Abstract: In order to suitably supplement the degree one Stokes coefficients (C₁₀, C₁₁, C₁₁) in the Gravity Recovery and Climate Experiment (GRACE) time-variable gravity data and substitute the C₂₀ coefficient, we review two commonly applied methods based on different combination models of GRACE data and an Ocean Bottom Pressure (OBP) model, which we call Chambers model and Sun model. We compare the results of the degree one coefficients and derive global surface mass anomalies from the two methods, respectively. We also do the similar comparison for the C₂₀ coefficient and derived global surface mass anomalies from Satellite Laser Ranging (SLR) based estimates and the Sun model. There are obvious differences, particularly in the trend rates of the degree one coefficients, C₂₀ and the derived mass anomalies between the Sun model and both the Chambers model and SLR estimates. However, there are relatively small differences for annual signals. As the GRACE gravity models are released by different institutions applying different data processing and may also have different maximum degrees, the choice of GRACE data additionally affects the annual signals and especially the trend rates. Therefore, for the derivation of terrestrial water storage changes from GRACE, the results of degree one coefficients and C₂₀ from the Sun model are recommended to be used together with the best GRACE time variable gravity data with the highest maximum degree available.