Geodetic vertical velocities affected by recent rapid changes in polar motion

Abstract

Secular motion of Earth's rotation pole results in large-scale secular deformation of Earth. Here, we investigate the magnitude of the deformation that has resulted from the rapid motion of the rotation pole to the east since ~2005. We show that geodetic (GNSS, DORIS, VLBI and SLR) estimates of vertical velocity since ~2005 have been biased by up to ±0.38 mm yr-1 relative to the longer-term deformation pattern. The largest signals occur within regions that include the U.S. Pacific Coast, Europe and South Pacific islands where geodetic measurements provide essential measurements of tide-gauge vertical movement and important constraints on models of glacial isostatic adjustment. Consequently, geodetic vertical velocities based on recent data should not be interpreted as being identical to centennial or longer term vertical land movement. Since 2010 the effect is further amplified by the overprediction of the IERS polar motion model relative to the ongoing secular change in pole position-during this time geodetic vertical velocities based on the IERS pole tide model are not just biased relative to the long-termrates but also from actual post-2010 Earth deformation. For geophysical or reference frame studies seeking geodetic vertical velocities that are representative of decadal timescales, where interannual variation is considered noise, the correction for this non-linear effect is straightforward, requiring an elastic computation using a reference rate of polar motion that is linear over the timescales of interest.