中文摘要：

现有非点源污染模型在模拟我国零散种植、精耕细作和复种多熟耕地的非点源污染时存在一定局限性。针对此现状，开展基于遥感的能反映耕地零散种植、精耕细作和复种多熟特点的复种指数提取研究，可为现有非点源污染模型的改进提供基础数据。以鄱阳湖的源头区域一梅江流域作为研究区，以MODl3QI的NDVI数据作为数据源，采用HANTS方法去噪与重构NDVI时序数据，在GIS空间分析模块上设计二次差分算法建立复种指数提取模型，并结合梅江流域2010年的耕地利用现状数据提取梅江流域2010年和2011年的复种指数。经野外采样数据验证，复种指数的提取精度为83．3％，具有较高的精度，能够满足现有非点源污染模型改进的需要。

英文摘要：

According to the situation that the existing non-point pollution models exist a certain limi- tation when simulating in scattered planting, intensive cultivation and multiple cropping＇s cultivated land ,the paper carry out the multiple cropping index extraction research based on remote sensing that can reflect the cultivated land＇characteristics to provide the basis data for non-point pollution. This paper chose MeiJiang basin as the study area,with the MOD13QI NDVI data as data source, used the HANTS method to reconstruct the time series NDVI data, designed twi-differential algorithm to build multiple cropping index model in the GIS spatial analysis module, and combined with 2010 MeiJiang Basin＇s cultivated land-use status data to extract 2010 and 2011 ＇s MeiJiang basin multiple cropping index. Validated by sampling data, the accuracy of multiple cropping index is 83.3 %, hay-ing high precision,which can satisfy the improvement need of existing non-point pollution model.