Uncertainties of GRACE‐Based Terrestrial Water Storage Anomalies for Arbitrary Averaging Regions

The application of terrestrial water storage (TWS) data observed with GRACE and GRACE-FO often requires realistic uncertainties. For gridded TWS data, this the knowledge covariances, which can be derived from formal, i.e., formally estimated in parameter estimation, variance-covariance matrix provided together Stokes coefficients. However, propagation monthly matrices to is computationally expensive, so we apply a spatial covariance model for data. provides non-homogeneous (location depending), non-stationary (time anisotropic (orientation depending) covariances between any two given points. Further, accommodates wave-like behavior East-West-directed residuals striping errors cause. main such estimation uncertainties mean time series arbitrary regions as river basins. Alternatively, regional above mentioned formal This study compares modeled basin GFZ RL06 ITSG-Grace2018 ITSG-Grace 2018 solution. fit both temporal domain. We further evaluate by comparison empirical over arid regions. Here, again appropriateness shown. results, namely global basins, are available via GravIS portal. provide Python toolbox, allows computing matrices.