Satellite-based Inversion Of Soil Water Content In Baisha River Basin

As an important part of surface water storage,soil moisture plays a key role in the hydrological processes of surface runoff,infiltration and evapotranspiration,and directly affects the surface water balance.In the flood season,the regional differences in soil moisture content play an important role in reflecting surface runoff and flood warning.Therefore,it is of great practical significance and scientific value to study the spatial distribution of soil moisture and estimate soil moisture content on a large scale.Taking the Baisha River Basin in Sichuan province as the study area,Landsat8,MOD05,DEM data and meteorological data were used in this study.Based on the Ts-EVI feature space,the soil moisture information in the study area was extracted from 2016 to 2018,and then the field measurement in 2019 was used.Based on the volumetric water content data and the TVDI index of the same period,a linear regression model for inverting the volumetric water content of the soil depth of 0-10 cm and 10-30 cm was constructed.Finally,the model was used to invert the soil volumetric water content of the study area in 8 December 2019,and the accuracy of the inverted result was verified.The main conclusions are as follows:(1)Using the Landsat8 data in the five periods from 2016 to 2019,the Ts-NDVI and Ts-EVI feature spaces of the study area were constructed based on the temperature vegetation drought index method.Both feature spaces show obvious triangle scatter plots,but the triangle features of Ts-EVI feature space are more obvious,and the coefficients of fit of the wet and dry edges are better than the Ts-NDVI feature space.(2)The Ts-EVI feature space was used to extract the soil moisture information from the 5th phase of the study area.The spatial distribution of soil moisture in the northern mountainous area is greater than that in the central and southern regions;in time,the rainy season(May to September)is greater than the non-rainy season,which is consistent with the climatic characteristics of the local subtropical humid monsoon.(3)Through correlation analysis of TVDI index and measured volumetric water content of soil layers at different depths,the results show that there is a significant negative correlation between them.By comparison,it is found that the correlation is better at 0-10 cm,the determination coefficient is 0.73,and the determination coefficient is reduced to 0.56 at 10-30 cm.The correlation coefficient also decreases with the increase of soil depth,and the correlation coefficient is-0.85 in 0-10 cm,decreased to-0.74 at 10-30 cm.(4)The regression equations of the inverted soil volumetric water content is constructed using the TVDI index and field measured volumetric water content data all pass the F test,and the regression equations are highly significant,indicating that TVDI can reflect the soil volumetric water content status.Therefore,TVDI can be a reasonable indicator for moisture.(5)The average accuracy of the inversion results of each soil layer is relatively high.At the depth of 0-10 cm,the highest accuracy of the soil moisture obtained by the volumetric water content regression model inversion is 93.15%,the lowest value is 59.58%,and the average accuracy is 79.57%.The difference between the highest precision and the lowest precision is 33.57%.At the depth of 10-30 cm,the highest accuracy is 99.30%,the lowest value is 59.54%,the average accuracy is 75.30%,and the difference between the highest and lowest accuracy is 39.76%.The inversion results show that the TVDI feature space can effectively monitor the soil moisture of the 0-30 cm soil layer,especially can stably reflect and indicate the soil moisture status of the surface layer(0-10 cm)soil layer.