Data for: Multi-climate mode interactions drive hydrological and vegetation responses to hydroclimatic extremes in Australia
GRACE-derived TWSA from ensembling six scaling factors: We used the standard TWSA products for the Australian continent during the entire GRACE period of April 2002 to January 2017 (Landerer and Swenson 2012; Swenson and Wahr 2006). These products were derived from spherical harmonics (release-5) at three independent processing centres: the Jet Propulsion Laboratory (JPL), the Center for Space Research at the University of Texas (CSR) and the German Research Center (GFZ). The derived monthly data grids (1° × 1°) represent anomalies relative to the 2004-2009 time-mean baseline, which have been pre-processed to remove signals from the atmosphere and oceans, and were retrieved as Equivalent Water Height (EWH) in units of centimeters (details in GRACE Tellus website). An ensemble mean of TWSA from the three GRACE products was further calculated to minimize uncertainties associated with different gravity-field solutions (Sakumura et al. 2014). In addition, scaling factors are essential for restoring much of the energy removed during GRACE data processing (e.g. destriping, Gaussian filtering and truncation) to the land grids. Here, we chose to use the scaling factor averaged from six Land Surface Models (CLM4.0, WGHM2.2, and GLDAS-1 models of Noah2.7, Mosaic, VIC, and CLM2.0) described in Long et al. (2015a). This ensemble scaling factor can reduce errors due to the differences among different scaling factors whilst also include land models that account for human activities as well as surface water and groundwater interactions (Long et al. 2015a; Long et al. 2015b).