基于CA-Markov模型的青海湖流域土地利用变化模拟

Simulation of Land Use Change in Qinghai Lake Watershed Based on CA-Markov Model

[目的]利用遥感影像在定量描述土地覆盖和环境变化方面具有的优越性,以青海湖流域为研究区,研究土地利用时空变化,以便对该地区的环境治理和生态恢复提供理论依据和技术支持。[方法]选取2001、2006、2011年3期TM遥感影像数据,利用支持向量机分类器分别对遥感影像进行分类得到每期的土地覆盖图,建立CA-Markov预测模型,经模型验证后,应用该模型对2016年青海湖流域土地利用结构进行预测。[结果]预测结果与实际解译结果接近。[结论]利用CA-MARKOV预测模型预测土地覆盖变化是可行的,能够有效揭示土地覆盖变化趋势,对决策规划、分析该地区土地荒漠化、生态环境变化规律及其成因具有重要意义。

[Objective] Using the advantages of remote sensing image in the quantitative description of land cover and environmental changes,with Qinghai Lake basin as the research area,land use temporal and spatial changes were studied in order to provide theoretical basis and technical support for the region＇s environmental governance and ecological restoration.[Method]Selecting TM remote sensing image data of three years in 2001,2006,2011,using method of support vector machine（ SVM） classifing for remote sensing image,as a result,every classification land cover map was obtained,through the establishment of the CA-MARKOV prediction model,model validation and application,the land use result in Qinghai Lake watershed in 2016 was predicted. [Result] The predicted result is close to the actual interpretation results. [Conclusion] It shows that CA-MARKOV prediction model is feasible in predicting land cover change,and can reveal the land cover change trend effectively,it is of great significance on decision-making planning,analysis of land desertification in the region,the ecological environment change rule and its formation.