Prediction and Mapping of Soil Depth at A Watershed Scale with Fuzzy-c-means Clustering Method
Soil depth of the west Tiaoxi catchment was predicted using fuzzy c-means clustering(FCM) based on the relationships between soil depth and landscape parameters. Six terrain factors, I.e., elevation, slope, planform curvature, profile curvature, runoff intensity and topographic wetness index were clustered, then the whole catchment was classified into eight combinations of these factors. Typical soil depths from the training soil dataset, combined with attribute of samples and expert knowledge, were assigned to each cluster center. Soil depth map was predicted with weighted average model. Results showed that, FCM method could rationally and effectively classify the combination of terrain factors, and it is a low cost and high efficiency mapping method with satisfactory prediction precision and model stability and could be possible applied to areas with the similar landscape conditions.
国家科技期刊平台
NSTL
万方数据
维普
