Surface Soil Moisture Retrieval In Wheat Covered Area Using Multi-temporal SAR And Optical Satellite Data

In agriculture,soil moisture is a basic condition for crop growth and development,which runs through all aspects of crop growth.Meanwhile,it is an important parameter for crop growth monitoring,drought warning,and yield estimation.Therefore,it is necessary to monitor the spatial and temporal distribution of soil moisture effectively and accurately.It has important practical significance and scientific value for crop yield estimation,drought monitoring,water resources distribution,ecological protection,and so on.It also helps to implement precision agriculture and promote the development of modern agriculture.Because the traditional methods of soil moisture monitoring are based on point measurement,it is difficult to obtain soil moisture in large-scale area economically and quickly.Remote sensing technology can obtain the surface information of land characteristics in real time,quickly and in a large range,which is widely used to obtain the regional surface soil moisture.Taking Dingxing County of Hebei Province as the study area,this study established a surface soil moisture inversion method suitable for different vegetation coverage conditions in farmland by using modified water cloud model（MWCM）and coupled empirical model（CEM）.The research was based on Radarsat-2 C-band synthetic aperture radar（SAR）data,Landsat-8 optical data and field measured data.Firstly,in view of the inapplicability of water cloud model（WCM）in areas with sparse vegetation coverage,this study introduced vegetation fraction to obtain the modified water cloud model.The effects of WCM and MWCM on vegetation removal were compared.Then,the dual polarization backscattering coefficient database was established based on the advanced integrated equation model（AIEM）and the Oh model.The response relationship between the bare soil backscattering coefficient and both soil moisture and surface roughness was analyzed,and the effective roughness parameter was simulated.Thus,the coupled empirical model（CEM）was established.Finally,combined with MWCM and CEM,MWCM-CEM,a method suitable for surface soil moisture inversion during the wheat growth cycle was proposed.Comparison experiments were designed to verify the effectiveness of the proposed method.MWCM-CEM was applied to soil moisture inversion in four important phenological periods of winter wheat in the study area,and the spatial distribution maps of soil moisture in different phases were obtained.The main research contents and conclusions of this paper are as follows:（1）It is difficult to use the AIEM theoretical model to retrieve soil moisture directly due to its complex expression form and various parameters.In this study,combined with field measured data,a dual polarization backscattering coefficient database was established based on the AIEM model and the Oh model.The response relationship between backscattering coefficient,soil moisture and surface roughness was obtained through regression analysis,and the CEM for bare soil moisture retrieval was established.（2）Aiming at the inaccuracy of surface roughness measurement in the field,the effective roughness parameters were simulated by look up table（LUT）based on the dual polarization backscattering coefficient database.The results indicated that using effective roughness parameters instead of measured surface roughness as model input can effectively avoid the uncertainty in the process of surface roughness measurement,and can realize the soil moisture retrieve in the absence of field measured surface roughness parameters.（3）In view of the limitation of WCM in the application of sparse vegetation coverage area,vegetation fraction was introduced into WCM as a weight factor,and the effects of removing vegetation influence by WCM and MWCM were compared.Aiming at the multi-temporal soil moisture inversion in wheat region,this study proposed MWCM-CEM,a new soil moisture inversion method.In order to verify the effectiveness of this method,this study set four soil moisture inversion methods and carried out comparative experiments.The results showed that the MWCM considering vegetation fraction can remove the vegetation backscattering contribution more effectively and improve the accuracy of soil moisture inversion.Compared with WCM-CEM_er,the R² of soil moisture inversion results under MWCM-CEM increased by 0.1273 and the RMSE decreased by 0.97%.（4）Through theoretical analysis and comparative experiments,it is verified that MWCM-CEM proposed in this paper is suitable for soil moisture retrieval in crop coverage area.MWCM-CEM was used to retrieve the soil moisture at the returning green,jointing,filling and milky maturity stages of wheat in the study site,and the spatial and temporal distribution maps of soil moisture were obtained.