GPS and Relative Sea-level Constraints on Glacial Isostatic Adjustment in North America

Abstract:

Recently, new GIA models have been developed for the Innuitian Ice Sheet and for the north-central portion of the Laurentide Ice Sheet (Simon, 2014; Simon et al., 2015). This new combined model, herein called Innu-Laur15, was developed from the ICE-5G model and load adjustments were made to improve the fit to relative sea-level observations and to GPS-constrained vertical crustal motion in the Canadian Arctic Archipelago and around Hudson Bay. Here, the predictions of Innu-Laur15 are compared to observations and other GIA models over an extended region comprising much of North America east of the Rocky Mountains. GIA predictions are made using compressible Maxwell Earth models with gravitationally self-consistent ocean loading, changing coastlines, and ocean-water inundation where marine ice retreats or floats. For this study, GPS time series are the NA12 solution (Blewitt et al., 2013) downloaded from http://geodesy.unr.edu/NGLStationPages/GlobalStationList and fit with a linear trend, annual and semi-annual terms, and offsets as indicated by station logs and by inspection of the time series. For example, a comparison of GPS observations of vertical crustal motion from the NA12 solution at 360 sites gives root-mean-square (RMS) residuals of 3.2 mm/yr (null hypothesis), 1.8 mm/yr (Innu-Laur15), and 2.9 mm/yr (ICE-5G) for the VM5a Earth model. Preliminary comparisons with other Earth models give similar patterns where Innu-Laur15 provides a better fit than ICE-5G. Further adjustments to the Innu-Laur15 ice sheet history could improve the fit to GPS rates in other regions of North America.