Mass balance of Graham Land at present-day and over the past 150 years using GRACE and GNSS station data

Abstract:

The Antarctic Peninsula has been warming at an anomalously high rate with respect to other parts of the globe for the past 115 years. The increased atmospheric and ocean warmth has caused both increased precipitation and loss of buttressing near the trunks of the numerous outlet glaciers that drain into the ocean. The measurements by space gravimetry (Gravity Recovery and Climate Experiment or ‘GRACE’) have been crucial to identifying the dominant role of losses in butressing in a decadal-scale total ice mass budget for the Peninsula north of about 69 °S. One of the major losses of butressing came at a time when the Larsen B Ice Shelf catastrophically disintegrated into the northernmost Weddell Sea in March of 2002, the same month that GRACE was launched into orbit. The solid Earth rebounded in phase with the speed-up of the unbutressed outlet glaciers as evidenced by GNSS station data. The process has been repeating itself throughout the northern Peninsula since the late 1980’s. Larsen A Ice Shelf breakup (LAISb) occurred, for example, in the Austral summer of 1993. Rott et al. (2011) have estimated a velocity acceleration at the trunks of 11 LBISb outlet glaciers for which flux-gate mass transport can be measured. Butress loss caused an increase in discharge of 4.3 ± 1.6 Gt/a in 2008, relative to measurements that span 1995-1999 for 11 Larsen B feeding glaciers. GRACE analysis centers have now released 12 years of monthly solutions (RL05) for global mass change. Solutions for the region of the Peninsula north of 69 ° S (Graham Land) using both constrained and unconstrained mascon methods show 2003-2011 net ice loss to the oceans with a trend of greater than 32 Gt/yr (Ivins et al., 2011; Luthcke et al., 2013). A new set of mascon solutions developed by JPL (Wiese et al., 2014) provide a high-resolution look at the evolution of mass in the Antarctic Peninsula from 2003-2014. We employ a time-series from these solutions and discuss the evolution of mass changes over the past 12-150 years, the GNSS uplift rates and the GIA correction that can be retrieved from modeling. The solutions for Graham Land yield a trend of -29.8 ± 4.8 Gt/yr using a GIA correction of 6.9 ± 1.0 Gt/yr. We infer that much of this mass loss began at least 50 years ago, and we attempt to quantify this ice mass history as it drives a substantial viscoelastic flow in the mantle.