[Remote sensing quantitative retrieval of soil water content in desertification region]

An empirical model was set up by using TM data to get the reflectance of the earth surface in Naiman Banner of Kerqin Sandy Land, and the influence of vegetation on the pixel reflectance was eliminated by optical vegetation coverage model. The regression model between soil water content and earth surface reflectance was set up to retrieve soil water content, and the accuracy of the retrieved results were validated by in situ data. The results showed that the theoretical and practical precision was 81.81% and 92.17%, respectively, suggesting that the remote sensing quantitative retrieval of soil water content was scientifically available.