Crustal uplift due to ice mass variability on Upernavik Isstrøm, west Greenland

Abstract

We estimate the mass loss rate of Upernavik Isstrøm (UI) using surface elevation changes between a SPOT 5 Digital Elevation Model (DEM) from 2008 and NASA's Airborne Topographic Mapper (ATM) data from 2010. To assess the validity of our mass loss estimate, we analyze GPS data between 2007 and 2011 from two continuous receivers, UPVK and SRMP which are established on bedrock and located $\sim 65$ and $\sim 2\mathrm{km}$ from the front of UI, respectively. We construct along-track elevation changes on UI for several time intervals during 2005–2011, based on ATM, SPOT 5 and Ice, Cloud, and land Elevation Satellite (ICESat) data to assess temporal changes of UI. We estimate a mass loss rate of −6.7±4.2 Gt/yr, over an area of $\sim 1600{\mathrm{km}}^2$. The ice mass loss occurs primarily over the northern glacier of UI. This pattern is also observed $\sim 40\mathrm{km}$ upstream, where we observe glacier thinning at a rate of −1.6±0.3 m/yr across the northern portion of UI and −0.5±0.1 m/yr across the southern portion. GPS measurements suggest bedrock uplift rates of 7.6±0.6 mm/yr (UPVK) and 16.2±0.6 mm/yr (SRMP). The modeled ice mass loss of UI causes bedrock uplift rates of 1.3±0.6 mm/yr (UPVK) and 8.3±4.2 mm/yr (SRMP). Including additional contributions from ice mass changes outside UI and from Glacial Isostatic Adjustment (GIA), we obtain total modeled uplift rates of 4.7±0.6 mm/yr (UPVK) and 13.8±4.2 mm/yr (SRMP). The modeled uplift rates from our UI ice mass loss are substantially lower, indicating that additional mass loss is taking place outside of UI. We obtain a difference of 0.6 mm/yr between the modeled and observed relative uplift rates (SRMP relative to UPVK), suggesting that the mass loss of UI is well captured in the model. We observe elevation changes from −15 to −40 m/yr near the front during the period 2005–2011, indicating that UI undergoes large variations in thinning pattern over short time spans.

Introduction

Over the last decade many outlet glaciers in Greenland have experienced an increase in flow rate (Rignot and Kanagaratnam, 2006, Joughin et al., 2010), initially causing frontal thinning, and later followed by increased thinning inland (Howat et al., 2007, Pritchard et al., 2009). Early in the last decade the Greenland mass loss was most pronounced in the southeast coast of Greenland, with Helheim and Kangerdlugssuaq as the largest individual contributing glaciers. This increase in mass loss later ($\sim 2005$) shifted to the northwest and northern coasts of Greenland (Pritchard et al., 2009, Khan et al., 2010a, Schrama et al., 2011). Frontal and inland thinning of the many small glaciers along the northwest coast contribute to $\sim 20\%$ of the annual mass loss rate (2003–2008) of the Greenland ice sheet (Van Den Broeke et al., 2009), making it an important region to understand.

Upernavik Isstrøm (UI), located at the northwest coast of Greenland, is a system of four glaciers (Fig. 1). UI has retreated by up to 20 km since the end of the Little Ice Age (LIA) (Weidick, 1958, Kollmeyer, 1980), with the maximum retreat between 1931 and 1946. Here, we present a mass loss estimate for UI, based on elevation changes between a SPOT 5 DEM from 3 June 2008 and on NASA's ATM altimetry data from May 2010 (Krabill, 2011). To assess our UI mass loss estimate, we analyse GPS data during the period July 2007 to April 2011 from two permanent sites SRMP and UPVK (Fig. 1). We compare observed and modeled rates of vertical crustal displacements, using previously published methodology (Khan et al., 2007, Khan et al., 2010b). Modeled rates consist of three components: elastic uplift caused by UI ice mass changes, elastic uplift caused by ice mass changes outside UI, and GIA, due to changes in the ice load since the last glacial maximum. Here, we use the estimated gridded mass change values to model uplift rates caused by UI. The contribution from ice mass changes outside of UI is based on a mass change grid derived from ICESat data (Zwally et al., 2010) during 2006–2009. The GIA contribution is based on ICE-5G(VM2) (Peltier, 2004). To analyze thinning upstream we use time series of ice elevations obtained from a combination of ICESat, ATM, and GPS measurements during 2003–2010. The time series are created at three GPS stations located on ice $\sim$ 40 km from the glacier front (marked as A–C in Fig. 1). We quantify changes on UI from 2005–2011, by estimating along-track elevation changes based on SPOT 5, ATM and ICESat data.