新疆戈壁煤矿露天开采对生态环境扰动定量分析  

Quantitative analysis of coal mining disturbance on environmentin Xinjiang Gobi Open-pit mining area

作　　者：刘英[1,3] 许萍萍 毕银丽 岳辉[1,3] 彭苏萍[4] 韩勇[5] 姜凯升 LIU Ying;XU Pingping;BI Yinli;YUE Hui;PENG Suping;HAN Yong;JIANG Kaisheng(College of Geomatics,Xi’an University of Science and Technology,Xi’an 710054,China;College of Geology and Environment,Xi’an University of Science and Technology,Xi’an 710054,China;Institute of Ecological Environment Restoration in Mine Areas of West China,Xi’an University of Science and Technology,Xi’an 710054,China;State Key Laboratory of Coal Resource and Safety Mining,China University of Mining Technology-Beijing,Beijing 100083,China;Guoneng Xinjiang Hongshaquan Energy Co.,Ltd.,Urumqi 838100,China)

机构地区：[1]西安科技大学测绘科学与技术学院,陕西西安710054 [2]西安科技大学地质与环境学院,陕西西安710054 [3]西安科技大学西部矿山生态环境修复研究院,陕西西安710054 [4]中国矿业大学(北京)煤炭资源与安全开采国家重点实验室,北京100083 [5]国能新疆红沙泉能源有限责任公司,新疆乌鲁木齐838100

出　　处：《煤炭学报》2023年第2期959-974,共16页Journal of China Coal Society

基　　金：国家重点研发计划资助项目(2022YFF1303300);陕西省自然科学基础研究计划资助项目(2023-JCYB-440,2023-JC-YB-266)。

摘　　要：新疆戈壁矿区生态环境脆弱,极容易受到自然和人为因素的影响,其生态环境也逐渐被广泛关注。选取新疆五彩湾露天矿区为研究对象,利用1990—2020年Landsat影像,在遥感生态指数IRSE基础上,引入煤尘指数、盐度指标、土地退化指数构建干旱煤矿生态指数(IACME),并运用M-K检验与Sen趋势分析法揭示研究区IACME时空演变特征与规律,评价干旱矿区的生态环境质量状况,同时利用多元回归、神经网络、随机森林和支持向量机方法分别建立五彩湾矿区气候因素与IACME的关系模型,预测得到基于气候条件下的I′_(ACME),结合残差分析与双重差模型定量评估采矿活动的影响。结果表明:(1)近30 a来研究区IACME均值为0.28,总体上IACME呈下降趋势,由1990年的0.30下降到2020年的0.22。变化幅度方面,退化区面积占比为62.63%,远高于改善区28.3%,由于采矿活动的影响,显著退化面积占比增加12.96%,主要集中在以矿井为中心向外辐射区域,显著改善面积占比下降17.28%。(2)利用多元回归、神经网络、随机森林、支持向量机4类方法分别建立气候因素与IACME的关系模型,得到基于麻雀搜索算法的随机森林(SSA-RF)精度最高,验证集R^(2)=0.82、均方根误差E_(RMS)=0.109,该模型在所使用的建模方法中可靠性更高,适用性较强;(3)通过SSA-RF预测基于气候条件的I′_(ACME),气候影响下,研究区生态环境以较差(0.2~0.4)和中等(0.4~0.6)为主。假设不存在采矿活动,研究区2006—2020年I′_(ACME)在0.26~0.59,I′_(ACME)平均值为0.40,与同年份实际的IACME空间分布相比,环境显著改善的区域明显增多,主要集中在研究区中部矿井所在区域。(4)利用残差定量分析得知研究区采矿活动对生态环境具有负向效应,平均残差δ为-0.116,通过双重差分模型得出采矿对生态环境的定量影响为-0.112,进一步表明煤炭开采对生态环境存在负面扰动,导致生态环境质�[JP2]The ecological environment in Xinjiang Gobi open-pit mining area is fragile and easily affected by natural and man-made factors.Therefore,the ecological environment in Xinjiang mining area has been paid more attention.Taking Wucaiwan open-pit mining area in Xinjiang Uygur Autonomous Region as study area,coal dust index,salinity index and land degradation index were introduced to construct arid coal mine ecological index(I ACME)based on the remote sensing ecological index(I RSE)and Landsat images from 1990 to 2020.M-K test and Sen trend analysis were used to reveal the spatial and temporal evolution of I ACME in the study area,which evaluated the eco-environmental quality of arid mining area.Multiple regression,neural network,random forest and support vector machine methods were employed to establish the relationship between climate factors and I ACME since mining and the prediction of I′ACME based on only climatic conditions was obtained.The influence of mining activity was quantitatively assessed by the residual analysis and difference-in-difference method.The results showed that:①The average I ACME of the study area was 0.28 during the past 30 years,indicating the quality of ecological environment was poor.The I ACME of study area showed a decreasing trend from 1990(0.30)to 2020(0.22).The percentage of degraded area was 62.63%,which was much higher than that of improved area(28.3%).Due to the influence of mining activities,the percentage of significantly degraded area increased by 12.96%,mainly located in the outward radiation area around the mine.The proportion of significantly improved area decreased by 17.28%.②The relationship model between climate factors and I ACME was established using multiple regression,neural network,random forest and support vector machine methods in Wucaiwan mining area.The results indicated that the random forest of sparrow search algorithm(SSA-RF)had the highest precision in modeling with R^(2) being 0.82 and E_(RMS) being 0.109 in verification set.SSA-RF had higher reli

关 键 词：干旱煤矿生态指数 煤尘指数 麻雀搜索算法 随机森林 采矿活动 

分 类 号：X171.4[环境科学与工程—环境科学]