Ice shelf tidal motion derived from ERS altimetry

Abstract

Ocean tides beneath floating ice shelves can affect heat and water circulations within the subshelf cavity, which may influence patterns of basal melting. Accurate tidal predictions are essential to eliminate vertical tide motions when mapping the flow speeds and thickness changes of floating ice. Global ocean tide solutions are typically accurate to 2-3 cm in amplitude. However, their coarse resolution limits their utility in coastal regions. Moreover, the absence of satellite and extensive gauge station data in the polar regions has resulted in significant differences between recent tide model solutions around the Antarctic continent. We use 9 years of high-resolution European Remote Sensing (ERS) satellite radar altimeter range measurements and a modified form of the orthotide method to generate tidal solutions for a series of ice shelves around the Antarctic Peninsula. We compare our tidal solution at George VI Ice Shelf on the western coast of the Antarctic Peninsula to direct measurements from British Antarctic Survey tide gauge records and contrast the results with solutions from the FES99 and Kantha2.0 (global) and the CATS01.02 (Antarctic) ocean tide models. The average root mean square vector difference between the tide gauge and tide model predictions of amplitude and phase was 10.9, 17.1, 4.2, and 15.4 cm for the ERS ice shelf, Kantha2.0, CATS01.02, and FES99 models, respectively. Tidal predictions derived from the ERS ice shelf model are closer to the tide gauge recordings than those of the global models, and the inclusion of ERS altimeter data may lead to improved global model performance.