An Intercomparison of Grounding Line Location from Differential Interferometry and CryoSat-2
Hogg, Anna; Shepherd, Andrew; McMillan, Malcolm
University of Leeds, UNITED KINGDOM

Ice sheet Grounding Lines (GL's) are defined as the boundary between floating ice in hydrostatic equilibrium with the ocean, and grounded glacial ice. Over short, sub-daily timescales the GL is a transitory feature with a location that fluctuates due to the effect of ocean tides (Rignot, 1998). Over longer timescales permanent GL migration has been observed and this provides a key indicator of change in mass balance and internal instability in marine terminating ice masses. Changes in GL location have been observed on a sub-decadal timescale as exemplified by the rapid retreat of Pine Island Glacier (PIG) GL between 1992 and 2011 (Park et al, in review). Current predictions of the ice sheet contribution to sea level rise over the next 100 years are principally limited by a lack of understanding about the ice sheet responses to GL retreat, in conjunction with the effect of additional ice ocean interactions (IPCC, 2010). As a result of this it is important to regularly measure the change in GL location to determine the mass loss of the most rapidly changing regions of the ice sheets and thus their contribution to global sea level rise.

This study utilises two independent Earth Observation techniques to locate ice sheet GL's namely, the effect of tidal motion in Differential Interferometric Synthetic Aperture Radar (DInSAR) (Goldstein et al, 1993, Rignot, 1998) and repeat track surface elevation change in CryoSat-2 (CS-2) SARIn mode altimeter data. An intercomparison is presented between the GL location identified using temporally coincident SAR and CS-2 data over Petermann Glacier in NW Greenland. DInSAR is used to determine the GL location with high accuracy and fine spatial resolution. However, the temporal sampling frequency of the technique is limited by the availability of SAR image triplets with a short temporal baseline. A minimum of three SAR images acquired on the order of a few days apart are required in order to maintain sufficient coherence for application of the DInSAR GL determination technique. Consequently, data well suited to this task have largely only been acquired during the three European Remote sensing Satellite (ERS) 3-day repeat Ice Phase's and 1-day repeat Tandem campaigns in 1991 - 1992, 1993 - 1994, 2011 and 1995 - 1996 respectively. Assessment of the change in GL location at an annual temporal resolution is therefore not possible using the DInSAR technique with the current SAR data archive.

In this study we also develop a new method for GL determination using CS-2 data based upon the Fricker and Padman (2006) laser altimetry technique. Fricker and Padman (2006) demonstrated that satellite altimetry can be used to determine GL location using data acquired by the Geoscience Laser Altimeter System (GLAS) instrument on board the Ice, Cloud and land Elevation Satellite (ICESat). Adaptation of the altimetry GL location technique to suit CS-2 SARIn mode data enables the exploitation of this relatively new dataset to investigate recent changes in GL location. Furthermore, the new method allows an independent reassessment of change in GL location from 2010 to the present day at a finer temporal resolution than the DInSAR technique, and at a finer spatial resolution than alternative radar altimeter sensors (i.e. ERS or ENVISAT). The 369 day repeat orbit of CS-2 provides a ground track separation of approximately 2-3 km, although these tracks are not evenly spaced because in SARIn mode the satellite tracks targets that can wander either side of the reference orbit. Overall this new technique extends in time the record of GL locations determined by altimetry and provides new information on the present day rate and timing of GL migration. In the future this can be compared with auxiliary data to facilitate a new understanding of the physical variables causing change in GL location.