Application Of Soil Moisture Monitoring In Wheat Field Using Radar And Optical Remote Sensing Data

The soil moisture directly affects the water and energy exchanges of the earth surface and atmosphere.It is the key input parameter for hydrological model,climate model,ecological model,and land surface processing model.It plays an important role in global water cycle,energy balance and climate changing research.In agricultural applications,soil moisture is the basic condition of crops growing and developing.At the same time,it is also an important parameter in the research of crop growth monitoring,estimation and drought monitoring.It is difficult to achieve high-frequency monitoring of soil moisture at the regional scale on the strength of the traditional survey sites.In addition,it is time consuming and laborious.Optical remote sensing is difficult to measure soil moisture,which is easily impacted by weather conditions such as cloud,rain,haze,as well as the limitation of surface vegetation coverage.The active microwave remote sensing,with all-time and all-weather monitoring capabilities,has certain penetration of sparse vegetation and it is sensitive to soil moisture.Therefore,it can make up for the deficiency of optical remote sensing and it is an effective method for soil moisture monitoring in vegetation cover area.This study selects the winter wheat fields as the research area,which is part of the main winter wheat production area in North China Plain,in Wuji County,Shijiazhuang City,Hebei Province.And the winter wheat and soil conditions were observed with high frequency from January to June in 2017.Then the multi-period Sentinel-1 radar data and Sentinel-2 optical data were used to inverse wheat soil moisture at different growth stages.Firstly,the relationship between soil backscattering coefficient,soil moisture and land surface roughness were established based on Advanced Integrated Equation Model(AIEM).Secondly,this study used the relationship between D-value of VH and VV polarization and the combination roughness to calculate the soil moisture at low vegetation cover period.Then,the optical images were used to calculate the water content of the vegetation and the water cloud model was used to calculate the contribution of canopy backscattering.Finally,the backscattering contribution of wheat canopy was removed from the backscattering coefficient of the radar images and the soil moisture during high vegetation cover period was obtained,based on the empirical relationship between soil moisture,backscattering coefficient and combined roughness.Comparing the soil moisture simulation results and measured values over the multi-period,we can conclude that:(1)The moisture inversion results of most pixels are better during the period of low vegetation coverage(R~2=0.84,RMSE=2.27%).(2)The inversion accuracy in the period of high vegetation coverage is slightly worse than the low vegetation cover period,because the influence of vegetation canopy on backscattering is difficult to eliminate completely in high vegetation cover period(R~2=0.58,RMSE=3.36%).(3)The results had a high inversion accuracy for the entire winter wheat growing season(R~2=0.73,RMSE=2.95%).This indicates that the soil water inversion method proposed in this study has practical applicability.At the same time,it is proved that this method is feasible to calculate the combined roughness by using the same polarization and cross polarization data.