Study On Soil Moisture Retrieved By Remote Sensing In Jinghe Watershed Oasis

Soil moisture is basic important in researches on agriculture, water resources,meteorology and ecology. It is also an important indicator in soil degradation monitoring andenvironment supervision. Study soil moisture retrieved by remote sensing is significant forimproving the utilization of soil moisture.In this research, after evaluation of all the models, TVDI (Temperature VegetationDryness Index) is selected for soil moisture retrieved by remote sensing in oasis of Jinghewatershed. This work aims to conduct temporal soil moisture retrieval in Jinghe watershed,explore its spatial and temporal distribution characteristics and analysis of influencingfactors. Major ones are as follows:(1) Comparing surface temperature retrieval algorithm, single window algorithm anduniversal single-channel algorithm were selected to retrieve surface temperature. Regressionanalysis was made between results of surface temperature retrieval and contemporanceousfield measured data, the correlation coefficient were0.8604and0.8027. Results shows thatsingle window algorithm and universal single-channel algorithm could all be used toretrieval surface temperature, except that the single window algorithm had a bettercorrelation with suface temperature compared to the universal single-channel algorithm.Therefore, single window algorithm was selected to retrieval the near suface temperature.(2)Surface temperature and MSAVI were used to establish the Ts-MSAVI characteristicspaces. The TVDI was calculated from Ts-MSAVI characteristic spaces, which was used toretrieve soil moisture. Regression analysis was made between results of soil moistureretrieval and contemporanceous field measured data, the correlation coefficient was0.8006.Therefore, TVDIm can effectively reflect the soil moisture.(3)The result was calculated from Landsat TM/8data could be blended with MODISLST data and NDVI. By using the temporal and spatial extension method could improvetemporal resolution of TVDI from MODIS data, but also improve the spatial resolution ofTVDIm from landsat TM/8data.The method established the high resolution and long timeseries of were soil moisture in Jinghe watershed.(4)Regression analysis was made results of surface temperatureand field measured data,precipitation data and contemporanceous TVDI. Results shows that rising temperaturescould reduce soil moisture, and rising precipitation would increase soil moisture. Throughcomparative analysising the different land use types corresponding to average of TVDI,there was a similar trend on the temporal and spatial. Soil moisture were reducting in thegrassland, woodland, farmland, urban and unused land. Work through more than that retrieval soil moisture, explore its spatial and temporaldistribution characteristics and analysis of influencing factors were completed in Jinghewatershed. The result that TVDI ways to direct the use of remote sensing data of soilmoisture inversion, inversion of the accuracy of medium-related, the method is fit tomoisture by remote sensing in the medium-arid areas.