Integrating satellite soil-moisture estimates and hydrological model products over Australia

Abstract

Accurate soil-moisture monitoring is essential for water-resource management and agricultural applications, and is now widely undertaken using satellite remote sensing or terrestrial hydrological models’ products. While both methods have limitations, e.g. the limited soil depth resolution of space-borne data and data deficiencies in models, data-assimilation techniques can provide an alternative approach. Here, we use the recently developed data-driven Kalman–Takens approach to integrate satellite soil-moisture products with those of the Australian Water Resources Assessment system Landscape (AWRA-L) model. This is done to constrain the model’s soil-moisture simulations over Australia with those observed from the Advanced Microwave Scanning Radiometer-Earth Observing System and Soil-Moisture and Ocean Salinity between 2002 and 2017. The main objective is to investigate the ability of the integration framework to improve AWRA-L simulations of soil moisture. The improved estimates are then used to investigate spatiotemporal soil-moisture variations. The results show that the proposed model-satellite data integration approach improves the continental soil-moisture estimates by increasing their correlation to independent in situ measurements (∼10% relative to the non-assimilation estimates). Highlights Satellite soil-moisture measurements are used to improve model simulation. A data-driven approach based on Kalman–Takens is applied. The applied data-integration approach improves soil-moisture estimates.