RS、GIS支持下北京市昌平区耕地分等与评价研究.pdf

摘 要 以北京市昌平区为研究区，在RS 、GIS 技术支持下，以土壤学、土壤肥力 学等知识为理论基础，对该区耕地进行分等研究，并将其结果应用于该区基本 农田划分、高标准基本农田建设。以两个时相的Landsat TM 遥感影像为数据源， 充分利用 DEM 、2000 年现状图等辅助数据，采用分层分类法、掩膜分析逐一 将昌平区土地类型分为 6 类，结果表明，分层分类法、掩膜分析并结合辅助数 据，分类精度较高。通过遥感影像地表反射率与对应样点土壤有机质含量的相 关性分析，建立土壤有机质定量反演模型，预测该区土壤有机质空间分布，并 对该模型的精度进行验证。利用标准粮实地调查数据，建立耕地质量评价指标 体系，在 ArcGIS10.0 平台下建立昌平区耕地质量评价数据库，对该区耕地进行 分等，分等结果与标准粮样点调查数据回归结果表明，AIS-DEA 分等结果可靠， 符合实情。将分等结果用于划分该区基本农田，结合昌平区土地整治数据、2012 年高标准基本农田建设任务，划分高标准基本农田建设类型，并有针对性地指 出建设内容。 关键词：土地利用，耕地分等，基本农田划分，高标准基本农田，区划 Abstract The cultivated land quality in Changping District, Beijing was evaluated supported by RS GIS technology and soil science knowledge, and the classification results were applied to demarcate high standard prime farmland in Changping District. Two multi-temporal TM satellite remote sensing imagery were selected to extract Changpings land use information by stratified classification method and Mask analysis, and the results showed the method had a higher accuracy.Based on soil field-survey data, TM imagery on 18th November in 2010 was selected as data source to build retrieved model by analysing the correlation between the SOM samples and spectral reflect characteristic, with which the SOM spatial distribution were predicted, the accuracy of the model were tested. Cultivated land evaluation database was established in ArcGIS10.0 with spatial data such as cultivated land distribution data, SOM data, etc.The cultivated land in Changping District was evaluated. Combined with cultivated land field-survey data, regression analysis results indicated that the evaluation results were reliable. At last, prime farmland wa