Summary
Despite the increasing accuracies of GRACE/GRACE-FO gravity field models through worldwide endeavors, the temporal aliasing effect caused by the imperfect background models used in gravity field modelling is still a crucial factor that degrades the quality of gravity field solutions. Since the important role of temporal resolution of atmospheric de-aliasing models, this paper specifically investigates the influence of temporal resolution on gravity field modeling from the perspectives of frequency, spectral, and spatial domains. To this end, we introduced the gravitational acceleration and geoid height derived from the static gravity field GOCO06s in the inner integral. The introduction of the static gravity field has a comparable impact on LRI range-rate residuals as the accuracy of the LRI range-rate data, despite its magnitude of being less than 0.1mm in the spatial domain. This finding also highlights the significance of error level in existing de-aliasing products as a crucial factor that restricts the current accuracy of gravity field solutions. Further analyses show that increasing the temporal resolution from 3 hours to 1 hour has an insignificant impact on the gravity solutions in both the frequency and spectral domains, which is also smaller than that caused by using different atmospheric datasets. However, in the spatial domain, LRI range-rate residuals can be effectively mitigated in certain regions of the Southern Hemisphere at mid- and high-latitudes by increasing the temporal resolution. Particularly, the discrepancies of mass change estimates brought about by enhancing temporal resolution have distinct characteristics, especially in the Congo River and the Amazon River Basins. The mass changes in terms of EWH derived by using P4M6 filtering show that the maximum RMS value of spatial differences caused by improving the temporal resolution of the atmospheric de-aliasing models can reach ∼13.4mm in the sub-region of the Congo River Basin. However, using different atmospheric datasets can lead to a maximum difference of ∼16.5mm. For the Amazon River Basin, the corresponding maximum discrepancy is ∼18.1mm, and that caused by improving temporal resolution is ∼9.4mm. We further divide the Congo River Basin into several sub-regions using a lat-lon regular grid with a spatial resolution of 3 degrees. The subsequent time series results of mass changes reveal that the maximum contribution of temporal resolution and changes in the atmospheric datasets can reach 11.09% and 21.24%, respectively. This suggests that it is necessary to consider the temporal resolution of de-aliasing products when studying mass changes at a regional scale.