Improving soil water representation in the Australian Water Resources Assessment landscape model through the assimilation of remotely-sensed soil moisture products

Abstract

Many researchers over the last decade have demonstrated how the assimilation of satellite soil moisture data can improve the accuracy of soil water representation in land surface models, and result in improved estimates of evaporative flux, drainage, and runoff. In this study we investigated whether similar benefits are achievable for the Australian Water Resources Assessment landscape (AWRA-L) model through the assimilation of AMSR-E and ASCAT soil moisture (SM) products. The AWRA-L model was co-developed by CSIRO and Bureau of Meteorology researchers to support the Bureau's mandated reporting requirements on national water accounts and water resource assessments. AWRA-L represents the soil column as three conceptual storage layers: a top-layer (equivalent to the emitting soil layer for C- or L-band microwave radiometry); and separate layers for shallow- and deep-rooted vegetation respectively. AWRA-L was run cell-wise (i.e. no lateral flow) across the continent at 0.05-degree resolution providing estimates of daily water fluxes and stores. We used perturbed meteorological forcing and the ensemble Kalman filter (EnKF) to assimilate AMSR-E and ASCAT SM products into AWRA-L. Evaluations to-date have been conducted using the OzNet network of in situ moisture sensors, but will be extended to other parts of the continent via the network of cosmic ray probes and (indirectly) through evaluation of independent satellite SM retrievals. Preliminary results clearly show an improvement in AWRA-L top-layer SM estimation compared to open-loop simulations. Results for the impact on root-zone soil layers are mixed, but appear to be linked to combinations of prescribed SM error and/or strength of vertical coupling between soil layers. Further investigations will identify where and when the assimilation of satellite SM benefits AWRA-L estimation in terms of soil water status and runoff estimation across Australia.