Combination of GRACE, gravity and GPS data determination of long-term geoid changes in North America

Abstract

Two independent solutions for the rate of change of geoid in North America, assumed to be due to PGR, are presented. The first solution is based on principal component analysis of a GRACE time series of geoid change. The second solution is derived via least-squares collocation using absolute gravity rates and GPS vertical velocities available in Canada. The GRACE geoid rate shows good agreement with postglacial rebound models based on the ICE5G de-glaciation history, in contrast with the geoid rate derived from gravity and GPS data, which agrees better with the ICE3G and ICE4G models. In view of the still short time span of the GRACE time series, and acknowledging the fact that the errors of the geopotential coefficients have not been taken into account, it is found that the GRACE geoid rate should be down-weighted when the two geoid rates are combined.