Estimation of GRACE water storage components by temporal decomposition

This manuscript version is made available under the CC-BY-NC-ND 4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0/ This author accepted manuscript is made available following 24 month embargo from date of publication (July 2017) in accordance with the publisher’s archiving policyThe Gravity Recovery and Climate Experiment (GRACE) has been in operation since 2002. Water storage estimates are calculated from gravity anomalies detected by the operating satellites and although not the true resolution, can be presented as 100 km × 100 km data cells if appropriate scaling functions are applied. Estimating total water storage has shown to be highly useful in detecting hydrological variations and trends. However, a limitation is that GRACE does not provide information as to where the water is stored in the vertical profile. We aim to partition the total water storage from GRACE into water storage components. We use a wavelet filter to decompose the GRACE data and partition it into various water storage components including soil water and groundwater. Storage components from the Australian Water Resources Assessment (AWRA) model are used as a reference for the decompositions of total storage data across Australia. Results show a clear improvement in using decomposed GRACE data instead of raw GRACE data when compared against total water storage outputs from the AWRA model. The method has potential to improve GRACE applications including a means to test various large scale hydrological models as well as helping to analyse floods, droughts and other hydrological conditions