Australia: The Land Where Time Began

Antarctic Ice Sheet – Mass Balance from 1992 to 2017

According to the IMBIE Team the Antarctic Ice Sheet is an important indicator of climate change and a sea level rise driver. In this paper satellite observations of the changing volume of the Antarctic Ice Sheet, flow and gravitational attraction with modelling of its surface mass balance shows that it lost 2,720 ± 1,390 billion tonnes of ice between 1992 and 2017, which corresponds to a mean sea level increase of 7.6 ± 3.9 mm (errors are 1 standard deviation). Rates of the loss of ice from the West Antarctic Ice Sheet (WAIS) has increased from 53 ± 29 billion to 159 ± 26 billion tonnes per year as a result of ocean-driven melting over this period and ice shelf collapse has increased the rate of ice loss from the Antarctic Peninsula from 7 ± 13 billion to 33 ± 16 billion tonnes per year. It was found by the IMBIE Team that large variations in and among the estimates of models of surface mass balance and for East Antarctica, glacial isostatic adjustments with its average rate of mass gain over the period 1992-2017 (5 ± 44 billion tonnes per year) being the least certain.

Enough water is held in the Antarctic ice sheets to raise global sea level by 58 m (Fretwell et al., 2013). Ice is channelled to the oceans through a network of glaciers and ice streams (Rignot, Mouginot & Scheuchl, 2011), each of which has a substantial catchment (Zwally, Giovinetto, Beckley and Saba, 2012). Fluctuations of the mass of grounded ice sheets arise due to differences between the net accumulation of snow at the surface, meltwater runoff and the discharge of ice into the ocean. Reductions in the thickness (Shepherd, 2010) and extent (Cook & Vaughn) of floating ice shelves have disturbed the flow of inland ice flow, and this has triggered retreat (Rignot et al., 2014; Konrad et al., 2018), acceleration (Joughin et al., 2002; Rignot et al., 2004) and drawdown (Shepherd, Wingham & Mansley, 2002; Scambos et al., 2002) of many ice streams that terminate at the ocean. There have been various techniques that have been developed to measure ice sheet mass changes, based on satellite observations of their speed (Rignot & Thomas, 2003), volume (Wingham et al., 1998) and gravitational attraction (Velicogna & Wahr, 2006) combines with modelled surface mass balance (SMB) (van Wessem et al., 2018) and glacial isostatic adjustment (GIA; the ongoing vertical movement of land associated with ice load changes) (King et al., 2012). There have been more than 150 assessments of the loss of ice from Antarctica based on these approaches (Briggs et al., 016) since 1989. It was demonstrated by the inter-comparison of 12 such estimates (Shepherd et al., 2012) that the 3 principal satellite techniques provide similar results at continental scale and, lead, when combined, to an estimated loss of mass of 71± 53 billion tonnes of ice per year (Gt yr^-1) averaged over the period 1992-2011 (errors are 1 standard deviation unless otherwise stated). In this paper the assessment has been extended to include twice as many studies, which doubles the overlap period and extends the record to 2017.

Satellite Observations

According to the IMBIE Team they collated 24 estimates that were derived independently of ice sheet mass balance that had been determined in the period 1992-2017 and were based on the techniques of satellite altimetry (7 estimates), gravimetry (15 estimates) or the input-output method (2 estimates). The computation involved a total of 24, 24 and 23 estimates of mass change within defined geographical limits (Zwally, Giovinetto, Beckley & Saba, 2012; Rignot, Mouginot & Scheuchl, 2011) for the East Antarctic Ice Sheet (EAIC), WAIS and Antarctic Peninsular Ice Sheet (APIS), respectively. The rates of mass change of ice sheets were compared over common intervals of time (Shepherd et al.,2012). The rates of ice sheet mass balance was then averaged using the same class of satellite observations to produce time series of mass change that were independent of technique in each geographical region. The uncertainty in the annual mass rate was computed within each class as the mean uncertainty of the individual contributions. The final reconciled estimated ice sheet mass change was computed for each region as the mean of the values that are dependent on the technique that are available for each epoch. The IMBIE Team assumed that the errors for each technique are independent in computing the associated uncertainty. In order to estimate the cumulative mass change and its uncertainty, estimates for each ice sheet were integrated and the annual uncertainty was weighted by ¹/_√n, where n is the number of years since the start of each time series. Mass trends of the Antarctic Ice Sheet (AIS) were computed as the linear sum of the regional trends and the uncertainties in the mass trends as the root-sum-square of the regional uncertainties.

Trends in Antarctic ice sheet mass

The disagreement level between individual estimates of the mass balance of ice sheets increases with the area of each region of the ice sheet, the average standard deviations per epoch being

· APIS 11 billion tonnes of ice per year,

· WAIS 21

· EAIS 37

Gravimetric estimates are the most abundant among the techniques, and are also the most closely aligned, though in East Antarctica their spread increases, where the Glacial Isostatic Adjustment remains poorly constrained (Bentley et al., 2014) is at its least uncertain when integrated spatially (Peltier, 2004; Wahr & Zhong, 2013; Sasgen et al., 2013; Peltier, Argus & Drummond, 2015; King, Whitehouse & van der Wal, 2016; Whitehouse et al., 2012; Spada, Melini & Colleoni, 2015; Konrad, Sasgen, Pollard & Klemann, 2015; Briggs, Pollard & Tarasov, 2014; Ivins & James, 2005; Ivins et al., 2013; Nield et al., 2014), because of the vast extent of the region. Solutions that are based on satellite altimetry, and the input-output method that is run for the full record is roughly twice the duration of the gravimetry time series. Though most (59%) estimates are within 1 standard deviation of the technique-dependent mean, and a few (6%) depart by more than 3 standard deviations.

For the Antarctic Peninsula, the 25-year average rate of mass balance is -20 ± 15 billion tonnes per year, with an increase of about 15 billion tonnes per year that has been lost since 2000. In West Antarctica the strongest signal and trend has occurred, where mass loss rates increased from 53 ± 29 to 159 ± 26 billion tonnes per year over the period between the first and final 5 years of the survey by the IMBIE Team; the largest of the increases occurred during the late 2000s when there was an acceleration of the discharge of ice from the Amundsen Sea sector (Mouginot, Rignot & Scheuchl, 2014). These regional losses are both driven by reductions in the thickness and extent of floating ice shelves, and this has triggered the retreat, acceleration and drawdown of glaciers that terminate at the ocean (Shepherd, Fricker & Farrell, 2018). East Antarctica is where the least certain result was obtained, where the average 25-year mass trend is 5 ± 46 billion tonnes per year. The Antarctic Ice Sheet, overall, lost 2,720 ± 1,390 billion tonnes per year between 1992 and 2017, which is an average rate of 109 ± 56 Billion tonnes per year.

Surface mass balance

An essential component of the input-output method is knowledge of the Surface Mass Balance of the icesheet, which subtracts the discharge of solid ice from the net accumulation of snow, and aids in the interpretation of mass trends that are derived from satellite altimetry and Gravimetry. The main driver of temporal and spatial variability in the Antarctic Ice Sheet surface mass change (Boening et al., 2012; Medley et al., 2018) is snowfall. Though locally important, spatially integrated sublimation and runoff of melt water are typically 1 and 2 orders of magnitude smaller, respectively. The Surface Mass Balance of the Antarctic Ice Sheet is usually taken from the atmospheric models in the absence of maps that are observationally based, evaluated with observations that are remotely sensed and in situ (van Wessem et al., 2018; Favier et al., 2013; van de Berg & Medley, 2016; Palerme et al., 2017; Van Wessem et al., 2014). The IMBIE Team compared 2 global reanalysis products (JRA55 and ERA-Interim) and 2 regional climate models (RACMO2 and MARv3.6) in order to assess the Antarctic Surface Mass Balance. The best performing reanalysis product over Antarctica is usually regarded to be ERA-Interim, albeit with a dry bias in the interior and overestimated rain fraction (Palerme et al. 2017; Bromwich, Nicolas & Monaghan, 2011; Behrangi et al., 2016). Accumulation rates that are averaged spatially peak at the Antarctic Peninsula, and are roughly 3 times lower in West Antarctica and roughly 7 times lower in East Antarctica. Values are given by the regional climate models of 4.7% higher and the reanalysis 7% lower. According to the INBIE Team these differences can be attributed to the higher resolution of the regional models, which resolve the steep coastal precipitation gradients in greater detail and to their improved representation of the polar processes. In all products the temporal variability is similar and they all agree that there is an absence of a trend that is ice sheet wide in the Surface Mass Balance over the period 1979-2017, which implies that the recent loss of mass from the Antarctic Ice Sheet is dominated by increased discharge of solid ice into the ocean.

Glacial isostatic adjustment

The method used to correct for Glacial Isostatic Adjustment (King et al., 2012) strongly influences the gravimetric estimates (Peltier, 2004; Peltier, Argus & Drummond, 2015; Whitehouse et al., 2012; Ivins & James, 2005; Ivins et al., 2013; Shepherd, Fricker & Farrell, 2018). There were also 9 continent-wide forward-model simulations and 2 regional simulations that were assessed in this study to understand better uncertainties in the Glacial Isostatic Adjustment signal; the gravimetry estimates of mass balance were reprocessed using W12a (White et al., 2012) and IJO5_r2 (31) Glacial Isostatic Adjustment models for comparison with earlier work (Shepherd et al., 2012). Across Antarctica the net gravitational effect is positive, and the mean and standard deviation of the continent-wide Glacial Isostatic Adjustment models (54 ± 18 billion tonnes per year is very close to that of W12a (56 ± 27 billion tonnes per year) and IJO5_R2 (55± 13 billion tonnes per year) models. The difficulty of quantifying the timing and extent of past ice sheet change and the absence of lateral variations in Earth rheology within some models (van der Wal, Whitehouse & Schrama, 2015), is probably a reflection of the narrow spread. In areas where the Glacial Isostatic Adjustment is a substantial component of the regional mass change, such as the Amundsen, Ross and Filchner-Ronne sectors of West Antarctica, though it is also the area of greatest variability (e.g. standard deviation of more than 10 mm/yr in the Amundsen sector, the models predict the greatest rates of solid Earth uplifts (on average, 5-7 mm/yr). There is a low variance among the models and broad agreement with the GPS observations (Martín-Español et al., 2016), away from areas of large Glacial Isostatic Adjustment signals. Most models that are considered in this study, nevertheless, do not account for ice sheet change over the past few thousand years, as it is not well known. Gravimetric calculations (Caron et al., 2018) can also be biased by inaccurate treatment of low degree harmonics that are associated with the global Glacial Isostatic Adjustment signal. If a transient component that is associated with recent ice sheet change is included in the Glacial Isostatic Adjustment signal, it will bias the mass trend estimates and should be accounted for in future work.

Outlook

It is still possible that assessments in ice sheet mass balance can be achieved. A powerful tool for evaluating models of Surface Mass Balance and compaction of firn over large spatial (thousands of kilometres) and temporal (centennial) scales is airborne snow radar (Medley et al., 2014; Lewis et al., 2017), as well as the ice cores that have been used traditionally (Thomas et al., 2017). Glacial Isostatic Adjustment models are enabled to be scrutinised and calibrated by geological constraints of the ice sheet history (Bentley et al., 2014) and GPS measurement of contemporary uplift (Martín-Español et al., 2016; Thomas et al., 2011). The IMBIE Team suggest more of these types of datasets are needed, especially in East Antarctica. The spread of models such as Glacial Isostatic Adjustment and Surface Mass Balance should be evaluated in concert with the satellite gravimetry, altimetry and velocity measurements, given their apparent diversity. The imbalance that is present in the current record would be addressed by a reassessment of the satellite measurements that were acquired during the 1990s. Alternative techniques (e.g. Wahr, Wingham & Bentley, 2000) for combining satellite datasets should be explored, and satellite measurements with common temporal sampling should be contrasted. The mass balance of the ice sheet record should now be separated into continuations due to short term fluctuations in in Surface Mass Balance and to trends on longer terms in glacier ice. Continued satellite observations are essential, ss well as these improvements.

Sources & Further reading