【作者】 孟磊；

【导师】 冯启言；

【作者基本信息】 中国矿业大学 ， 环境科学， 2010， 博士

【摘要】 流域是地表水系的集水区域,流域内生态系统相对完整。在自然状态下,受气候、水文、地质活动的影响,流域生态演变是缓慢和渐进的,但是,在人类活动的干扰下,尤其是水利水电开发、矿山开采、毁林造田等活动,流域生态在短时期内发生了强烈的变化。其中,在煤炭资源蕴藏丰富的流域,煤炭资源的强烈开采严重破坏了水土资源,对小流域水文、土地利用、植被覆盖、景观等生态要素产生巨大影响,显著改变了流域生态系统。深入、系统地研究采煤驱动下流域生态演变规律,评价采煤对流域生态的影响,对于保护流域生态平衡,实现水土资源可持利用和煤炭资源绿色开采具有重要意义。基于上述背景,在国家环境保护公益性行业科研专项“煤炭井工开采的地表沉陷监测、预报及生态环境损害累积效应研究”(200809128)和国家863计划项目“矿山复杂地表环境下地物信息自动提取与目标识别的若干关键技术”(2007AA12Z162)支持下,本文以淮南泥河流域为例,根据流域生态学、地理学、开采沉陷学、景观生态学、生态影响评价等理论,采用遥感与地理信息技术,运用遥感影像、数字高程模型、矿井采掘工程资料等多源数据,从小流域尺度系统地研究了采煤驱动下流域地表水体、土地利用与景观格局演变规律,综合评价了采煤对流域生态的影响。论文取得以下主要成果:(1)采用Landsat MSS/TM数据进行了泥河流域水体提取和土地利用分类,引入NDVI、NDBI、MNDWI等指数特征以及TM Band 4的Mean、Variance、Entropy 3种纹理测度,建立了基于SVM的分层分类方案,结果表明所构建的分类方案具有较高精度,各年份分类结果Kappa系数均达到了0.8以上。(2)根据采煤对水体的影响特征,将流域水体划分为塌陷水体、受采煤影响水体和其它水体3种类型,提出了井下采场信息支持下流域不同类型水体识别方法与技术流程。泥河流域的识别结果表明塌陷水体自1987年的6.17hm~2增加至2009年的1031.55hm~2,泥河受影响范围由1987年的170.49hm~2增加至2009年的475.94hm~2,塌陷预计结果显示,2020年和2030年塌陷水体面积分别达到5879.43hm~2和7736.2hm~2,分别占流域水体总面积59.31%和66.25%。(3)提出了遥感影像分类后水体变化检测方法,综合采用动态度、水体变化检测、水体分布重心分析了采煤驱动下泥河流域地表水体时空演变特征,采用水体密度、区位熵指数比较了矿区内、外水体聚集特征差异。结果显示,1987年至2009年流域水体重心向塌陷水体扩展方向迁移了4.67km,矿区水体密度和区位熵指数逐年升高,2009年达到1987年的1.6倍以上。矿区水体密度和区位熵指数高于矿区外,2009年达到矿区外的2倍以上,说明矿区是流域内水体最为密集、演变最为剧烈的区域。构建了一种简单有效的流域水体演变采煤驱动指数(CMWDI),泥河流域年均CMWDI由1987-1994年的0.87升高至2006-2009年的4.13,表明采煤对流域水体演变的驱动力在逐渐增强。(4)采用土地利用变化动态度与面积变化幅度分析了泥河流域土地利用变化特征,构建了单一地类和综合土地利用变化采煤驱动指数(CMLDI),以泥河流域为例分别评价了流域和矿区尺度下土地利用变化的采煤驱动力。结果表明在流域尺度下采煤对水体演变的驱动力最强,其次是耕地、建设用地、林地和园地,矿区尺度下的单一地类与综合采煤驱动指数均大于流域尺度,无论矿区尺度还是流域尺度,综合采煤驱动指数呈波动性上升,采煤对流域和矿区土地利用变化的驱动力在逐渐增强。(5)采用RUSLE模型,提出了基于DEM的采煤塌陷盆地土壤侵蚀LS因子计算方法,并对潘北矿塌陷盆地土壤侵蚀进行了计算。结果表明,塌陷盆地最大土壤侵蚀模数比塌陷前增加78%,侵蚀模数增加显著的地区位于塌陷盆地边缘,在塌陷不积水的情况下,区域总侵蚀量比塌陷前增加23%,塌陷积水减少了区域内土壤侵蚀发生面积,使区域总侵蚀量仅增加0.4%。该方法适用于计算平原地区采煤塌陷盆地土壤侵蚀。(6)基于景观指数分析了采煤驱动下泥河流域景观格局演变规律。结果表明:在景观尺度上,泥河流域1987至2006年景观格局向破碎化、异质化和低连通性发展,2006年至2009年流域景观向连续化、均一化及高连通性方向发展;在类型尺度上,耕地是流域景观的模地,但是其优势度指数由1987年的63.38下降至2009年的58.85,塌陷水体优势度由1987年的0.12上升至2009年的2.69。塌陷水体斑块数量、边缘密度也逐渐升高,采煤对流域景观的异质化、破碎化和边缘效应的影响逐渐增强。采煤活动形成的斑块连通性大于流域其它斑块类型连通性。(7)筛选了采煤生态影响的关键因子,构建了平原小流域采煤生态响应指数(ERIcum),并提出了生态影响评价技术流程。泥河流域评价结果表明2009年至2030年流域植被覆盖指数、生物丰度指数逐渐降低,土地退化指数与水体密度指数逐渐升高,水体密度指数增长幅度大于其它指数变化幅度,导致生态响应指数升高。但是,采煤导致的生物多样性降低、植被覆盖减少等负面效应还应当引起重视。更多还原

【Abstract】 A watershed is the land area that drains into a stream system. Within the watershed, the ecological system is relatively integrated. In the watershed with rich coal resources, mining subsidence makes great damage to land resources; coal mine industrial site construction occupies a lot of eco-land, which have significantly changed the land use, vegetation cover and landscape pattern. Studying the ecological evolution rules of watershed driven by coal mining systematically and intensively and evaluating the eco-impact of coal mining have an important significance to protect watershed ecological system and realize sustainable utilization of water resource and green mining.This thesis presents a case study in Nihe watershed of Huainan, which focuses on the evolution rules of surface water bodies, land use and landscape pattern and the comprehensive ecological impact assessment of coal mining on small watershed scale, using Remote Sensing (RS) and Geographic Information System (GIS) techniques with multi-source data. This thesis is jointly supported by National Environmental Protection Specialized Fund for Commonweal Industry (200809128) and the National High Technology Research and Development Program of China (“863 Program”) (2007AA12Z162). Main contents of the thesis are listed as follows:(1) Water body and land use information was extracted from Landsat MSS and TM image data. A hierarchical classification strategy based on SVM was built. The result shows that the strategy has higher accuracy.(2) According to the features of coal mining impact on water bodies, three kinds of water bodies in a watershed are classified which are water bodies formed by subsidence (WBS), water bodies affected by coal mining (WBCM) and other water bodies (OWB). One kind of water body type identification technique was proposed, which is supported by mining field information.(3)Based on the water body extraction and land use classified results, a method for detecting water body change was proposed. Dynamic degree index, water body change detecting and spatial distribution center were used for systematically analyzing temporal-spatial evolution features. Using water body density and location entropy, the spatial clustering features within and without mining area were analyzed. A simple and effective water body change driving index of coal mining (CMWDI) was constructed.(4) The land use change characteristics of Nihe watershed were analyzed using land use dynamic degree and area change value. The single land use class and comprehensive land use change driving index of coal mining (CMLDI) were built and applied in Nihe watershed at different scales. The CMLDI of Nihe watershed shows the driving strength of water body change is the biggest of all land use classes. The single and comprehensive CMLDI at mining area scale are larger than that at watershed scale.(5) Using RUSLE model, a method for calculating LS factors of subsidence basin based on DEM was proposed and applied in Panbei coal mine. The result indicates that the maximum soil erosion modulus of subsidence basin increases. The significant increasing area is located on the edge of subsidence basin. Comparing with normal landscape, the total erosion value increases by 23 percent after subsidence without water logging. Under the condition of subsidence with water logging, the total erosion amount only increased by 0.4 percent compared with normal landscape.(6)The characteristics of landscape pattern change of Nihe watershed were analyzed based on landscape metrics. The landscape pattern became more fragmented, heterogeneous and lower connective from 1987 to 2006. However, it became more continuous, homogeneous and higher connective from 2006 to 2009. Farm land is the matrix of Nihe watershed, while its dominance declined from 1987 to 2009 and the dominance of SWB ascended from 1987 to 2009. The patch number and edge density of SWB increased from 1987 to 2009. The connectivity of the patch classes formed by coal mining is the highest of all classes.(7)The key factors of coal mining impact on plain small watershed ecosystem were selected and an eco-response index (ERIcum) was constructed. The technique process of assessment was proposed and applied in Nihe watershed. The result shows the ERIcum increases from 2009 to 2030. But more attention should be paid on the biodiversity destroy, vegetation cover decreasing and other adverse effects induced by coal mining in the future.更多还原