Preliminary results of the GFZ daily GRACE/GRACE-FO Kalman filter solution

Abstract

Within the framework of the Research Unit NEROGRAV (New Refined Observations of Climate Change from Spaceborne Gravity Missions), we focus on the computation of daily gravity field solutions using a Kalman filter approach. We use the RL07p GRACE/GRACE-FO daily normal equations, which incorporate refined stochastic modeling of the observations and background models. The stochastic information on the geophysical processes governing submonthly gravity field variations is derived from HISerr time series.In this study, we present GFZ daily gravity field solutions for the five-year period from 2004 to 2008. To investigate the impact of different processing strategies, several Kalman filter configurations are tested. These include the use of different a priori gravity field models, different approaches to constructing the required hydrological variance-covariance matrix, and the application of a Kalman smoother in comparison with the corresponding raw Kalman filter solutions. The resulting solutions are assessed in terms of their consistency and sensitivity to the different processing choices.Validation is performed in two ways. First, the GFZ daily solutions are compared with independent daily Kalman filter solutions computed at TU Graz. Second, the daily gravity field solutions are evaluated by comparison with the vertical components of GNSS stations from the global IGS network. This provides an external assessment of the temporal variations captured by the daily gravity field solutions.