College of Geoscience and Surveying Engineering,China University of Mining and Technology-Beijing

矿区生态环境遥感可为矿区土地复垦或生态修复活动提供必要的监测数据,也可为相关部门提供便捷的监管工具,具有重要的实用价值和研究意义。然而,面向矿区或矿山场景,大多数生态环境遥感综合指数还未能有效顾及稀疏植被区裸土和裸岩的质量差异问题(简称“岩土差异”),以及浓密植被区植物多样性导致的生态质量差异问题(简称植物多样性)。为了更全面、准确地监测与评价矿区生态环境,建立了矿区植被-土壤-不透水层(或裸露岩石)框架,并据此发展出一种面向矿区环境的遥感综合生态指数(简称REM)。研究以Sentinel-2多光谱数据为驱动,以抚顺西露天矿为研究区域,实现了2016—2021年10m分辨率的年度REM监测,并结合野外调查数据(95个采样点)和地表覆盖类型分类数据对比分析了REM模型的性能。结果表明:①REM值与土地覆盖类型表征的生态环境质量具有较强的一致性,并能够区分裸土和不透水层(或裸露岩石)的质量差异;②REM值与野外采样点调查的生态环境质量也具有较强的一致性(相关系数0.8994,显著性水平P<0.01),并能够顾及植物多样性问题;③REM值有效表征了2016—2021年研究区生态环境质量的时空格局和演变,通过2021与2016年REM差值也可以有效表征生态修复区域的质量变化。最后,探讨了REM值的分级问题,给出了REM模型的简化应用策略。综上,构建的REM模型在准确描述矿区生态环境质量的同时,能够有效兼顾“岩土差异”和“植物多样性”问题。当然,REM模型中的参数有可能需要适时调整,以适应“岩土差异”和“植物多样性”问题在不同地理区域、数据条件或矿区场景下的变化规律。

Remote sensing ecological environment for mining areas can provide necessary monitoring informationfor mining land reclamation or ecological restoration. It can also provide required management tool for related manage⁃ment. Therefore,it is essential to develop a comprehensive remote sensing ecological index of mining areas. However,in terms of mining areas,most of the existing methods have not fully considered the difference between soil and sockover sparse vegetation area (abbreviated as rock⁃soil difference) and the role of plant diversity over dense vegetationarea ( abbreviated as plant diversity ). In order to monitor and evaluate the ecological environment of miningareas more comprehensively and accurately,a framework called Vegetation⁃ Impervious surface ( or bare rock)⁃Soilfor Mining area ( VIS⁃M) was constructed. Then a Remote sensing Ecological index of Mining areas consideringplant diversity and rock⁃soil difference (abbreviated as REM) was proposed based on the VIS⁃M. Taking the Fush⁃un Western Open⁃pit coal mine in Liaoning province,China as the study area,the REM was implemented at 10 m spa⁃tial resolution and one⁃year temporal resolution from 2016 to 2021 with Sentinel⁃2 data. The field survey data with 95sampling sites and land cover classification data in 2 years were also used to evaluate the REM. Results indicated:① there was well consistence between the REM value and the ecological environment quality indicated by theland cover type. Also,the REM can distinguish the difference between soil and rock. ② The REM values presentedwell consistence with the field survey data ( correlation coefficient is 0. 899 4 at a significant level P < 0. 01).Moreover,the REM can demonstrate the role of plant diversity over dense vegetation area. ③ The REM illustrated thespatiotemporal pattern of ecological environment for the study area from 2016 to 2021. Furthermore,through the differ⁃ence between 2021 REM and 2016 REM,the ecological restoration area was clearly mapped. Finally,the grading crite⁃ria of REM was discussed and some convenient ways of determining weights and plant diversity were suggested. Insummary,the REM model can accurately describe the quality of the ecological environment in the mining area,andat the same time, it can effectively take into account the issues of rock⁃soil difference and plant diversity. Ofcourse,the parameters in the REM model may need to be adjusted in time to adapt to the changing laws of rock⁃soil difference and plant diversity in different geographical areas,data conditions or mining scenarios.

北京市自然科学基金面上资助项目(6222045);国家自然科学基金面上资助项目(41871338);中国矿业大学(北京)“越崎青年学者”资助项目(CUMTB2018)

孙灏,胡佳琪,蒋金豹,等. 顾及“岩土差异”和植物多样性的矿区生态环境遥感监测模型[J].煤炭学报,2023,48(S1):219-232.

SUN Hao,HU Jiaqi,JIANG Jinbao,et al. REM:A remote sensing ecological index of mining areas consideringplant diversity and rock⁃soil difference[J]. Journal of China Coal Society,2023,48(S1):219-232