CGPS as a Tool to Measure Short-Term Mass Balance Variations of Icelandic Ice Caps

Abstract:

As the global climate changes, understanding the variations in water storage is increasingly important. In Iceland, glaciers cover approximately 11% of the land surface and comprise the country’s largest reservoir of freshwater. Seasonal variations in winter snowfall and summer melting can impact river discharge and water availability for utilities such as hydropower. A dense network of continuously operating GPS sites record rapid crustal uplift exceeding 30 mm/year and accelerations up to 1-2 mm/yr² reflecting the recent accelerating ice mass loss from the major Icelandic ice caps. Additionally, CGPS records seasonal motion: the elastic response to annual snow and melt seasons. Amplitudes of vertical seasonal motion approach 10 mm for those sites in the Central Highlands region in close proximity to the Vatnajökull ice cap. We demonstrate the utility of CGPS-measured displacements for estimating seasonal ice mass changes using a simple least squares inversion. We utilize all 3 components of motion, taking advantage of the seasonal motion recorded in the horizontal. We remove secular velocities and accelerations as well as seasonal motions due to atmospheric, hydrologic, and non-tidal ocean loading. We model GPS noise as a flicker process and set the standard deviation for that noise to be 1 mm for the horizontal components and 3 mm for the vertical. We calculate unit responses to each of the 5 largest ice caps in central Iceland at each of the 62 CGPS locations using both the RELAX and REAR codes and investigate the impact that Earth’s structure has on our modeling results, noting here that the standard PREM Earth model results in calculated displacements smaller than the observations by an order of magnitude. Our preliminary inversion results match available summer and winter mass balance measurements well. We are often able to reproduce the year-to-year variations in loading and melting within our calculated 95% confidence bounds.