Comment on “An Assessment of the ICE‐6GC (VM5a) Glacial Isostatic Adjustment Model” by Purcell et al.

Abstract The most recently published model of the glacial isostatic adjustment process in the ICE‐NG (VMX) sequence from the University of Toronto, denoted ICE‐6GC (VM5a), was originally developed to degree and order 256 in spherical harmonics and has been shown to provide accurate fits to a voluminous database of GPS observations from North America, Eurasia, and Antarctica, to time dependent gravity data being provided by the GRACE satellites, and to radiocarbon‐dated relative sea level histories through the Holocene epoch. The authors of the Purcell et al. (2016, https: //doi. org/10. 1002/2015JB012742) paper have suggested this model to be flawed. We have produced a further version of our model, denoted ICE‐6GD (VM5a), by employing the same BEDMAP2 bathymetry for the Southern Ocean as employed in their analysis which has somewhat reduced the differences between our results. However, significant physically important differences remain, including the magnitude of present‐day vertical crustal motion in the embayments and in the spectrum of Stokes coefficients for present‐day geoid height time dependence which continues to “flatten” at high spherical harmonic degree. We explore the reasons for these differences and trace them to the use by Purcell et al. of a loading history for the embayments that differs significantly from that tabulated for both the original and modified versions of our model.