Item

Abstract

The Gravity Recovery and Climate Experiment/Follow-On (GRACE/GRACE-FO) mission provide an important way of estimating Greenland icesheet (GrIS) mass change. However, filtering and coarse resolution can introduce leakage errors that impede the analysis at higher resolutions. The commonly used scaling factors are computed in terms of the total mass change, which can lead to bias due to the complex and spatially variable mass change signals in different regions of the globe that comprise multiple frequency signals, e.g., in Greenland. This paper developed a method to calculate the frequency-dependent scaling factors (FDSFs) for estimating Greenland ice mass change based on GRACE observations. The FDSFs are computed based on frequency components of total signals from combination of regional atmospheric model and ice discharge data. Our findings suggest that FDSFs for inter-annual and long-term mass change exhibit substantial differences compared to the traditional non-FDSF, with a ratio of 23% and 39%, respectively. This discrepancy is particularly pronounced in the north and southwest Greenland, with an average ratio of 28%. Moreover, in terms of the root mean square errors (RMSEs), the FDSF-based result shows significant improvement compared to the non-FDSF-based and mascon solutions, particularly for the long-term mass change. Specifically, we obtain an average RMSE of 37.2 Gt and 93.4 Gt for the FDSF-based and non-FDSF-based/mascon comparisons, respectively, across the six basins studied. This study emphasizes the importance of including frequency-dependent scaling factors to obtain accurate estimates of mass changes in different regions of the world, particularly in areas where significant frequency discrepancies are present.