Monitoring soil moisture through assimilation of active microwave remote sensing observation into a hydrologic model

Soil moisture can be estimated from point measurements, hydrologic models, and remote sensing. Many researches indicated that the most promising approach for soil moisture is the integration of remote sensing surface soil moisture data and computational modeling. Although many researches were conducted using passive microwave remote sensing data in soil moisture assimilation with coarse spatial resolution, few researches were carried out using active microwave remote sensing observation. This research developed and tested an operational approach of assimilation for soil moisture prediction using active microwave remote sensing data ASAR (Advanced Synthetic Aperture Radar) in Heihe Watershed. The assimilation was based on ensemble Kalman filter (EnKF), a forward radiative transfer model and the Distributed Hydrology Soil Vegetation Model (DHSVM). The forward radiative transfer model, as a semi-empirical backscattering model, was used to eliminate the effect of surface roughness and vegetation cover on the backscatter coefficient. The impact of topography on soil water movement and the vertical and lateral exchange of soil water were considered. We conducted experiments to assimilate active microwave remote sensing data (ASAR) observation into a hydrologic model at two field sites, which had different underlying conditions. The soil moisture ground-truth data were collected through the field Time Domain Reflectometry (TDR) tools, and were used to assess the assimilation method. The temporal evolution of soil moisture measured at point-based monitoring locations were compared with EnKF based model predictions. The results indicated that the estimate of soil moisture was improved through assimilation with ASAR observation and the soil moisture based on data assimilation can be monitored in moderate spatial resolution.