在国家资源战略西移的背景下，新疆将成为我国十分重要的矿产资源接替区和战略储备区。矿区土壤碳排放已成为陆地生态系统碳循环的重要环节，间接影响了全球温室气体排放进程，以往研究主要依靠土壤原位监测且区域集中于华北及云贵川地区，针对新疆露天矿区的土壤碳排放监测研究较少，因此基于遥感技术探究新疆红沙泉露天煤矿土壤碳排放的高光谱响应特征，采集研究区103组土壤样品，获取土壤碳排放速率、土壤反射率、土壤温度（Soil Temperature，ST）、土壤湿度（Soil Moisture，SM）数据；采用Savitzky-Golay（SG）滤波平滑，数学变换、相关系数-连续投影算法（CC-SPA）以及竞争自适应重加权采样算法（CARS）进行数据变换与特征波段筛选，筛选土壤碳排放速率敏感波段；结合研究区裸岩石砾地指数NGLI、归一化差异植被指数NDVI等，选择偏最小二乘（PLSR）、支持向量机（SVM）、随机森林（RF）与遗传优化神经网络（GA-BP）建立研究区不同土地利用类型下土壤碳排放最优反演模型。结果表明： 

（1）自然条件下土壤的反射率明显高于采矿活动影响下的土壤，平均反射率在0.35左右，受矿区开采影响最大的南线地区土壤反射率最低，平均反射率在0.15左右，曲线整体在400-760nm内急剧上升，在760-2000nm范围内上升速率减慢，在2000-2400nm范围内具有明显特征峰。

（2）光谱特征筛选方面，基于CC-SPA法提取的碳排放特征波段数远小于相关系数法（CC）与连续投影算法，且特征波段呈现一定的聚集式分布，主要集中于1600~2200nm波长范围内。

（3）添加光谱指数及ST、SM的反演模型对土壤碳排放速率的估测精度明显提升，其中矿区综合土地利用类型下土壤的一阶微分变换（1st）的SVM反演效果最好，测试集R²=0.851、RMSE=0.095gm^-2hr^-1，5种不同特征土地利用类型下土壤的最佳指数组合方式存在差异，不同的光谱指数对各土壤碳排放速率的估测精度均有不同程度的提升效果，其中测试集R²均在0.8以上，可以较为准确地估测红沙泉矿区内不同特征土壤类型的碳排放速率。

（4）昼-夜环境下不同土地利用类型下土壤碳排放敏感波段分布具有差异性，白天敏感波段少量分布于可见光及红光波段（404~773nm），主要分布于近红外波段范围（822~1348nm）；夜间敏感波段集中于近红外及中红外（943~1346nm、1669~2320nm）波段范围，相比于白天，夜晚敏感波段具有明显向长波方向“后移”特征，排土场、人工林、南线、柽柳碳排放速率与特征波段间相关系数r绝对值大于0.8。

该研究可为荒漠化矿区土壤碳排放遥感反演提供依据，实现矿区碳排放的无损探测，为我国双碳目标实现提供数据支撑。

In the context of the westward shift of the national resource strategy, Xinjiang will become a very important mineral resources replacement area and strategic reserve area in my country. Soil carbon emissions in mining areas have become an important link in the carbon cycle of terrestrial ecosystems, indirectly affecting the process of global greenhouse gas emissions. Previous studies mainly relied on soil in-situ monitoring and concentrated in North China and Yunnan-Guizhou-Sichuan regions. Soil carbon emissions in open-pit mining areas in Xinjiang There are few monitoring studies, so based on remote sensing technology to explore the hyperspectral response characteristics of soil carbon emissions in Hongshaquan open-pit coal mine in Xinjiang, 103 groups of soil samples were collected in the study area to obtain soil carbon emission rate, soil reflectance, soil temperature (Soil Temperature, ST ), soil moisture (Soil Moisture, SM) data; using Savitzky-Golay (SG) filter smoothing, mathematical transformation, correlation coefficient-continuous projection algorithm (CC-SPA) and competitive adaptive reweighted sampling algorithm (CARS) for data transformation Screening with characteristic bands, screening the sensitive bands of soil carbon emission rate; combined with the bare rock gravel index NGLI, normalized difference vegetation index NDVI, etc., select partial least squares (PLSR), support vector machine (SVM), random forest (RF) and genetic optimization neural network (GA-BP) to establish the optimal inversion model of soil carbon emissions under different land use types in the study area. The results show:

(1) The reflectance of the soil under natural conditions is significantly higher than that of the soil under the influence of mining activities, with an average reflectance of about 0.35. The soil reflectance of the southern line, which is most affected by mining, is the lowest, with an average reflectance of about 0.15. The overall curve It rises sharply in the range of 400-760nm, the rising rate slows down in the range of 760-2000nm, and has obvious characteristic peaks in the range of 2000-2400nm.

(2) In terms of spectral feature screening, the number of carbon emission characteristic bands extracted based on the CC-SPA method is much smaller than that of the correlation coefficient method (CC) and the continuous projection algorithm, and the characteristic bands present a certain clustered distribution, mainly concentrated in 1600~2200nm wavelengths within range.

(3) The estimation accuracy of the soil carbon emission rate is significantly improved by adding the spectral index and the inversion model of ST and SM, and the SVM inversion effect of the first-order differential transformation (1st) of the soil under the comprehensive land use type of the mining area is the best. The test set R²=0.851, RMSE=0.095gm^-2hr^-1, there are differences in the optimal index combinations of soils under the five different characteristic land use types, and different spectral indices have different degrees of improvement in the estimation accuracy of soil carbon emission rates. Among them, the R² of the test set is above 0.8, which can accurately estimate the carbon emission rate of different characteristic soil types in the Hongshaquan mining area.

(4) The distribution of sensitive bands of soil carbon emission under different land use types in the day-night environment is different. During the day, a small amount of sensitive bands are distributed in the visible light and red light bands (404~773nm), and mainly distributed in the near-infrared band range (822~773nm). 1348nm); the sensitive bands at night are concentrated in the near-infrared and mid-infrared (943~1346nm, 1669~2320nm) bands. The absolute value of the correlation coefficient r between the carbon emission rate of, Nanxian and Tamarix and the characteristic bands is greater than 0.8.

This research can provide a basis for remote sensing inversion of soil carbon emissions in desertified mining areas, realize non-destructive detection of carbon emissions in mining areas, and provide data support for the realization of my country's dual carbon goals.

[1]Schimel D S. Terrestrial ecosystems and the carbon cycle[J]. Global change biology, 1995, 1(1): 77-91.

[2]Bond-Lamberty B, Thomson A. Temperature-associated increases in the global soil respiration record[J]. Nature, 2010, 464(7288): 579-582.

[3]Schlesinger W H, Andrews J A. Soil respiration and the global carbon cycle[J]. Biogeochemistry, 2000, 48(1): 7-20.

[4]Raich J W, Potter C S, Bhagawati D. Interannual variability in global soil respiration, 1980–94[J]. Global Change Biology, 2002, 8(8): 800-812.

[5]Zhou Z, Zhang Z, Zha T, et al. Predicting soil respiration using carbon stock in roots, litter and soil organic matter in forests of Loess Plateau in China[J]. Soil Biology and Biochemistry, 2013, 57: 135-143.

[6]张萌,卢杰,任毅华.土壤呼吸影响因素及测定方法的研究进展[J].山东林业科技,2021,51(02):100-106+92.

[7]赵宁伟, 郜春花, 李建华.土壤呼吸研究进展及其测定方法概述[J].山西农业科学,2011,39(1):4.

[8]陈宝玉, 王洪君, 杨建,等.土壤呼吸组分区分及其测定方法[J].东北林业大学学报,2009,37(1):96-99.

[9]郑永红, 张治国, 胡友彪,等.煤矿复垦重构土壤呼吸季节变化特征及其环境影响因子[J].煤炭学报,2014,39(11):2300-2306.

[10]郭春燕, 李晋川, 岳建英,等.安太堡露天煤矿复垦区不同人工林土壤呼吸特征[J].水土保持通报,2015,35(03):55-61.

[11]周育智, 王芳, 陈孝杨,等.生态修复模式对淮南矿区重构土壤CO2通量日变化的影响[J].水土保持通报,2016,36(6):8.

[12]张玉龙, 蔡旭, 张凤华.干旱区盐渍化弃耕地复垦后农田土壤呼吸及碳平衡[J].江苏农业科学,2019,47(9):284-288.

[13]渠俊峰, 葛梦玉, 王坤,等.邹城东滩矿区复垦土壤微生物量碳与基础呼吸的特征[J].长江流域资源与环境,2018,27(08):1858-1865.

[14]Kowalska A, Singh B R, Grobelak A. Carbon footprint for post-mining soils: The dynamic of net CO2 fluxes and SOC sequestration at different soil remediation stages under reforestation[J]. Energies, 2022, 15(24):9452.

[15]Ahirwal J, Maiti S K. Assessment of soil carbon pool, carbon sequestration and soil CO 2 flux in unreclaimed and reclaimed coal mine spoils[J]. Environmental earth sciences, 2018, 77: 1-13.

[16]Ngaba M J Y, Uwiragiye Y, Bol R, et al. Global cross-biome patterns of soil respiration responses to individual and interactive effects of nitrogen addition, altered precipitation, and warming[J]. Science of the Total Environment, 2023, 858: 159808.

[17]De Long J R, Heinen R, Heinze J, et al. Plant-soil feedback: incorporating untested influential drivers and reconciling terminology[J]. Plant and Soil, 2023: 1-37.

[18]Huete A, Didan K, Miura T, et al. Overview of the radiometric and biophysical performance of the MODIS vegetation indices[J]. Remote sensing of environment, 2002, 83(1-2): 195-213.

[19]Justice C, Belward A, Morisette J, et al. Developments in the ‘validation’ of satellite sensor products for the study of the land surface. International Remote Sensing, 2000, 21(17):3383–3390.

[20]Asgari N, Ayoubi S, Jafari A, et al. Incorporating envi-ronmental variables, remote and proximal sensing data for digital soil mapping of USDA soil great groups. Int J Remote Sensing,2020, 41(19):7624–7648.

[21]Gitelson A A, Viña A, Ciganda V, et al. Remote estimation of canopy chlorophyll content in crops[J]. Geophysical research letters, 2005, 32(8).

[22]Suárez L, Zarco-Tejada P J, González-Dugo V, et al. Detecting water stress effects on fruit quality in orchards with time-series PRI airborne imagery[J]. Remote Sensing of Environment, 2010, 114(2): 286-298.

[23]Wylie B K, Johnson D A, Laca E, et al. Calibration of remotely sensed, coarse resolution NDVI to CO2 fluxes in a sagebrush–steppe ecosystem[J]. Remote Sensing of Environment, 2003, 85(2): 243-255.

[24]Cicuendez V, Rodriguez-Rastrero M, Huesca M, et al. Assessment of soil respiration patterns in an irrigated corn field based on spectral information acquired by field spectroscopy[J]. Agriculture, ecosystems & environment, 2015, 212: 158-167.

[25]Huang N, Niu Z, Zhan Y, et al. Relationships between soil respiration and photosynthesis-related spectral vegetation indices in two cropland ecosystems[J]. Agricultural and Forest Meteorology, 2012, 160: 80-89.

[26]Huang N, Niu Z. Estimating soil respiration using spectral vegetation indices and abiotic factors in irrigated and rainfed agroecosystems[J]. Plant and Soil, 2013, 367: 535-550.

[27]Stocker T F, Qin D, Plattner G K, et al. Contribution of working group I to the fifth assessment report of the intergovernmental panel on climate change[J]. Climate change, 2013, 5: 1-1552.

[28]Zhou Z, Zhang Z, Zha T, et al. Predicting soil respiration using carbon stock in roots, litter and soil organic matter in forests of Loess Plateau in China[J]. Soil Biology and Biochemistry, 2013, 57: 135-143.

[29]姚雪雯, 陈书涛, 王君,等. 不同作物农田的土壤呼吸与高光谱的关系[J]. 农业环境科学学报, 2020, 39(5):10.

[30]崔慧瑾, 刘菊, 杜自强,等. 生物炭和地膜覆盖下高粱高光谱特征及其与土壤呼吸的关系[J]. 水土保持研究, 2020, 27(3):8.

[31]Liang Y, Cai Y, Yan J, et al. Estimation of soil respiration by its driving factors based on multi-source data in a sub-alpine meadow in north China[J]. Sustainability, 2019, 11(12): 3274.

[32]Cicuendez V, Rodriguez-Rastrero M, Huesca M, et al. Assessment of soil respiration patterns in an irrigated corn field based on spectral information acquired by field spectroscopy[J]. Agriculture, ecosystems & environment, 2015, 212: 158-167.

[33]Huang N, Niu Z, Zhan Y, et al. Relationships between soil respiration and photosynthesis-related spectral vegetation indices in two cropland ecosystems[J]. Agricultural and Forest Meteorology, 2012, 160: 80-89.

[34]Ding S, Yao X, Wang J, et al. Relationships between soil respiration and hyperspectral vegetation indexes and crop characteristics under different warming and straw application modes[J]. Environmental Science and Pollution Research, 2021, 28(30): 40756-40770.

[35]Huang N, Gu L, Niu Z. Estimating soil respiration using spatial data products: A case study in a deciduous broadleaf forest in the Midwest USA[J]. Journal of Geophysical Research: Atmospheres, 2014, 119(11): 6393-6408.

[36]Wu C, Gaumont-Guay D, Black T A, et al. Soil respiration mapped by exclusively use of MODIS data for forest landscapes of Saskatchewan, Canada[J]. ISPRS journal of photogrammetry and remote sensing, 2014, 94: 80-90.

[37]Huang N, Niu Z, Zhan Y, et al. Relationships between soil respiration and photosynthesis-related spectral vegetation indices in two cropland ecosystems[J]. Agricultural and Forest Meteorology, 2012, 160: 80-89.

[38]Johnson J A, Runge C F, Senauer B, et al. Global food demand and carbon-preserving cropland expansion under varying levels of intensification[J]. Land Economics, 2016, 92(4): 579-592.

[39]王琼杰. “双碳”目标下难改煤炭主体能源地位[N]. 中国矿业报, 2021-07-19(001).

[40]Caspersen J P, Pacala S W, Jenkins J C, et al. Contributions of land-use history to carbon accumulation in US forests[J]. Science, 2000, 290(5494): 1148-1151.

[41]Schimel D S. Terrestrial ecosystems and the carbon cycle[J]. Global change biology, 1995, 1(1): 77-91.

[42]任志胜, 齐瑞鹏, 王彤彤,等. 风化煤对晋陕蒙矿区排土场新构土体土壤呼吸的影响[J]. 农业工程学报, 2015, 31(23): 230-237.

[43]陈孝杨, 王芳, 严家平,等. 覆土厚度对矿区复垦土壤呼吸昼夜变化的影响[J]. 中国矿业大学报, 2016, 45(01): 163-169.

[44]李君剑, 曹杰, 严俊霞,等. 矿区不同复垦措施下土壤呼吸与环境因子关系的研究[J].环境科学学报, 2014, 34(08): 2102-2110.

[45]Pandey B, Mukherjee A, Agrawal M, et al. Assessment of seasonal and site-specific variations in soil physical, chemical and biological properties around opencast coal mines[J]. Pedosphere, 2019, 29(5): 642-655.

[46]Mukhopadhyay S, Maiti S K. Soil CO2 flux in grassland, afforested land and reclaimed coalmine overburden dumps: a case study[J]. Land Degradation & Development, 2014, 25(3): 216-227.

[47]Ahirwal J, Maiti S K, Singh A K. Changes in ecosystem carbon pool and soil CO2 flux following post-mine reclamation in dry tropical environment, India[J]. Science of the Total Environment, 2017, 583: 153-162.

[48]Valente F D A, Gomes L C, Castro M F, et al. Influence of different tree species on autotrophic and heterotrophic soil respiration in a mined area under reclamation[J]. Land Degradation & Development, 2021, 32(15): 4288-4299.

[49]刘英, 魏嘉莉, 毕银丽,等. 红沙泉露天煤矿碳储量时空动态变化分析[J]. 煤炭学报, 2022, 47(S1): 214-224.

[50]张俊华, 马天成, 贾科利. 典型龟裂碱土土壤光谱特征影响因素研究[J]. 农业工程学报, 2014, 30(23): 8.

[51]陈颂超, 冯来磊, 李硕,等. 基于局部加权回归的土壤全氮含量可见-近红外光谱反演[J]. 土壤学报, 2015, 52(2): 312-320.

[52]王动民, 纪俊敏, 高洪智. 多元散射校正预处理波段对近红外光谱定标模型的影响[J]. 光谱学与光谱分析, 2014, 34(09): 2387-2390.

[53]戴昭华. 统计数学在环境科学中的应用[J]. 环境科学丛刊, 1987(01): 1-44.

[54]董春旺, 梁高震, 安霆,等. 红茶感官品质及成分近红外光谱快速检测模型建立[J]. 农业工程学报, 2018, 34(24):8.

[55]牛芳鹏, 李新国, 麦麦提吐尔逊·艾则孜,等. 基于连续投影算法的博斯腾湖西岸湖滨绿洲土壤有机碳含量的高光谱估算[J]. 浙江大学学报(农业与生命科学版), 2021, 47(05): 673-682.

[56]Jackson R D, Slater P N, Pinter P J. Discrimination of growth and water stress in wheat by various vegetation indices through clear and turbid atmospheres[J]. Remote Sensing of Environment, 1983, 13(3): 187-208.

[57]Rouse Jr J W, Haas R H, Schell J A, et al. Monitoring vegetation systems in the Great Plains with ERTS[J].NASA special publication, 1974, 1(1): 351-309.

[58]王树果, 李新, 韩旭军,等. 利用多时相ASAR数据反演黑河流域中游地表土壤水分[J]. 遥感技术与应用, 2009(5):7.

[59]Allbed A, Kumar L, Aldakheel Y Y. Assessing soil salinity using soil salinity and vegetation indices derived from IKONOS high-spatial resolution imageries: Applications in a date palm dominated region[J]. Geoderma, 2014, 230: 1-8.

[60]Price J C. Thermal inertia mapping: A new view of the Earth[J]. Journal of Geophysical Research, 1977. 82(18): 2582-2590.

[61]韩万强, 靳瑰丽, 岳永寰,等. 基于高光谱成像技术的伊犁绢蒿荒漠草地主要植物识别参数的筛选[J]. 草地学报, 2020, 28(04): 1153-1163.

[62]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[J]. Science of the Total Environment, 2018, 644: 1232-1243.

[63]Meng X, Bao Y, Ye Q, et al. Soil organic matter prediction model with satellite hyperspectral image based on optimized denoising method[J]. Remote Sensing, 2021, 13(12): 2273.

[64]张超, 蔡焕杰, 李志军. 高光谱特征参量的冬小麦吸收性光合有效辐射分量估算模型[J]. 光谱学与光谱分析, 2015, 35(9):6.

[65]罗批, 郭继昌, 李锵,等. 基于偏最小二乘回归建模的探讨[J]. 天津大学学报:自然科学与工程技术版, 2002, 35(006): 783-786.

[66]郭澎涛, 李茂芬, 罗微,等. 基于多源环境变量和随机森林的橡胶园土壤全氮含量预测[J]. 农业工程学报, 2015, 31(5): 9.

[67]都月, 孟晓辰, 祝连庆. 遗传算法和神经网络的重叠光谱解析[J]. 光谱学与光谱分析, 2020, 40(7): 7.

[68]孙亚楠,李仙岳,史海滨,等. 河套灌区土壤水溶性盐基离子高光谱综合反演模型[J]. 农业机械学报, 2019, 50(5): 344-355.

[69]Chen Y W, Lin C J. Combining SVMs with various feature selection strategies[J]. Feature extraction: foundations and applications, 2006: 315-324.

[70]Keerthi S S, Lin C J. Asymptotic behaviors of support vector machines with Gaussian kernel[J]. Neural computation, 2003, 15(7): 1667-1689.

[71]Willmott C J, Matsuura K. Advantages of the mean absolute error (MAE) over the root mean square error (RMSE) in assessing average model performance[J]. Climate research, 2005, 30(1): 79-82.

[72]Zornoza R, Guerrero C, Mataix-Solera J, et al. Near infrared spectroscopy for determination of various physical, chemical and biochemical properties in Mediterranean soils[J]. Soil Biology and Biochemistry, 2008, 40(7): 1923-1930.

[73]Rossel R A V, Behrens T. Using data mining to model and interpret soil diffuse reflectance spectra[J]. Geoderma, 2010, 158(1-2): 46-54.

[74]李俊超, 党廷辉, 郭胜利, 等. 植被重建下煤矿排土场土壤熟化过程中碳储量变化[J]. 环境科学, 2014 (10): 3842-3850.

[75]Mukhopadhyay S, Maiti S K. Soil CO2 flux in grassland, afforested land and reclaimed coalmine overburden dumps: a case study[J]. Land Degradation & Development, 2014, 25(3): 216-227.

[76]Gholizadeh A, Borůvka L, Saberioon M M, et al. Comparing different data preprocessing methods for monitoring soil heavy metals based on soil spectral features[J]. Soil and Water Research, 2015, 10(4): 218-227.

[77]Zhang Z, Ding J, Zhu C, et al. Combination of efficient signal pre-processing and optimal band combination algorithm to predict soil organic matter through visible and near-infrared spectra[J]. Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy, 2020, 240: 118553.

[78]Yang F, White M A, Michaelis A R, et al. Prediction of continental-scale evapotranspiration by combining MODIS and AmeriFlux data through support vector machine[J]. IEEE Transactions on Geoscience and Remote Sensing, 2006, 44(11): 3452-3461.

[79]Xie X, Liu W T, Tang B. Spacebased estimation of moisture transport in marine atmosphere using support vector regression[J]. Remote Sensing of Environment, 2008, 112(4): 1846-1855.

[80]成倬治, 王文君, 闫卓璇,等. 优化的神经网络算法设计及其在地表发射率反演中的应用[J]. 科技创新与应用, 2020(17): 3.

[81]Jägermeyr J, Gerten D, Lucht W, et al. A high‐resolution approach to estimating ecosystem respiration at continental scales using operational satellite data[J]. Global change biology, 2014, 20(4): 1191-1210.

[82]F Fan L C, Yang M Z, Han W Y. Soil respiration under different land uses in eastern China[J]. Public Library of Science, 2015, 10(4): e0124198.

[83]Wu C, Gaumont-Guay D, Black T A, et al. Soil respiration mapped by exclusively use of MODIS data for forest landscapes of Saskatchewan, Canada[J]. ISPRS journal of photogrammetry and remote sensing, 2014, 94: 80-90.

[84]张玉龙, 蔡旭, 张凤华.干旱区盐渍化弃耕地复垦后农田土壤呼吸及碳平衡[J].江苏农业科学,2019,47(09):284-288.

[85]Raychaudhuri B. Spatial distribution of atmospheric CO2 absorption detected around 2 µm on the reflectance spectra derived from Hyperion images[J]. International Journal of Remote Sensing, 2017, 38(11): 3415-3429.