Soil Moisture Retrieval Of Grassland In Guizhou Plateau Based On Multisource Remote Sensing

Soil moisture is the key factor of land and atmosphere energy exchange,plays an important role in the water cycle,and also plays an important role in hydrological climate,drought forecast,crop yield estimation and other aspects.Active microwave remote sensing has the advantages of all-weather,all-weather and strong penetrating ability,which enables it to obtain relatively complete ground feature information under cloudy and rainy climate conditions,and provides reliable data for the quantitative and timed research on crop growth,soil moisture change monitoring,biomass inversion and many other aspects.In recent years,with the radar remote sensing data is more and more rich,the use of radar for soil moisture inversion research more and more,but for this kind of guizhou plateau more cloud region related research is relatively lag,based on the detailed full polarization of Radarsat-2 radar data and Landsat 8 multispectral data,the measured soil moisture,soil texture,soil bulk density and the surface roughness parameters such as data,using the advanced integral equation model(AIEM),TOPP model,Dobson ratio model and model,Two methods were proposed for the inversion of soil moisture at 0-5cm and 5-10 cm depths in the assixi steppe in the mountainous area of the Guizhou plateau.The first method is based on the soil moisture inversion of AIEM model,TOPP model and ratio model.First of all,were obtained through field soil samples of different soil depth,and the surface roughness of image data and Landsat-8 multispectral data for soil volumetric water content,soil parameters such as surface roughness,and four vegetation index are(SAVI,DVI,NDWI,NDVI),combined with radar system parameters(incident Angle and the incident frequency and wave number),then use AIEM model,TOPP model as well as the ratio model bare backscatter coefficient on the surface of the earth,after the AIEM model is mainly used to simulate the bare surface to the scattering coefficient;TOPP model is mainly used to transform soil volume water content into soil dielectric constant to parameterize AIEM model.The ratio model is used to eliminate the contribution of vegetation to the backscattering coefficient.Finally,the relationship between backscattering coefficient of bare surface and measured soil moisture is established to obtain the soil moisture content in the study area.Among them,the ratio model adopts the ratio of the simulated backscattering coefficient of bare surface to the total backscattering coefficient observed from the image and the four-planting cover index for parameterization,and the unknown coefficient is solved by the least square method.The second method is based on the soil moisture inversion of AIEM model,Dobson model and ratio model.This method is similar to the first method in general,except that the Dobson soil dielectric model is used in the conversion of soil water content to soil dielectric constant.Compared with TOPP model of the first kind,Dobson model is better than TOPP model in applicable frequency range and soil texture range.The above two methods were used to obtain the soil moisture content in the soil depth of 0-5cm and 5-10 cm respectively in the study area,and the following conclusions were drawn: first,the inversion effect of VV polarization on soil moisture inversion is better than that of HH polarization.Secondly,for the two methods,the inversion effect of method 2 is better than that of method 1.Then,for soil depth,the inversion result of 0-5cm soil depth is better than that of 5-10 cm.Finally,the NDWI inversion was better than the other three vegetation indexes for the four planting indexes.In conclusion,when the soil depth is 0-5cm,the polarization mode is VV polarization,and the vegetation parameter is NDWI,the inversion accuracy of method 1 is higher.