with an overall accuracy of 93.21%. The overall area of China’s PCL has kept growing since 1990, with significant differences across different time intervals. During 2000–2010, the growth rate of areas was 2.32 and 6.13 times that during 1990–2000 and 2010–2020, respectively. The agglomeration of China’s PCL intensified and enlarged its sphere, and rendered the phenomenon of transferring from the N and SE zones to the NW, SW, and QTP zones during 1990–2020. The spatial pattern and transformation of China’s PCL are joint effects of policies, mineral resources, economy, and other factors over the past 30 yr, among which policy and economy factors have contributed more prominently to the long time-sequence transition of areas.

      The ecological-environmental problems in the PCL have attracted great attention in China. The results of this study provide scientific information about PCL and provide directions for local environmental protection departments and ecological governance. Subsequent research could begin with the comparation of transport models of different compounds and sites in different industries. The more detailed analyses of ecosystem services, such as comparing soil conservation and biodiversity protection functions with those of the pre-state to improve the ecological environment of the PCL in a more targeted way.

参考文献 (57)

目录

    /

    返回文章
    返回