Propagation of atmospheric model errors to gravity potential harmonics-impact on GRACE de-aliasing

Abstract

SUMMARY High-frequency, time-varying mass redistributions in the ocean and atmosphere have an impact on GRACE gravity field solutions due to the space–time sampling characteristics of signal and orbit. Consequently, aliasing of these signals into the GRACE observations is present and needs to be taken into account during data analysis by applying atmospheric and oceanic model data (de-aliasing). As the accuracy predicted prior to launch could not yet be achieved in the analysis of real GRACE data, the de-aliasing process and related geophysical model uncertainties are regarded as a potential error source in GRACE gravity field determination. Therefore, this study aims to improve the de-aliasing process in order to obtain a more accurate GRACE gravity field time-series. As these time-series provide estimates for the integrated mass transport in the Earth system, like the global water cycle and solid Earth geophysical processes, any increase in accuracy will lead to improvements in the geophysical interpretation of the results. So in conclusion, improving the de-aliasing is of relevance for a better understanding of geophysical processes. By no longer regarding the atmosphere and ocean model output as error-free, deeper insight into the impact of such uncertainties on the de-aliasing and on the resulting GRACE gravity field models can be obtained. For this purpose, in a first step, a full error propagation of the atmospheric and oceanic model parameters up to the de-aliasing gravity field coefficients is performed and the GRACE K-Band-Satellite-to-Satellite Tracking (KBR-SST) residuals, as an intermediate gravity field result, are analysed. The paper reviews the standard GRACE de-aliasing process and presents the mathematical model applied for the error propagation. Specifically, the effect of uncertainties in the atmospheric input parameters (temperature, surface pressure, specific humidity, geopotential) on the gravity field potential coefficients used for de-aliasing is shown in several scenarios. Finally, the impact of dealiasing products (with and without error propagation) on a GRACE gravity field solution is investigated on the level of observation residuals. From the results obtained in this study it can be concluded that with respect to the current GRACE error budget, atmospheric model uncertainties do not play a prominent role in the error budget of current GRACE gravity field solutions. Nevertheless, in order to fully exploit the GRACE measurements towards the baseline accuracy, an optimized de-aliasing is needed. In this case, GRACE gravity field solutions are sensitive to uncertainties in atmospheric and oceanic models. Thus, the associated geophysical model errors shall be taken into account in the de-aliasing process.