准确掌握土壤重金属浓度的空间分布信息是开展土壤修复与治理工作的前提。传统方法费时、费力、代价高且极易造成对环境的二次污染，如何迅速高效获取土壤重金属含量信息已成为学术界研究热点与难点之一。高光谱遥感具有光谱分辨率高、代价低、不会对环境造成污染、波谱信息丰富且时空连续性强等特点，为高效开展土壤重金属监测提供了可能。本文以陕西省蒲白煤矿某复垦农田为研究区，基于实测光谱数据、原位土壤重金属含量数据、高分五号（GF-5）高光谱遥感影像，利用BP神经网络构建实测光谱数据和GF-5高光谱遥感影像反射率之间的光谱转换模型，得到校正后的GF-5反射率数据；采用连续小波变换（CWT）和分数阶导数（FOD）对校正后GF-5反射率进行光谱变换，引入Boruta算法提取两种超标重金属（Cr和Ni）的高光谱特征波段；在此基础上，引入偏最小二乘回归（PLSR）、极限学习机（ELM）和随机森林（RF）模型分别构建两种重金属含量（Cr和Ni）与特征波段之间的高光谱遥感反演模型，利用交叉验证方法和精度评价指标R²和RMSE评价反演模型精度以选取最优遥感反演模型；根据最优遥感反演模型，对GF-5高光谱影像进行波段运算得到研究区重金属Cr和Ni的空间分布图。全文主要研究结果如下：

（1）研究区土样中5种重金属Cr、Ni、Zn、Cu和Mn的平均值浓度分别为76.42mg/kg、27.40mg/kg、67.37mg/kg、19.57mg/kg和580.34mg/kg。Cr（62.50mg/kg）和Mn（557.00mg/kg）超过了陕西省土壤背景值，Cr（61.00mg/kg）和Ni（26.00mg/kg）超过了全国土壤背景值，因Mn多属于自然来源，故本文选择Cr和Ni这两种重金属进行高光谱遥感反演研究；从变异系数（CV）来看，Cu元素属于高度变异（63.42%），除Cu以外，所有元素含量均符合正态分布。

（2）Cr元素特征波段大多数集中在可见光及近红外区域，主要包括连续小波变换的第2尺度（CWT_2）和0.75阶分数阶导数变换（FOD_0.75），其中CWT_2的特征波段分别为484、497、514、660和1528nm，FOD_0.75的特征波段分别为891、899、904、921和959nm；Ni元素在CWT_2选取的特征波段数量最多，分别是814、981、985、1291、1974和2193nm。

（3）基于所选特征波段，利用PLSR、ELM和RF模型构建两种重金属的遥感反演模型，通过交叉验证和精度评价指标（R²和RMSE）对不同遥感反演模型进行精度评价，Cr的最优反演模型为CWT_2变换下的RF模型，其中建模精度为R²_c=0.97，RMSE_c=1.80，验证精度为R²_v=0.94，RMSE_v=2.66；Ni的最优遥感反演模型为连续小波变换的第8尺度（CWT_8）的ELM模型，其中建模精度为R²_c=0.83，RMSE_c=1.07，验证精度为R²_v=0.36，RMSE_v=4.56。

（4）基于最优反演模型，对GF-5高光谱影像进行波段运算得到研究区重金属Cr和Ni的空间分布图。并与空间插值结果进行对比后发现，整体分布趋势基本一致，其中Cr含量较高的区域主要分布在研究区的东南部，东北部和中部偏西区域，Ni含量较高的区域主要分布在研究区的东南部，其余位置均为低值区域，研究区Ni含量由东南到北部呈现降低的趋势，由东南部到西北部呈现先降低后增加的趋势。

It is a precondition for soil pollution prevention and control work to grasp accurately spatial distribution information of soil heavy metal. The traditional methods are time-consuming, laborious and costly, and it is easy for the environment to produce secondary pollution. It is one of research hotspots and difficulties about how to acquire soil heavy metal content efficiently and quickly. Hyperspectral remote sensing technology provides the feasibility to monitor soil heavy metal pollution due to many advantages, including high spectral resolution, low price, more bands and a strong continuity. It will not pollute the environment. We selected an abandoned reclamation region of a Pubai coal mine in Shaanxi province as the study area in this study. Soil samples are collected in field and spectral reflectance data and heavy metal are measured and Gaofen5 (GF-5) hyperspectral remote sensing image. Spectral transformation models are established between measured spectral data and GF-5 hyperspectral remote sensing image using BP neural network model to get corrected GF-5 reflectance data. Spectrum is transformed using continue wavelet transform (CWT) and fractional derivative (FOD) for corrected GF-5 reflectance data. Feature bands are selected based on tow heavy metals content using Boruta algorithm. The inversion models are calibrated between Cr and Ni elements content and feature bands reflectance using PLSR, ELM and RF, respectively. The optimal inversion model was selected through using cross validation and accuracy evaluation index (R² and RMSE) to evaluate model accuracy. According to the optimal inversion model, the spatial distribution of Cr and Ni content are mapped by band math of GF-5 hyperspectral remote sensing image. The main results in this study are as follows:

(1) The average content of Cr, Ni, Zn, Cu and Mn are 76.42mg/kg, 27.40mg/kg, 67.37mg/kg, 19.57mg/kg and 580.34mg/kg in study area. Compared with Shaanxi soil background value, Cr (62.50mg/kg) and Mn (557.00mg/kg) exceed Shaanxi soil background value. Compared with China soil background value, Cr (61.00mg/kg) and Ni (26.00mg/kg) exceed China soil background value. This study selects Cr and Ni to achieve Hyperspectral inversion research due to Mn belongs to natural source. According to variable coefficient, Cu belong to highly variable (63.42%). All elements meet normal distribution except Cu.

(2) A majority feature bands of Cr are located in visible and near-infrared light area, the most bands are selected in the second dimension of continue wavelet transform (CWT_2) and 0.75 order of fractional derivative (FOD_0.75). Feature bands of CWT_2 is 484, 497, 514, 660 and 1528nm, respectively, and FOD_0.75 is 891, 899, 904, 921 and 959nm. The most bands of Ni are selected in CWT_2, including 814, 981, 985, 1291, 1974 and 2193nm.

(3) According to selected feature bands, the remoting sensing inversion models of Cr and Ni are calibrated using PLSR, ELM, and RF. The different models are evaluated through cross validation and accuracy evaluation index (R² and RMSE). The optimal inversion model of Cr is RF of CWT_2 transformation, and the accuracy of calibration R²_c=0.97 and RMSE_c=1.80, the accuracy of verification R²_v=0.94 and RMSE_v=2.66. The optimal inversion model of Ni is ELM of CWT_8 transformation, R²_c=0.83, RMSE_c=1.07, R²_v=0.36, RMSE_v=4.56.

(4) Based on the optimal inversion model, the spatial distribution of Cr and Ni content are mapped by band math of GF-5 hyperspectral remote sensing image, and compare with spatial interpolation result. The results indicate that the total trend is consistent. The high area of Cr is in the southeast, the northeast and west-central region. The high content area of Ni is located in lower right corner of study area, and the other area is low value. The trend of Ni content is reduced from southeast to north, and reduced and increased from southeast to northwest.

[1] Hou D, O'Connor D, Nathanail P, et al. Integrated GIS and multivariate statistical analysis for regional scale assessment of heavy metal soil contamination: A critical review. Environmental Pollution, 2017, 231(pt.1): 1188-1200.

[2] Li P, Wu J, Qian H, et al. Origin and assessment of groundwater pollution and associated health risk: a case study in an industrial park, northwest China. Environmental Geochemistry & Health, 2014, 36(4): 693-712.

[3] Solgi E, Esmaili-Sari A, Riyahi-Bakhtiari A, et al. Soil Contamination of Metals in the Three Industrial Estates, Arak, Iran. Bulletin of Environmental Contamination & Toxicology, 2012, 88(4): 634-638.

[4] N R, M M, D P. Soil pollution: a hidden reality. 2018: FAO. 142.

[5] Alshahri F, El-Taher A. Assessment of Heavy and Trace Metals in Surface Soil Nearby an Oil Refinery, Saudi Arabia, Using Geoaccumulation and Pollution Indices. Archives of Environmental Contamination & Toxicology, 2018, 75: 390-401.

[6] Liu K, Wang F, Li J, et al. Assessment of trends and emission sources of heavy metals from the soil sediments near the Bohai Bay. Environmental Science and Pollution Research, 2019, 26(7): 29095–29109.

[7] Gall J E, Boyd R S, Rajakaruna N. Transfer of heavy metals through terrestrial food webs: a review. Environmental Monitoring and Assessment, 2015, 187(4): 1-21.

[8] Bergthorson J M, Goroshin S, Soo M J, et al. Direct combustion of recyclable metal fuels for zero-carbon heat and power (vol 160, pg 368, 2015). Applied Energy, 2017, 202(sep.15): 784-784.

[9] Li K, Gu Y, Li M, et al. Spatial analysis, source identification and risk assessment of heavy metals in a coal mining area in Henan, Central China. International Biodeterioration & Biodegradation, 2017, 128: 148-154.

[10] Ma W, Tai L, Zhi Q, et al. Contamination source apportionment and health risk assessment of heavy metals in soil around municipal solid waste incinerator: A case study in North China. Science of the Total Environment, 2018, 631–632(1): 348-357.

[11] Liu P, Zhang Y, Wu T, et al. Acid-extractable heavy metals in PM2.5 over Xi'an, China: seasonal distribution and meteorological influence. Environmental Science and Pollution Research, 2019, 26(33): 34357–34367.

[12] Kai, Zhang, Fahe, et al. Size distribution and source of heavy metals in particulate matter on the lead and zinc smelting affected area. Journal of Environmental Sciences, 2018, 71(09): 191-199.

[13] Li, Siyue, Jia, et al. Heavy metals in soils from a representative rapidly developing megacity (SW China): Levels, source identification and apportionment. Catena An Interdisciplinary Journal of Soil Science Hydrology Geomorphology Focusing on Geoecology & Landscape Evolution, 2018, 163: 414-423.

[14] Ren Z, Xiao R, Zhang Z, et al. Risk assessment and source identification of heavy metals in agricultural soil: a case study in the coastal city of Zhejiang Province, China. Stochastic Environmental Research and Risk Assessment, 2019, 33(11-12): 2109-2118.

[15] Zereini F, Alt F, Messerschmidt J, et al. Concentration and Distribution of Heavy Metals in Urban Airborne Particulate Matter in Frankfurt am Main, Germany. Environmental Science & Technology, 2005, 39(9): 2983-2989.

[16] Yang Z, Li B. Effects of Cd contamination and physiological and biochemical characteristics on Brassica juncea. Proceedings of 2011 International Conference on Remote Sensing, Environment and Transportation Engineering: 2011: 1147-1150.

[17] Gu Y W, Li S, Gao W, et al. Hyperspectral estimation of the cadmium content in leaves of Brassica rapa chinesis based on the spectral parameters. Acta Ecologica Sinica, 2015, 35(13): 4445-4453.

[18] Wang F, Gao J, Zha Y. Hyperspectral sensing of heavy metals in soil and vegetation: Feasibility and challenges. ISPRS Journal of Photogrammetry and Remote Sensing, 2018, 136: 73-84.

[19] Liu F. Ministry of Land and Mineral Resources of China. Acta Geologica Sinica, 2007, 81(2): 204-225.

[20] Zhang S, Shen Q, Nie C, et al. Hyperspectral inversion of heavy metal content in reclaimed soil from a mining wasteland based on different spectral transformation and modeling methods. Spectrochimica Acta Part A: Molecular & Biomolecular Spectroscopy, 2019, 211(15): 393-400.

[21] Wu Y, Chen J, Wu X, et al. Possibilities of reflectance spectroscopy for the assessment of contaminant elements in suburban soils. Applied Geochemistry, 2005, 20(6): 1051-1059.

[22] Cui L, Wu J, Ju H. Electrochemical sensing of heavy metal ions with inorganic, organic and bio-materials. Biosensors & Bioelectronics, 2015, 63: 276-286.

[23] Lu Y, Liang X, Niyungeko C, et al. A review of the identification and detection of heavy metal ions in the environment by voltammetry. Talanta, 2017, 178(1): 324-338.

[24] Soodan R K, Pakade Y B, Avinash, et al. Analytical techniques for estimation of heavy metals in soil ecosystem: A tabulated review. Talanta, 2014, 125(11): 405-410.

[25] Hou X, He Y, Jones B T. Recent advances in portable X-ray fluorescence spectrometry. Appl. Spectrosc. Rev, 2004, 39(1): 1-25.

[26] Weindorf D, Bakr N, Zhu Y. Advances in portable X-ray fluorescence (PXRF) for environmental, pedological, and agronomic applications. In Advances in Agronomy, New York: Academic Press, 2014, 128: 1-45.

[27] Gates W P. Chapter 12.3 X-ray Absorption Spectroscopy. Developments in Clay Science, 2006, 1(05): 789-864.

[28] Kaniu M I, Angeyo K H, Mwala A K, et al. Direct rapid analysis of trace bioavailable soil macronutrients by chemometrics-assisted energy dispersive X-ray fluorescence and scattering spectrometry. Analytica Chimica Acta, 2012, 729: 21-25.

[29] Butler O T, Cairns W, Cook J M, et al. Atomic spectrometry update. Environmental analysis. Journal of Analytical Atomic Spectrometry, 2011.

[30] Block C N, Shibata T, Solo-Gabriele H M, et al. Use of Handheld X-ray Fluorescence Spectrometry Units for Identification of Arsenic in Treated Wood. Environmental Pollution, 2007, 148(2): 627-633.

[31] Min J. Application of portable X-ray fluorescence (pXRF) for heavy metal analysis of soils in crop fields near abandoned mine sites. Environmental Geochemistry and Health, 2010, 32(3): 207-216.

[32] Radu T, Diamond D. Comparison of soil pollution concentrations determined using AAS and portable XRF techniques. Journal of Hazardous Materials, 2009, 171(1-3): 1168-1171.

[33] Kilbride C, Poole J, Hutchings T R. A comparison of Cu, Pb, As, Cd, Zn, Fe, Ni and Mn determined by acid extraction/ICP-OES and ex situ field portable X-ray fluorescence analyses. Environ. Pollut, 2006, 143(1): 16-23.

[34] Kruse F A, Boardman J W, Huntington J F. Comparison of airborne hyperspectral data and EO-1 Hyperion for mineral mapping. IEEE Transactions on Geoence and Remote Sensing, 2003, 41(6): 1388-1400.

[35] Shi T, Chen Y, Liu Y, et al. Visible and near-infrared reflectance spectroscopy-An alternative for monitoring soil contamination by heavy metals. Journal of Hazardous Materials, 2014, 265: 166-176.

[36] Mohamed E S, Ali A M, Shirbeny M A E, et al. Near infrared spectroscopy techniques for soil contamination assessment in the Nile Delta. Eurasian Soil Science, 2016, 49(6): 632-639.

[37] Kokaly R F, Clark R N. Spectroscopic Determination of Leaf Biochemistry Using Band-Depth Analysis of Absorption Features and Stepwise Multiple Linear Regression. Remote Sensing of Environment, 1999, 67(3): 267-287.

[38] Grzegorz S, Mccarty G W, Stuczynski T I, et al. Near and mid-infrared diffuse reflectance spectroscopy for measuring soil metal content. Journal of Environmental Quality, 2004, 33(6): 2056-2069.

[39] Vohland M, Bossung C, Fründ H. A spectroscopic approach to assess trace–heavy metal contents in contaminated floodplain soils via spectrally active soil components. Journal of Plant Nutrition & Soil Science, 2010, 172(2): 201-209.

[40] Singh A N. Estimation of As and Cu Contamination in Agricultural Soils Around a Mining Area by Reflectance Spectroscopy: A Case Study. Pedosphere, 2009, 19(6): 719-726.

[41] Wang J, Cui L, Gao W, et al. Prediction of low heavy metal concentrations in agricultural soils using visible and near-infrared reflectance spectroscopy. Geoderma, 2014, 216(4): 1-9.

[42] Zhang C, Kovacs J M. The application of small unmanned aerial systems for precision agriculture: A review. Precision Agriculture, 2012, 13(6): 693–712.

[43] Colomina I, Molina P. Unmanned aerial systems for photogrammetry and remote sensing: A review - ScienceDirect. ISPRS Journal of Photogrammetry and Remote Sensing, 2014, 92(2): 79-97.

[44] Faye E, Rebaudo F, Cajo D Y, et al. A toolbox for studying thermal heterogeneity across spatial scales: from unmanned aerial vehicle imagery tolandscape metrics. Methods in Ecology and Evolution, 2016, 7(4): 12488.

[45] Suomalainen J, Anders N, Iqbal S, et al. A Lightweight Hyperspectral Mapping System and Photogrammetric Processing Chain for Unmanned Aerial Vehicles. Remote Sensing, 2014, 6(11): 11013-11030.

[46] Paz-Kagan T, Zaady E, Salbach C, et al. Mapping the Spectral Soil Quality Index (SSQI) Using Airborne Imaging Spectroscopy. Remote Sensing, 2015, 7(11): 15748-15781.

[47] Kemper T, Sommer S. Use of airborne hyperspectral data to estimate residual heavy metal contamination and acidification potential in the Guadiamar floodplain Andalusia, Spain after the Aznacollar mining accident.Remote Sensing,2004.

[48] Choe E, Meer F, Ruitenbeek F V, et al. Mapping of heavy metal pollution in stream sediments using combined geochemistry, field spectroscopy, and hyperspectral remote sensing: A case study of the Rodalquilar mining area, SE Spain. Remote Sensing of Environment, 2008, 112(7): 3222-3233.

[49] Riaza A, Garcia-Melendez E, Mueller A. Spectral identification of pyrite mud weathering products: a field and laboratory evaluation. International Journal of Remote Sensing, 2011, 32(1-2): 185-208.

[50] Ben-Dor E, Chabrillat S, Demattê J, et al. Using Imaging Spectroscopy to study soil properties. Remote Sensing of Environment, 2009, 113: S38-S55.

[51] Lagacherie P, Mcbratney A B. Spatial soil information systems and spatial soil inference systems: perspectives for digital soil mapping. Developments in Soil Science, 2006, 31: 3-22.

[52] Mirik M, Norland J E, Crabtree R L, et al. Hyperspectral One-Meter-Resolution Remote Sensing in Yellowstone National Park, Wyoming: I. Forage Nutritional Values. Rangeland Ecology & Management, 2005, 58(5): 452-458.

[53] Iman T, Xu Z, Boyd S, et al. Laboratory-based hyperspectral image analysis for predicting soil carbon, nitrogen and their isotopic compositions. Geoderma, 2018, 330: 254-263.

[54] 胡玉玲. 基于HJ-1高光谱数据的矿区植被污染监测. 2011，山东科技大学.

[55] Yang L Y, Gao X H, Wei Z, et al. Estimating heavy metal concentrations in topsoil from vegetation reflectance spectra of Hyperion images: A case study of Yushu County,Qinghai,China. Ying yong sheng tai xue bao = The journal of applied ecology / Zhongguo sheng tai xue xue hui, Zhongguo ke xue yuan Shenyang ying yong sheng tai yan jiu suo zhu ban, 2016, 27(6): 1775-1784.

[56] Zhang T, Li L, Zheng B. Estimation of agricultural soil properties with imaging and laboratory spectroscopy. Journal of Applied Remote Sensing, 2013, 7(1): 073587-073587.

[57] 朱亚星，于雷，洪永胜，等. 土壤有机质高光谱特征与波长变量优选方法. 中国农业科学. 2017(22)：4325-4337.

[58] Leardi R, González A L. Genetic algorithms applied to feature selection in PLS regression: how and when to use them. Chemometrics & Intelligent Laboratory Systems, 1998, 41(2): 195-207.

[59] Araújo M C U, Saldanha T C B, Galvao R K H, et al. The successive projections algorithm for variable selection in spectroscopic multicomponent analysis. Chemometrics and Intelligent Laboratory Systems, 2001, 57(2): 65-73.

[60] Vohland M, Besold J, Hill J, et al. Comparing different multivariate calibration methods for the determination of soil organic carbon pools with visible to near infrared spectroscopy. Geoderma, 2011, 166: 198–205.

[61] Vohland M, Emmerling C. Determination of total soil organic C and hot water‐extractable C from VIS‐NIR soil reflectance with partial least squares regression and spectral feature selection techniques. European Journal of Soil Science, 2011, 62(4): 598-606.

[62] Sun W, Zhang X, Sun X, et al. Predicting nickel concentration in soil using reflectance spectroscopy associated with organic matter and clay minerals. Geoderma, 2018, 327: 25-35.

[63] Kancheva R, Georgiev G. Assessing Cd-induced stress from plant spectral response. Proceedings of SPIE - The International Society for Optical Engineering, 2014, 9239: 923926-923926-12.

[64] Dunagan S C, Gilmore M S, Varekamp J C. Effects of mercury on visible/near-infrared reflectance spectra of mustard spinach plants (Brassica rapa P.). Environmental Pollution, 2007, 148(1): 301-311.

[65] 夏芳，彭杰，王乾龙，等. 基于省域尺度的农田土壤重金属高光谱预测. 红外与毫米波学报. 2015，34(5)：593-598+605.

[66] 夏军. 准东煤田土壤重金属污染高光谱遥感监测研究. 2014，新疆大学.

[67] Zhou C, Wang D, Chen S, et al. Vegetation Corrected Continuum Depths Model and Its Application in Mineral Extraction from Hyperspectral Image. Earth Science(Journal of China University of Geosciences), 2015, 40(8): 1365-1370.

[68] Breiman L. Random forests. Machine learning. 2001, 45(1): 5-32.

[69] Huang S. A remote sensing ship recognition using random forest.n Proceedings of The fourth International Conference on Information Science and Cloud Computing, Guangzhou, China: Sissa Medialabsrl Partita: 2015: 18-19.

[70] Wang L, Zho X, Zhu X, et al. Estimation of biomass in wheat using random forest regression algorithm and remote sensing data. The Crop Journal, 2016, 4(3): 212-219.

[71] Belgiu, Dragut. Random forest in remote sensing: A review of applications and future directions. ISPRS J PHOTOGRAMM, 2016, 114: 24-31.

[72] Wang Q, Xie Z, Li F. Using ensemble models to identify and apportion heavy metal pollution sources in agricultural soils on a local scale. Environmental Pollution, 2015, 206: 227-235.

[73] Rodriguez-Galiano V, Sanchez-Castillo M, Chica-Olmo M, et al. Machine learning predictive models for mineral prospectivity: An evaluation of neural networks, random forest, regression trees and support vector machines. Ore Geology Reviews: Journal for Comprehensive Studies of Ore Genesis and Ore Exploration, 2015, 71: 804-818.

[74] Kemper T, Sommer S. Estimate of heavy metal contamination in soils after a mining accident using reflectance spectroscopy. Environmental Science & Technology, 2002, 36(12): 2742.

[75] Behrens T, Scholten T. Chapter 25 A Comparison of Data-Mining Techniques in Predictive Soil Mapping. Developments in Soil Science, 2006, 31(06): 353-364.

[76] Yi P, Rania K, Kabindra A, et al. Digital Mapping of Toxic Metals in Qatari Soils Using Remote Sensing and Ancillary Data. Remote Sensing, 2016, 8(12).

[77] Meng X, Bao Y, Liu J, et al. Regional soil organic carbon prediction model based on a discrete wavelet analysis of hyperspectral satellite data. International Journal of Applied Earth Observation and Geoinformation, 2020, 89: 102111.

[78] Ren H Y, Zhuang D F, Singh A N, et al. Estimation of As and Cu Contamination in Agricultural Soils Around a Mining Area by Reflectance Spectroscopy: A Case Study. Pedosphere, 2009, 19(6): 719-726.

[79] Wu Y, Chen J, Ji J, et al. Feasibility of reflectance spectroscopy for the assessment of soil mercury contamination. Environmental Science & Technology, 2005, 39(3): 873-878.

[80] Khan S, Cao Q, Zheng Y M, et al. Health risks of heavy metals in contaminated soils and food crops irrigated with wastewater in Beijing, China. Environmental Pollution, 2008, 152(3): 686-692.

[81] Markey A M, Clark C S, Succop P A, et al. Determination of the feasibility of using a portable X-ray fluorescence (XRF) analyzer in the field for measurement of lead content of sieved soil. Journal of Environmental Health, 2008, 70(7): 24-29.

[82] Cai L, Wang Q, Luo J, et al. Heavy metal contamination and health risk assessment for children near a large Cu-smelter in central China. Science of the Total Environment, 2018, 650: 725-733.

[83] Li S, Lei Y, Chen L, et al. Spatial distribution of heavy metal concentrations in peri-urban soils in eastern China. Environmental Science and Pollution Research, 2019, 26(9): 1615–1627.

[84] Lu Q, Wang S, Bai X, et al. Rapid inversion of heavy metal concentration in karst grain producing areas based on hyperspectral bands associated with soil components. Microchemical Journal, 2019, 148: 404-411.

[85] 叶元元. 多金属矿区土壤重金属的高光谱定量估算研究[D]. 2014，中国矿业大学.

[86] 李剑锋. 基于近红外光谱漫反射技术检测鸡蛋新鲜度的研究. 2009，华中农业大学.

[87] Chen Y Q. Savitzky-Golay Smoothing Filter. 2010.

[88] 戴彬，吕建树，战金成，等. 山东省典型工业城市土壤重金属来源、空间分布及潜在生态风险评价. 环境科学. 2015，36(02)：507-515.

[89] Zhou Y, Xue J, Chen S, et al. Exploring the potential of airborne hyperspectral image for estimating topsoil organic carbon: Effects of fractional-order derivative and optimal band combination algorithm. Geoderma, 2020, 365: 114228.

[90] Lin X, Su Y, Shang J, et al. Geographically Weighted Regression Effects on Soil Zinc Content Hyperspectral Modeling by Applying the Fractional-Order Differential. Remote Sensing, 2019, 11(6): 636.

[91] 程蓉. 基于小波的神经网络研究及其在信息处理中的应用. 2015，中山大学.

[92] Lei Y, Hong Y S, Zhou Y, et al. Inversion of Soil Organic Matter Content Using Hyperspectral Data Based on Continuous Wavelet Transformation. Guang pu xue yu guang pu fen xi, 2016, 36(5): 1428-1433.

[93] Blackburn G A, Ferwerda J G. Retrieval of chlorophyll concentration from leaf reflectance spectra using wavelet analysis. Remote Sensing of Environment, 2008, 112(4): 1614-1632.

[94] 于雷，洪永胜，周勇，等. 连续小波变换高光谱数据的土壤有机质含量反演模型构建. 光谱学与光谱分析. 2016，36(05)：1428-1433.

[95] 赵英时. 遥感应用分析原理与方法. 2003: 遥感应用分析原理与方法.

[96] F V E, D T, L D J. 6S User Guide Version. Vol. 2. 1997. 1-54.

[97] Tanre D, Deuze J L, Herman M, et al. Second Simulation of the Satellite Signal in the Solar Spectrum, 6S: an overview.Geoscience and Remote Sensing Symposium, 1990. IGARSS '90. 'Remote Sensing Science for the Nineties'., 10th Annual International,2002.

[98] 张安定. 遥感原理与应用题解. 2016, 北京: 科学出版社.

[99] CNEMC. The background values of elements in Chinese soils. 1990, Environmental Science Press of China: Beijing. 15-505.

[100] Wang L, Lin Q Z, Dong J, et al. Study on the Prediction of Soil Heavy Metal Elements Content Based on Reflectance Spectra. Journal of Remote Sensing, 2007.

[101] Kooistra L, Wehrens R, Leuven R, et al. Possibilities of visible-near-infrared spectroscopy for the assessment of soil contamination in river floodplains. Analytica Chimica Acta, 2001, 446(1): 97-105.

[102] Ji G, Xu B. REFLECTANCE OF SOIL CLAY MINERALS AND ITS APPLICATION IN PEDOLOGY. Acta Pedologica Sinica, 1987, (01): 67-76.

[103] 黄应丰，刘腾辉. 华南主要土壤类型的光谱特性与土壤分类. 土壤学报. 1995，032(001)：58-68.

[104] Ou D, Tan K, Lai J, et al. Semi-supervised DNN regression on airborne hyperspectral imagery for improved spatial soil properties prediction. Geoderma, 2021, 385: 114875.

[105] 江振蓝，杨玉盛，沙晋明. GWR模型在土壤重金属高光谱预测中的应用. 地理学报. 2017，72(3)：533-544.

[106] Jiang Q, Liu M, Wang J, et al. Feasibility of using visible and near-infrared reflectance spectroscopy to monitor heavy metal contaminants in urban lake sediment. CATENA, 2018, 162: 72-79.

[107] Dorini F A, Cecconello M S, Dorini L B. On the logistic equation subject to uncertainties in the environmental carrying capacity and initial population density. Communications in Nonlinear Science & Numerical Simulation, 2016, 33: 160-173.

[108] Zhang X, Sun W, Cen Y, et al. Predicting cadmium concentration in soils using laboratory and field reflectance spectroscopy. Science of The Total Environment, 2018, 650(1-834): 321-334.

[109] Lan Z Y, Liu Y. Research on Indirect Hyperspectral Estimating Model and the Spatial Distribution Characteristics of Heavy Metal Contents in Basin Soil of Lean River. Geography and Geo-Information Science, 2015, 31(3): 26-31+2.

[110] Hang C, Shen R, Chen Y, et al. Estimating heavy metal concentrations in suburban soils with reflectance spectroscopy. Geoderma, 2019, 336: 59-67.

[111] Hong Y, Chen S, Zhang Y, et al. Rapid identification of soil organic matter level via visible and near-infrared spectroscopy: Effects of two-dimensional correlation coefficient and extreme learning machine. Science of The Total Environment, 2018, 644: 1232-1243.

[112] Ji W J, Zhou S, Zhou Q, et al. VIS-NIR reflectance spectroscopy of the organic matter in several types of soils. JOURNAL OF INFRARED AND MILLIMETER WAVES, 2012, 31(3): 277-282.

[113] 徐彬彬，季耿善，朱永豪. 中国陆地背景和土壤光谱反射特性的地理分区的初步研究. 环境遥感. 1991(02)：142-151.

[114] Liu J, Yang Z, Wang H, et al. Study on the prediction of soil heavy metal elements content based on visible near-infrared spectroscopy. Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy, 2018, 199: 43-49.

[115] Fei W, Li C, Wang J, et al. Concentration estimation of heavy metal in soils from typical sewage irrigation area of Shandong Province, China using reflectance spectroscopy. Environmental Science & Pollution Research, 2017, 24(3): 16883-16892.

[116] Xu S, Zhao Y, Wang M, et al. Comparison of multivariate methods for estimating selected soil properties from intact soil cores of paddy fields by Vis-NIR spectroscopy. Geoderma, 2018, 310: 29-43.

[117] Nawar S, Buddenbaum H, Hill J, et al. Estimating the soil clay content and organic matter by means of different calibration methods of vis-NIR diffuse reflectance spectroscopy. Soil & Tillage Research, 2016, 155: 510-522.

[118] Douglas R K, Nawar S, Cipullo S, et al. Evaluation of vis-NIR reflectance spectroscopy sensitivity to weathering for enhanced assessment of oil contaminated soils. Science of The Total Environment, 2018, 626: 1108-1120.

[119] Zhou X, Sun J, Tian Y, et al. A deep learning-based regression method on hyperspectral data for rapid prediction of cadmium residue in lettuce leaves. Chemometrics and Intelligent Laboratory Systems, 2020, 200: 103996.

[120] Zhou X, Sun J, Tian Y, et al. Development of deep learning method for lead content prediction of lettuce leaf using hyperspectral images. International Journal of Remote Sensing, 2019, 41(6): 2263-2276.

[121] Zhou X, Jun S, Yan T, et al. Hyperspectral technique combined with deep learning algorithm for detection of compound heavy metals in lettuce. Food Chemistry, 2020, 321: 126503.