耕地轮作模式遥感监测

Monitoring the Pattern of Crop Rotation Through Remote Sensing

研究目的：探索耕地轮作模式空间格局的遥感监测方法。研究方法：基于变化向量分析技术，分析多时相NDVI变化的大小、方向与耕地轮作模式的响应规律，构建基于变化向量分析的耕地轮作模式遥感监测模型。研究结果：利用多时相NDVI信息进行变化向量分析，能有效地监测研究区耕地轮作模式空间格局；研究区以冬小麦一夏玉米轮作模式为主，其次是春玉米、休闲一春玉米一冬小麦模式，露天菜地与大棚菜地所占比例较少；以野外实测样本进行精度验证，本研究提出的耕地轮作模式监测方法的总体精度为80．36％，其中冬小麦一夏玉米一年两熟模式和春玉米一年一熟模式的监测精度最高，分别可达90％和85．71％。研究结论：多时相NDVI变化向量分析法为区域耕地轮作模式分布面积与空间格局监测提供了一种新的思路。

The purpose of this paper is to explore the monitoring method of the crop rotation pattern through remote sensing. Method employed was the NDVI analysis based on the technology of vector change analysis. The model of monitoring crop rotation modes based on the NDVI analysis was further developed. Results showed that the spatial pattern of crop rotation could be effectively monitored by analyzing the changed vector of multi-temporal NDVI information. Among the crop rotation patterns in the studied area, crop rotation of winter wheat and summer maize accounted for the major part. The other pattern such as spring maize-recreation-winter wheat rotation, vegetable production with and without green house facility was not popular. Comparing with the in-situ samples information, the paper evaluated the accuracy of the proposed method, and the assessed accuracy was 80.36%. The accuracy of the pattern of winter wheat and summer maize was 90%, and that of the single cropping pattern of spring maize was 85.7%. It was concluded that the method of changed vector analysis based on NDVI information could provide a new idea for mapping the regional crop rotation pattern.