基于GlobeLand 30数据和CA__Markov模型的郑州市2000-2020年地表覆盖变化特征及预测分析  

Land Use and Land Cover Feature Analyses in Zhengzhou City During 2000 to 2020 Based on GlobeLand 30 and CA__Markov Model

作　　者：刘培[1,2] 贾守军 马朝阳 卢晓峰[1,2] 韩瑞梅[1,2] 贾函 

机构地区：[1]河南理工大学测绘与国土信息工程学院,河南焦作454000 [2]河南理工大学矿山空间信息技术国家测绘与地理信息局重点实验室,河南焦作454000 [3]北京奥腾岩石科技有限公司,北京101102

出　　处：《水土保持通报》2017年第4期282-287,共6页Bulletin of Soil and Water Conservation

基　　金：国家自然科学基金项目"基于半监督随机森林的城市地表覆盖主被动遥感数据协同分类研究"(41601450);河南省高等学校重点科研项目(16A420004);河南省高校基本科研业务费专项(NSFRF140113);河南理工大学博士基金项目(B2015-20);河南省基础与前沿项目(152300410098)

摘　　要：[目的]针对不同分类算法对地物识别结果的差异性导致地表覆盖变化分析难以为城市变化提供参考信息和决策依据问题。[方法]研究以GlobeLand 30标准产品为数据源,结合CA_Markov模型,对郑州地区2000—2020年土地利用变化速度、幅度、程度,以及熵值演变方向等时空变化特征进行分析与模拟预测。[结果](1)2000—2010年,郑州地区人造地表显著增加,湿地、耕地、草地有一定程度的减少,其中人造地表和湿地的变化速度相对较快;(2)2000—2010年,郑州地区土地利用程度变化量为8.04,土地利用信息熵和均衡度都有所提升,优势度相应降低;(3)预测2020年郑州地区地表覆盖状态同2010年相比,人造地表和草地分别增加68.88%和49.99%,水体、湿地增加幅度均在30%以上,耕地、林地有一定程度的减少。[结论]2000—2020年,郑州地区土地利用总体分布格局具有显著性差异,土地利用的复杂性增加,环境问题日益凸显。[Objective] The common problem in the processing of remote sensing images, that is different classification algorithms may lead to different identification results of ground objects, was discussed, to provide reference information of urban change and decision making based on LULC analysis. [Methods] In this research GlobeLand 30 product and CA_Markov model were selected to monitor and analyze land use degree, entropy change of land use, spatial-temporal change of land use in Zhengzhou area from 2000 to 2020. [Results] （1） There was a great land use degree change from 2000 to 2010. During this period human-made cover type was increasing greatly; on the contrary, wetland, farmland, grassland were in a decreasing trend.Where human-made earth surface and wetland changed significantly. （2） The amount of land use was 8.04, entropy of land use and equilibrium degree were promoted, while dominance index decreased. （3） There were 68.88% and 49. 99% increases for human-made earth surface and grass land respectively, and there were more than 30% increases for both of water surface and wetland. [Conclusion] There was a significant difference of landscape from 9,000 to 2020. The complexity of land use was increasing, and the proportionality of land use raised.

关 键 词：GlobeLand 30 CA-Markov模型 地表覆盖变化 预测分析 郑州市 

分 类 号：TP79[自动化与计算机技术—检测技术与自动化装置] Q948.1[自动化与计算机技术—控制科学与工程]