基于MODIS时间序列数据的作物季相信息提取  

Crop Phenology Information Extraction Using MODIS Multi-temporal Data

作　　者：颉继珍[1,2] 王红说[2] 黄敬峰[3] 

机构地区：[1]甘肃省测绘工程院,甘肃兰州730000 [2]香港中文大学 [3]浙江大学农业遥感与信息技术应用研究所,浙江杭州310029

出　　处：《遥感技术与应用》2010年第5期647-652,共6页Remote Sensing Technology and Application

基　　金：国家863项目(2006AA12010103)资助

摘　　要：基于MODIS NDVI时间序列数据对浙北平原单季稻区进行作物季相一致性分析。对NDVI时间序列数据进行离散傅立叶变换去除噪声,再利用土地利用现状图提取耕地区的NDVI影像图,根据时间序列曲线的最大值研究作物的季相。结果表明:水稻生长期对NDVI时间序列曲线的响应和季相一致性均较小麦和油菜好;8d合成的数据较16d合成的数据可以更详细地反映作物季相信息。研究证实了MODIS NDVI时间序列曲线对区域作物季相分析的意义。Phenology consistency information was extracted from MODIS NDVI time series data of plain sin-gle-rice district northern Zhejiang.NDVI time series data were treated using discrete fourier transformation method in order to rectify the wrong data.Land use map was used to derive the NDVI time series data of cultivated land in study area.Maximum values of time series data per pixel were extracted to make analysis about the regional crop phenology characteristics.The result showed that rice had better phenology consis-tency than rape and wheat,and phenology information got from 8days composite data was more detailed than that from 16days composite data.Growing period response of rice to NDVI time series was much bet-ter than that of rape and wheat.The result indicates that the potential senses of MODIS NDVI time series to analyzing regional crop phonologies.

关 键 词：作物季相 MODIS 离散傅立叶变换 时间序列 

分 类 号：S165.21[农业科学—农业气象学] TP79[农业科学—农业基础科学]