Use of Gravity Recovery and Climate Experiment terrestrial water storage retrievals to evaluate model estimates by the Australian water resources assessment system

Abstract

Terrestrial water storage (TWS) estimates retrieved from the Gravity Recovery and Climate Experiment (GRACE) satellite mission were compared to TWS modeled by the Australian Water Resources Assessment (AWRA) system. The aim was to test whether differences could be attributed and used to identify model deficiencies. Data for 2003–2010 were decomposed into the seasonal cycle, linear trends and the remaining de‐trended anomalies before comparing. AWRA tended to have smaller seasonal amplitude than GRACE. GRACE showed a strong (>15 mm yr−1) drying trend in northwest Australia that was associated with a preceding period of unusually wet conditions, whereas weaker drying trends in the southern Murray Basin and southwest Western Australia were associated with relatively dry conditions. AWRA estimated trends were less negative for these regions, while a more positive trend was estimated for areas affected by cyclone Charlotte in 2009. For 2003–2009, a decrease of 7–8 mm yr−1 (50–60 km3 yr−1) was estimated from GRACE, enough to explain 6%–7% of the contemporary rate of global sea level rise. This trend was not reproduced by the model. Agreement between model and data suggested that the GRACE retrieval error estimates are biased high. A scaling coefficient applied to GRACE TWS to reduce the effect of signal leakage appeared to degrade quantitative agreement for some regions. Model aspects identified for improvement included a need for better estimation of rainfall in northwest Australia, and more sophisticated treatment of diffuse groundwater discharge processes and surface‐groundwater connectivity for some regions.