Estimation of the ice mass change in Antarctica is a critical component of sea level rise determination. Satellite gravimetry measures mass distribution changes and allows to calculate ice mass loss. However, such estimates crucially depend on the correction for glacial isostatic adjustment (GIA), the vertical movement of the solid-Earth due to ice load changes in the past. In this thesis we investigate the determination of GIA by the combination of gravimetry, altimetry, and firn densification model (FDM) data products. The investigation encompasses the impact of various processing parameters of the GIA on rate level for the four regions West- and East Antarctic Ice Sheet (WEAIS), Amundsen Sea Embayment (ASE, West Antarctica), Atlantic Sector (ATS, East Antarctica), and Wilkes Land (WIL, East Antarctica). Furthermore, an analysis of GIA is conducted at time series level to reveal the differences in combinations using different altimetry and FDM products. The consideration of GIA at rate level reveals differences through the utilization of disparate altimetry products, FDM products and time periods. Variations of up to 50 % of the total GIA signal are observed in WEAIS and ASE, 100 % in WIL and 200 % in ATS, respectively. The investigation at time series level unravels limitations of the input data through unphysical seasonal amplitudes, interannual variations, jumps, and trend shifts in the GIA combination result. Deviations from a linear trend in the time series, i.e. temporal variations of the linear trend, are analyzed using a state space filter and leading to promising results comparing to a least squares method. The mean rate with 2σ uncertainty of the GIA mass effect combination ensemble using state space is 79 ± 29 Gt/a, 45 ± 6 Gt/a, 10 ± 16 Gt/a, and 1 ± 7 Gt/a for the WEAIS, the ASE, the ATS, and the WIL, respectively. The results indicate the substantial impact of the selection of the utilized input data product, as well as the challenges associated with data-based GIA estimation in regions exhibiting a low signal-to-noise ratio. The determination of the GIA time series offers a benchmark opportunity to assess the performance of various data products, thereby highlighting their respective limitations.