The land use and land cover change (LUCC) is one of the prime driving forces of climate change. Most attention has been paid to the influence of accuracy of the land cover data in numerous climate simulation projects. The accuracy of the temporal land use data from Chinese Academy of Sciences (CAS) is higher than 90%, but the high-precision land cover data is absent. We overlaid land cover maps from different sources, and the grids with consistent classification were selected as the sample grids. By comparing the results obtained with different decision tree classifiers with the WEKA toolkit for data mining, it was found that the C4.5 algorithm was more suitable for converting land use data of CAS classification to land cover data of IGBP classification. We reset the decision rules with Net Primary Productivity (NPP) and Normalized Difference Vegetation Index (NDVI) as the indicators. The accuracy of the reclassified land cover data was proven to reach 83.14% through comparing with the Terrestrial Ecosystem Monitoring Sites and high resolution images. Therefore, it is feasible to produce the temporal land cover data with this method, which can be used as the parameters of dynamical downscaling in the regional climate simulation.
Land Cover Mapping Based on Multisource Spatial Data Mining Approach for Climate Simulation: A Case Study in the Farming-Pastoral Ecotone of North China