Soil Moisture Retrieval from Remote Sensing Data in Arid Areas Using a Multiple Models Strategy

Due to the heterogeneity of land surface features in a large-scale region, this study examines the effectiveness of a multiple models strategy in soil moisture retrieval from MODIS in Xinjiang, and established a robust model for soil moisture retrieval from remote sensing data in arid and semi-arid areas where diverse land surface covers such as bare soil, sparsely vegetated and densely-vegetated lands exist. Specifically, the composed inversion models: ATI-based, TVDI-based and an averaged model were developed and used to estimate land surface soil water content in bare-soil, densely-vegetated and sparsely-vegetated areas, respectively. The models were verified and validated using the in-situ measured data, and the result indicates that, the proposed composite model performed better and achieved more accurate estimate of Xinjiang surface soil moisture than either ATI-based model or TVDI model.