Prospects of Satellite Gravimetry to Monitor Atmospheric Mass, Water Vapor Partial Pressure, and/or Atmospheric Net-Water Fluxes

<p align="justify">Global Atmospheric re-analysis efforts aim at the realistic reproduction of the time-variable atmospheric state and associated diagnostic quantities by assimilating a wide range of different types of observations into a numerical model. Satellite gravimetry as realized with GRACE and GRACE-FO is on the one hand sensitive to the three-dimensional distribution of mass in the atmosphere and on the other hand also to (rapid) variations in terrestrial water storage, that largely represent the time-integrated effect of precipitation and evapotranspiration. By means of an ensemble of three global atmospheric reanalyses (ERA5, MERRA2, JRA-55) and two operational numerical weather prediction models (IFS from ECMWF, ICON from the German Meteorological Service) we will discuss the current level of agreement among the atmospheric models for (i) hydrostatic surface pressure, (ii) water vapor partial pressure, and (iii) atmospheric net-water fluxes obtained from summing precipitation and evapotranspiration. For time-scales between a few days and several weeks, we find that atmospheric net-water fluxes show the highest spread among the different models, thereby indicating a need for additional observations of both precipitation and/or evapotranspiration that are not yet routinely assimilated. We also find that water vapor partial pressure is quite uncertain, but note that tropospheric products from dense GNSS station networks are promising to constrain the moisture distribution. Finally, we demonstrate that atmospheric mass is already well constrained in weather models due to the abundant availability of in situ barometers at the Earth&#8217;s surface so that utilizing a priori information from global atmospheric model data in satellite gravity data processing via the Atmosphere and Ocean De-Aliasing Level-1B Model is very well justified. In terms of meteorologic applications of future satellite gravity missions currently under consideration by both NASA and ESA, we thus recommend to further refine methodologies aiming at isolating signals of atmospheric net-water fluxes out of the time-variable gravity signatures instead of attempting to map any direct signatures of the atmospheric mass distribution.</p>