Frequency-Dependent Scaling Factors to Estimate Multiple Time-Period Global Mass Changes Observed by GRACE/GRACE-FO

The scaling factor method is commonly used to restore near-true time-variable gravity from gravity recovery and climate experiment/follow-on (GRACE/GRACE-FO). This study presents a novel method for computing frequency-dependent scaling factors (FDSFs) using spherical harmonic decomposition of GRACE/GRACE-FO Level-2 data at full frequency. Two applications based on a global hydrological model (land excluding Greenland and Antarctica) and a combined model (Greenland) demonstrate that FDSFs reduce the theoretical recovery residuals versus using a single scaling factor by ∼11.0% across short-term, seasonal, inter-annual, and long-term scales. Given the ability to capture more model details, the FDSFs improved the estimates at the basin scale and in glacier regions such as High Mountain Asia. In Greenland, the FDSF-scaled results revealed an enhanced amplitude with an averaged relative increase of 38% and improved resolution to 0.5 degrees below 1500 m, compared with the GRACE Level-3 mascon solution. Our results also imply that using FDSFs would cause uncertainties, particularly in scaled short-term mass change, which could be attributed to the large discrepancy between the hydrological model and GRACE/GRACE-FO. Our study provides insights into estimating mass changes using a downscaled GRACE/GRACE-FO solution and suggests that users select FDSFs for regions of interest based on a reliable model.