基于遥感和CA_Markov模型的煤矿区热环境与土地覆盖变化模拟评价

摘要

针对遥感影像在定量描述土地覆盖和热环境分布方面的优越性,利用支持向量机（SVM）分类器对遥感影像进行分类,获取土地覆盖图,利用单窗算法反演获取地表温度。通过将元胞自动机和马尔科夫模型结合,构建CA_Markov模型对土地覆盖和热环境时空格局进行模拟与分析。采用定量景观指数和土地利用转移矩阵进一步挖掘了研究区空间信息,综合模拟与预测了土地覆盖和热环境时空变化特征。实验结果表明：利用遥感影像定量反演热环境特征时空分布与预测是可行的,CA_Markov模型能够有效地揭示和预测矿区土地覆盖变化和热环境演变趋势。%According to the merits of remotely sensed data in depicting regional land cover and thermal environment, multi-objective information processing was employed to remote sensing images to analyze and simulate land cover and thermal environment in mining areas. Support Vector Machine (SVM) classifier was used to classify remote sensing images and derive regional land cover maps, mono-window algorithm was used to retrieve land surface temperature from thermal infrared images. And land cover change and thermal environmental trends were simulated by CA_Markov mod- el. Quality landscape indexes and land use and land cover transfer matrixes were selected to analysis and assess the dynamic changes trend of different land cover types as well. The results demonstrate that it is signification and suitable to retrieval thermal environmental, evaluate and analyze its distribution and trend using remote sensing methods. CA_ Markov model is able to predict land cover and thermal environmental trends with good performance.