1Department of Land Resource Management, China University of Geosciences, Wuhan, Hubei 430074, China
2Department of Geography and Center for Environment Sciences and Engineering, University of Connecticut, Storrs, CT 06269-4087, USA
The food and ecological security have been a universal and important issue taking place in China, current techniques for quantifying cropping intensity may not accurately map smallholder farms where the size of one field is typically smaller than the spatial resolution of readily available satellite data. The purpose of this study was to develop a downscaling method for classifying detailed agricultural land use classes with remote sensing data, which allows prediction at a finer spatial resolution than that of the input imagery. A local regression method called geographically weighted regression (GWR) is introduced for downscaling the Landsat ETM + imagery, and it is compared with cokriging for this purpose. The case studies show that the GWR approach is an effective technique to downscale the image, although the image classification is still not perfect. GWR performed marginally better than the complex cokriging method, which too has proven to be an effective method, but is considered fewer variables. The result also shows the different accuracies between different downscaling sizes. Which demonstrate that the GWR approach can improve great performance for downscaling.
Keywords:downscaling; cultivated land use classification; geographically weighted regression; cokriging; TM