基于多光谱遥感的矿区积水区水深反演  

Water Depth Inversion Based on Multi-spectral Remote Sensing in Mine Subsidence Area

作　　者：吴双 武瑞杰 吴侃[2] 王瑞[2,4] 周大伟[2] Wu Shuang;Wu Ruijie;Wu kan;Wang Rui;Zhou Dawei(Key Laboratory of Resources and Environmental Information Engineering of Jiangsu Province,Xuzhou 221116,China;School of Environment and Spatial Informatics,China University of Mining and Technology,Xuzhou 221116,China;School of Resources and Environmental Science,Wuhan University,Wuhan 430079,China;School of Applied Science,Jiangxi University of Science and Technology,Ganzhou 341000,China)

机构地区：[1]江苏省资源环境信息工程重点实验室,江苏徐州221116 [2]中国矿业大学环境与测绘学院,江苏徐州221116 [3]武汉大学资源与环境科学学院,湖北武汉430079 [4]江西理工大学应用科学学院,江西赣州341000

出　　处：《金属矿山》2020年第8期129-135,共7页Metal Mine

基　　金：国家自然科学基金项目(编号:51604266)。

摘　　要：采煤沉陷区积水会严重影响到矿区周围的生态环境和生产工作,为定量获取淮南潘一矿沉陷区的积水深度,基于多光谱遥感数据对沉陷区积水深度的反演方法进行了研究。首先选择两期Landsat 5卫星遥感影像进行辐射定标和积水区水体提取等预处理;然后基于像元绝对辐射亮度值与2005年研究区水深实测值,分别选取单波段模型、双波段比值模型进行建模分析,在此基础上,选择拟合优度最高的影响因子建立多波段模型并进行模型改进;最后将2010年反演结果与矿区沉陷预测预报系统(MSPS)软件计算的水深值进行对比分析。结果表明:改进后的水深反演模型拟合优度达到0.9以上,且2005年潘一矿沉陷区积水深度反演值与实测值相比平均绝对误差为0.21 m,平均相对误差为10.92%;2010年该区域水深反演值与预计值的平均绝对误差为0.63 m,平均相对误差为15.24%,二者差值呈正态分布,可以达到相互验证的目的。研究表明:利用多光谱遥感影像可以得到精度较高(误差在分米级)的水深反演模型,在一定程度上满足工程需求,可以为矿区生态环境治理和评价提供可靠依据。Water accumulation in coal mining subsidence area will affect the ecological environment and production work seriously around the mining area.In order to obtain the water accumulation depth quantitatively in the subsidence area of Panyi Mine in Huainan,the inversion method of water accumulation depth in subsidence area was studied based on multispectral remote sensing data.Firstly,two phases of Landsat 5 remote sensing images were selected for radiation calibration,water extraction and other pretreatment operations.Then,based on the absolute radiation luminance value of pixel and the measured depth of water in the study area in 2005,the single-band model and dual-band ratio model were selected for modeling and analysis respectively,besides that,the influencing factor with the highest fitting goodness was selected to establish the multi-band model and improve it.Finally,the inversion results in 2010 are compared with the water depth values calculated by MSPS software.The results show that the improved water depth inversion model has a good fitting degree of more than 0.9,and the average absolute error of the inversion value and measured value of water depth in the subsidence area of Panyi Mine in 2005 is 0.21 m,and the average relative error is 10.92%.In 2010,the average absolute error of annual water depth inversion value in this region is 0.63 m,and the average relative error is 15.24%compared with the predicted value.The difference between the two values is normally distributed,which can achieve the purpose of mutual verification.The study results show that the multi-spectral remote sensing image can be used to obtain the water depth inversion model with high accuracy(the error in the decimeter level),which can basically meet the engineering requirements,can provide reliable reference for ecological environment management and evaluation of mining area.

关 键 词：多光谱遥感 水深反演 多波段非线性模型 沉陷积水区 矿区沉陷预测预报系统 

分 类 号：TD325[矿业工程—矿井建设] TP79[自动化与计算机技术—检测技术与自动化装置]