Australia: The Land Where Time Began
Antarctic Glacier Grounding Lines Net Retreat
Changes in the position of grounding lines reflect imbalance with the surrounding ocean and affect the flow of inland ice, which makes grounding lines a key indicator of instability in an ice sheet. The grounding lines of several Antarctic glaciers have rapidly retreated as a result of melting that is ocean-driven, though the records are too scarce to assess the scale of the imbalance. In this paper Konrad et al. have combined satellite altimeter observations of changes in the elevation of ice and measurements of ice geometry to track the movement of grounding lines around the entire continent, which tripled the coverage of previous surveys. Between 2010 and 2016, 22%, 3% and 10% of the grounding lines that were surveyed in West Antarctica, East Antarctica and the Arctic Peninsula, respectively, showed that the grounding lines had retreated at more rapid rates than 25 m/yr (the typical pace since the Last Glacial Maximum (LGM) and there was a loss of 1,464 km2 ± 791 km2 of grounded ice from the continent. It was found that the Pine Island Glacier grounding line has stabilised, probably as a consequence of abated ocean forcing, though by far the fastest rates of retreat was in the Amundsen Sea Sector. Fast-flowing ice streams of Antarctica retreated on average by 110 m per metre of ice thinning.
Marine based ice sheets become buoyant enough to detach from the sea floor and float at grounding lines where they are joined to a land-based glacier (Weertman, 1974; Schoof, 2007). When quantifying ice discharge (Rignot et al., 2008) knowledge of this position is critical, as a boundary condition for numerical models of the flow of ice (Docquier, Perichon & Pattyn, 2011), as well as for an indicator of the state of the ice sheet during periods of advance or retreat (Scambos, Bohlander, Shuman & Skvarca, 2004; Shepherd et al., 2010; Rignot, Jacobs, Mouginot & Scheuchl, 2013). At several Antarctic ice streams the pace of retreat has been much higher during the era of satellites (Rignot, 1998; Park et al., 2013; Rignot et al., 2014; Christie et al. 2016; Li et al., 2015; Scheuchl et al., 2016), and it has been indicated by this rapid retreat it may be followed by collapse on a centennial scale of the inland catchment areas (Joughin, Smith & Medley, 2014; DeConto & Pollard, 2016), though grounding lines have retreated since the LGM (Bentley et al., 2014).
There are 3 general approaches that have been relied on to track the position of grounding lines by the use of satellite observations:
1) Identifying mismatch between surface elevation and freeboard determined by buoyancy calculations (Fricker et al., 2002);
2) Surface slope breaks associated with the transition between grounded ice to floating ice (Horgan & Anandakrishnan, 2006) and;
3) The contrast between the vertical motion of floating ice and grounded ice due ocean tides (Goldstein, Engelhardt, Kamb & Frolich, 1993).
The last method is by far the most accurate, as it relies on the mapping the hinge line, which is the limit of tidal flexure at the surface of the ice, which is more easily detectable than the grounding line itself (Gray et al., 2002). The satellite observations have not often been acquired, so that estimates exist at only few locations (Rignot et al., 2014; Christie et al.,2016; Li et al., 2015), though it is usual to quantify the migration of grounding lines by repeating the above techniques over time (Rignot et al., 2014). In this paper Konrad et al. extended a method (Park et al., 2013; Christie et al., 2016; Shepherd, Wingham & Mansley, 2002) for detecting motion of grounding lines from satellite measurements of changes in surface elevation and airborne surveys of the geometry of ice sheets to produce assessment of Antarctic grounding line migration on a continental scale.
Horizontal migration of the grounding line as the area, in which the ice is buoyant, grows or shrinks, results from the mass of the firn and ice column around the grounding line. Konrad et al. converted surface elevation rates, ∂S/∂t, that was obtained from observation by cryostat-2 (McMillen et al., 2014) into rates of migration of the grounding line, νGL, at known locations of grounding lines:
νGL = (-α + (ρo/ρi - 1] β)-1] ρm/ρi ∂S/ αt
The term in the square brackets, which was referred to as propensity for retreat, takes slopes of the surface (α, also from Cryosat-2) and the bedrock (Fretwell et al., 2013) topographies (β) in the direction of the migration of the grounding line as well as the contrast between the ocean (ρo) and densities of the ice (ρi) into account. The material density ρm allows changes in thickness to occur at densities between snow and ice. The direction of grounding line migration is defined empirically based on grounding line perpendiculars, direction of flow and inclinations of the bedrock. Konrad et al. restricted their solution to sections where the location of the grounding line had been derived from satellite Interferometric Synthetic Aperture Radar (InSAR) (Rignot, Mouginot & Scheuchl, 2011) and where the propensity for retreat was not excessively high. Uncertainties in the assumptions of density, surface elevation and rate of elevation, and topography of the bedrock were considered in order to estimate the overall error. According to Konrad et al. quantification of the migration of the grounding line along 33.4% of the 47,000 km grounding line of Antarctica was achieved, which included 612 glacier basins (Rignot et al., 2008), which is 3 times the combined length and 4 times as many glaciers as had been mapped by earlier surveys (Park et al., 2013; Rignot et al., 2014; Christie et al., 2016; Li et al., 2015; Horgan & Anandakrishnan, 2006).
It was estimated by Konrad et al. that, between 2010 and 2016, 10.7% of the grounding line of Antarctica had retreated and 1.9% advanced faster than 25 m/yr, the typical rate of glacier retreat during the retreat of the last deglaciation (Smith et al., 2014; Pollard, 2016). There were notable regional differences: while the Peninsula matched this overall picture quite well (9.5% retreat and 3.5% advance) in the West Antarctic Ice Sheet there was ice retreat of 21.7%, of which 59.4% was in the Amundsen Sea Sector, and advance of only 0.7%, whereas in the East Antarctic Ice Sheet the grounding line retreated by 3.3% and advanced by 2.2% along its length. These changes amounted to a net loss of 209 km2 ± 113 km2 of grounded ice area per year over the CryoSat-2 period in the sections that were surveyed, the major part of which took place along the West Antarctic Ice Sheet (WAIS), an area of 177 km2 ± 48 km2).
It was found that coinciding with sectors in which ice streams are known to be thinning large scale patterns of retreat of grounding lines, e.g., in West Antarctica at the Amundsen Sea and the Bellingshausen Sea coasts. The local ice sheet geometry (propensity) modulates this general link, which introduces higher spatial variability into the pattern of retreat. Along swathes of the English Coast and Wilkes Land bedrock slopes make these sectors unfavourable for either retreat or advance, In spite of the relatively large changes in thickness of the ice sheet that have occurred. In Dronning Maud Land, East Antarctica, where mass gains have been associated with increased accumulation of snow along sections that were in a state of advance (Lenaerts, et al., 2013; Boening et al., 2012).
Konrad et al. examined changes within 61 drainage basins (Rignot et al., 2008) in order to investigate regional patterns of migration of grounding lines. Between 500 m/yr and 1,200 m/yr highly localised extremes of retreat of grounding lines that was very rapid have occurred at Fleming, Thwaites, Haynes, Pope, Smith, Kohler, glaciers and Hull glaciers, at Ferrigno Ice Stream and glaciers feeding the Getz Ice Shelf, all of which are draining into the Amundsen Sea and the Bellingshausen Sea, where broad sections of the coast have retreated at rates of 300 m/yr and 100 m/yr, respectively. Frost Glacier and Totten Glacier, East Glacier have retreated, locally, up to 200 m/yr, whereas the Mercer and Dibble ice streams and glaciers of the Budd Coast have advanced locally at up to 60 m/yr to 230 m/yr.
Generally, rapidly migrating grounding lines occur in areas of fast ice flow, therefore they also computed average migration rates at places where the speed of the ice exceeds 25 m/yr and 800 m/yr. In total, there are 10 fast flowing ice streams where the central portions have retreated at rates of up to more than 50 m/yr on average, including those of the Amundsen Sea, Totten Glacier as well as several in the Bellingshausen Sea, areas where it is known that change is at least partially driven by warm ocean water (Jenkins et al., 2010; Rintoul et al., 2016; Holland, Jenkins & Holland, 2010; Hogg et al., 2017). The Lidke Glacier, the Berg Ice Stream and glaciers flowing into Venable and Abbot Ice shelves have also retreated; though at slower average rates of up to 50 m/yr. Migrations are mostly centred around zero elsewhere in East Antarctica, with the exception of the Frost, Denman and Recovery glaciers, which have retreated at between 19 and 45 m/yr, and the Mertz, Budd and Shirase glaciers and the Slessor Ice Stream, in which the average rate of retreat was between 14 and 48 m/yr.
Widespread retreat of the grounding lines has been recorded in the Amundsen Sea Embayment by the use of satellite InSAR (Park et al., 2013; Rignot et al., 2014), which contrasts with the altimetry-based results of Konrad et al. The average rate of retreat of the grounding line at the Thwaites Glacier has increased from 340 ± 240 m/yr between 1996 and 2011 (Rignot et al., 2014) to 420 ± 240 m/yr according to the method used by Konrad et al. At Pine Island Glacier retreat appears to have stagnated at 40 m/yr ± 30 m/yr during the period of CryoSat-2, after it had been migrating inland at about 1,000 m/yr between 1992 and 2011 as has been documented by previous studies (Park et al., 2013). According to Konrad et al. the stagnation coincides with a deceleration of thinning from 5 m/yr at around 2009 to less than 1 m/yr across a 20 km section inland of the grounding line in 2011 (Scheuchl et al., 2016), which in principle explains the reduced rate of retreat. The slowdown in surface lowering could, however, also result from further ungrounding, which is why Konrad et al. first examine this possibility. The grounding line would have had to retreat by at least 15 km since 2011, which is more than double that of the 2 previous decades (Park et al., 2013; Rignot et al., 2014), at a time when thinning has abated across the lower reaches of the glacier, in order to maintain contact with the upstream parts of the ⁓120 m long central trunk, which in the data of Konrad et al. are thinning at a maximum rate of 2 m/yr. This led to the conclusion that the grounding line of the main trunk has stabilised, which could potentially be due to the absence of warm subshelf water (Dutrieux et al., 2014) that drove retreat prior to 2011.
A continuation of retreat was also observed at other ice streams that were less frequently sampled. In the results of Konrad et al. high local rates of retreat of ⁓1,200 m/yr, e.g., on the Haynes, Smith and Kohler glaciers are comparable to peak rates of 1,800 to 2,000 m/yr that were detected by InSAR between 1992 and 2014 (Rignot et al., 2014; Scheuchl et al., 2016). Slower rates of retreat recorded over the last 40 years (Christie et al., 2016) in the Bellingshausen Sea are similar to those that were derived by Konrad et al.: at Ferrigno Ice Stream, retreat rates remain in the range of 50 to 200 m/yr; at Lidke Glacier, Berg Ice Stream and venable and Abbot Ice Shelves, the rates of retreat are in the range of 10 to 40 m/yr, which accords with the multidecadal range of 10 to 90 m/yr (Christie et al., 2016); and at the Cosgrove Ice Shelf no significant retreat was detected, which agrees with rates that were observed previously between -40 m/yr and +11 m/yr (Christie et al., 2016). The Totten Glacier in East Antarctica is the only location where retreat of the grounding line has been documented, and the results of Konrad et al. of 154 m/yr ± 24 m/yr in its fast flowing section is consistent with the maximum rate that was recorded between 1996 and 2013 (Christie et al., 2016).
Konrad et al. compared the rates of change in ice thickness and migration of the grounding line to assess the degree to which the processes are related. Within sections of the ice stream that were flowing faster than 800 m/yr, changes in thickness and migration of the grounding line were approximately proportional with 110 ± 6 m of retreat that was occurring with each metre by which the ice thinned. The remarkably consistent propensity for retreat that is geometry-driven at these ice streams is highlighted by this comparison, which leads to an intimate relation between thinning and retreat, in spite of the very different processes that drive them (Boening et al., 2012; Jenkins et al., 2010; Joughin et al., 2005). Konrad et al. suggest a possible reason for the stability of this relationship is that the geometry at the margins of fast moving ice sheets may be comparable due to the processes that are involved: The non-linear viscous flow of ice and the conditions of sliding at the bedrock (40) forms the shape of the surface topography. The topography of the bedrock is, in turn, produced by tectonic evolution and pre-glacial erosion, and is interactively shaped by the ice-dynamical environment by sediment erosion through overriding and subglacial hydrology (Smith et al., 2007) and by glacial isostatic adjustment of the solid Earth to the overlying ice mass (van den Berg, 2008). Though these processes occur on different spatial and temporal scales depending on many parameters, it appears that they are average propensity for retreat, nevertheless, can be approximated for different geological settings – a convenient proxy relationship that that may be used as a benchmark in investigations that are not able to rely on detailed glacial geometry or dynamics.
Konrad et al. have compiled a comprehensive record of grounding line migration rates around Antarctica which spans ⅓ of the margin of Antarctica. Their results complement and extend earlier assessments of the retreat of grounding lines (Park et al., 2013; Rignot et al., 2015; Christie et al., 2016; Li et al., 2015), in the Amundsen Sea and the Bellingshausen Sea sectors of West Antarctica and the Totten Glacier, as well as elsewhere providing the first observation of migration in key sectors, such as the Getz Ice Shelf and large portions of East Antarctica and the Antarctic Peninsula. It was estimated that 3.3%, 21.7% and 9.5% of East Antarctica, West Antarctica and the Antarctic Peninsula, respectively, are measurably in a state of retreat, though most of the grounding line is stable. The largest rates of retreat of grounding line, by far, >50 m/yr, occur at ice streams flowing into the Amundsen Sea and the Bellingshausen Sea, which, on average, are currently retreating at rates of 134 m/yr and 57 m/yr, as well as the Totten Glacier, all of which are being affected by glacial change that is driven by warm ocean water (Rintoul et al., 2016; Holland, Jenkins & Holland, 2010; Jacobs, Jenkins, Giulivi & Dutrieux, 2011), which indicates that as a driver the ocean is currently generating the fastest retreat. In Wilkes Land the alternation between retreat and advance of glaciers could be explained by regional drivers of migration being suppressed by local ones in places. At Antarctic outlet glaciers there is a robust relationship between ice thickness changes and migration of grounding lines, which indicates that the geometrical propensity for retreat is relatively uniform in areas of fast flow. Konrad et al. suggest that a more detailed map of the grounding line position, acquired ideally in the same period as the satellite altimetry observations, could substantially increase the extent of the record. According to Konrad et al. their method is, overall, a novel and potent approach to detecting and monitoring the imbalance in ice sheets in Antarctica; Locations meriting more detailed analysis through field campaigns and dedicated InSAR surveys, e.g., where fast migration occurs or a high geometric propensity for retreat prevails can be pinpointed by its use.
|Author: M.H.Monroe Email: firstname.lastname@example.org Sources & Further reading|