Evaluating Hydro-Meteorological Fluxes Using GRACE-Derived Water Storage Changes

Atmospheric and terrestrial water budgets, which represent important boundary conditions for both climate modeling and hydrological studies, are linked by evapotranspiration (E) and precipitation (P). These fields are provided by numerical weather prediction models and atmospheric reanalyses such as ERA5 and MERRA-2 and play an important role in the context of climate modeling for initialization and evaluation. Via the terrestrial water budget equation, water storage changes derived from products of the Gravity Recovery and Climate Experiment (GRACE) mission, combined with runoff (R) information can be used to assess atmospheric models.

Besides long-term and (sub-)seasonal time scales also higher temporal frequencies of only a few days are of interest as especially extreme events like floods mainly occur on short time scales. In this contribution we will investigate the agreement of GRACE data on different temporal scales with the P-E-R flux time series from recent atmospheric reanalyses (e.g. ERA5, MERRA-2). While the models exhibit large biases and fail to capture the long-term flux trends in P-E-R, modeled fluxes agree remarkably well with GRACE water storage changes on short time scales and significant correlations can be detected down to sub-monthly (daily) time scales.