Is the GIA signal in GRACE being treated correctly?

Abstract:

GRACE observes geoid anomalies due to surface mass change driven by water mass redistribution (hydrology and land ice melt) plus solid Earth processes such as GIA. A variety of GRACE products are available to users: level 2 geoid anomalies, level 3 gridded mass change fields (GIA included), water mass change fields (GIA removed). Furthermore, several GIA models are available and to provide users with options to choose a model, GIA fields are available both at level 2 (geoid trends) and level 3 (mantle mass trends). The relation used to convert GIA level 2 fields to level 3 is not the same as that used to convert the hydrology related geoid anomaly to water mass change near the surface of the Earth, but in many instances the relation meant for hydrology has been used with GIA level 2 fields, which leads to incorrect GIA level 3 fields. Interestingly, removing the correct GIA level 3 fields from GRACE level 3 gridded mass change fields results in an error, while using incorrect GIA fields does not. On the other hand, although using incorrect GIA products do not affect our water mass change estimates, it leads to erroneous GIA mantle mass change estimates, which we report alongside ice-sheet mass balance. Furthermore, it has serious implications for global inversion studies, wherein we use contemporary observations to either estimate GIA only or co-estimate several processes. In this presentation, we assess ongoing practices that lead to inconsistent treatment of GIA when using GRACE for Earth system science. We also recommend the best practice to ensure a consistent treatment. Using GIA model by Caron et. al., 2018, we find that inconsistent treatment of GIA can overestimate Antarctic ice mass loss by approx. 17.3 Gt/yr and underestimate ocean mass change by approx. 34.3 Gt/yr or 0.1 mm/yr Sea level equivalent.