高潜水位煤矿区开采扰动的长时序过程遥感监测与影响评价  

Remote sensing monitoring and impact assessment of mining disturbance in mining area with high undergroundwater level

作　　者：肖武 陈文琦[1] 何厅厅 赵艳玲 胡振琪[2] XIAO Wu;CHEN Wenqi;HE Tingting;ZHAO Yanling;HU Zhenqi(School of Public Affairs,Zhejiang University,Hangzhou310058,China;Institute of Land Reclamation and Ecological Restoration,China University of Mining and Technology-Beijing,Beijing100083,China)

机构地区：[1]浙江大学公共管理学院,浙江杭州310058 [2]中国矿业大学(北京)土地复垦与生态重建研究所,北京100083

出　　处：《煤炭学报》2022年第2期922-933,共12页Journal of China Coal Society

基　　金：国家自然科学基金资助项目(42071250);中国博士后科学基金资助项目(2021M702795)。

摘　　要：煤炭开采导致地表沉陷积水,是东部高潜水位平原煤矿区的主要特征,对矿区开采中沉陷水体变化的长时序检测,有助于定量评估煤炭开采对土地、生态与社会的综合影响效应。在没有地下采矿信息为先导的情况下,如何识别与区分自然水体、人类地面活动导致的挖掘水体,以及沉陷水体,同时量化开采沉陷影响的边界与程度是目前单纯采用遥感手段进行监测的难题。以山东兖州煤田为例,利用Google Earth Engine(GEE)云计算平台,基于1986年以来可以获取的所有Landsat时序影像数据,结合CCDC(Continue Change Detection and Classification)算法开发了基于轨迹数据检测变化的沉陷积水年份和复垦年份时空动态制图的方法,在此基础上,以提取的沉陷水体图斑为依据,利用Sentinel-2数据分别反演了积水缓冲区反映土壤含水量的VSDI(Visible and Shortwave Infrared Drought Index),LSWI(Land Surface Water Index),SMMI (Soil moisture monitoring index)3个指数,根据土壤水分空间变化规律,通过测度离沉陷积水区不同距离土壤水分空间分布特征,进一步对开采扰动的影响范围与程度进行定量分析。结果显示:基于CCDC算法识别1986—2017年地下煤炭开采引起的地表沉陷积水与复垦年度时空数据,精度分别为85%,77%。研究区自1990年出现沉陷水体,至2017年累计沉陷积水面积3 021.08 ha,其中75.80%的沉陷积水发生在2001—2011年;沉陷积水复垦从1993年开始出现,累计面积888.37 ha,占沉陷积水总面积的29.41%,主要集中于2007年之后。沉陷水体的影响主要集中在沉陷水体外围120 m范围内,该区域出现剧烈的土壤水分变化;120~300 m内存在扰动但是影响强度轻微;300 m之外几乎无影响。通过分析矿区长时序的沉陷积水变化过程,并基于高潜水位矿区土壤水分的遥感反演识别开采扰动的影响范围与程度,为类似矿区开采沉陷水体的监测识别与影响评估提供了�Coal mining leads to surface subsidence and waterlogging, which is the main feature of coal mining areas in eastern plains with high underground water levels. Long-term detection of changes in the subsidence water body is helpful to quantitatively evaluate the comprehensive impact of coal mining on land, ecology, and community. In the absence of underground mining knowledge as a guide, how to identify and distinguish the mining subsidence water body from the natural water body and the excavated water body caused by other human activities, and quantify the boundary and extent of the mining subsidence effects are difficult to be monitored by remote sensing only. This study takes the Yanzhou coalfield in Shandong Province as the study area. Based on all time-series Landsat images available since 1986,the authors have developed a spatiotemporal dynamic mapping method to detect the changes of subsidence water and land reclamation based on trajectory data under Google Earth Engine(GEE) platform using the CCDC(Continue Change Detection and Classification) algorithm. Three indices of VSDI(Visible and Shortwave Infrared Drought Index),LSWI(Land Surface Water Index),and SMMI(Soil moisture monitoring index) that reflect the soil moisture was inverted in the buffer zone of the subsidence water body using Sentinel-2 data respectively. According to the spatial variation law of soil moisture content, the authors measured the spatial distribution characteristics of soil moisture at different distances from the subsidence water area, further quantitatively analyzed the influence range and degree of mining disturbance. The results show that(1) The annual spatial and temporal data of surface subsidence water and reclamation caused by underground coal mining from 1986 to 2017 were identified based on the CCDC algorithm with an accuracy of 85% and 77%,respectively.(2) From 1990 to 2017,a waterlogged area of 3 021.08 hectares was accumulated in the study area, about 75.80% of which occurred from 2001 to 2011. Subsidence water reclamat

关 键 词：高潜水位煤矿区 开采沉陷积水 Google Earth Engine 云计算 变化监测 

分 类 号：TD88[矿业工程—矿山开采]