Australia: The Land Where Time Began

A biography of the Australian continent 

Greenland Ice Sheet Surface Melt Amplified by Snowline migration and bare Ice Exposure

Mass loss from the Greenland ice sheet has increased recently as a result of enhanced surface melt and runoff. As surface albedo critically modulates melt, a prerequisite for forecasting accurately mass loss is an understanding the processes and feedbacks that alter albedo. Ryan et al. demonstrated the importance of Greenland’s seasonally fluctuating snowline, which reduces the albedo and enhances the melt by exposing dark bare ice, by using satellite imagery. This process drove 53% of net shortwave radiation variability in the ablation zone and amplified melting of the ice sheet by 5 times more than hydrological and biological processes that darken bare ice itself. Fluctuations of the snowline will exert an even greater control of melt due to the flatter topography of the ice sheet at higher elevations in a warmer climate. However, current climate models predict inaccurately the elevations of the snowline during high melt years, which portends an uncertainty that has not been foreseen in the contribution to global sea level rise.

At the present, the Greenland Ice Sheet is the largest single cryospheric contributor to global sea level rise, contributing 25% of the total contribution of observed rise of the global ocean (Chen et al., 2017). For the 21^st century the enhanced loss of mass from Greenland has been attributed to the increased runoff of surface meltwater (van den Broeke et al., 2009; Enderlin et al., 2014; Van den Broeke et al., 2016), of which ~93% is derived from the relatively small ablation zone (~22% of the area of the ice sheet) along the margin of the ice sheet (Box et al., 2012). During summer, as the winter snowpack melts, bare glacial ice is exposed. Bare glacial ice absorbs more than twice as much solar radiation and retains less melt water, because the bare ice is darker and less porous than snow. Therefore, bare ice produces a large proportion (~78%) of the total runoff to the ocean from Greenland (Steger, Reijmer & van den Broeke, 2017), though exposed only across a small area of the ice sheet during summer. It is, therefore, critical to capture accurately the reduced albedo and the full extent of bare ice in climate models in order to determine the current and future runoff contributions from Greenland to sea level rise (van den Broeke et al., 2017).

According to Ryan et al. it is a challenge to represent bare ice albedo and extent in climate models, as both impart positive feedbacks that are nonlinear between net shortwave radiation and surface melt over seasonal time scales (Box et al., 2012; van den Broeke et al., 2017; Tedesco et al., 2011; Tedesco et al., , 2016). Bare ice albedo is reduced by a seasonal increase in downwards shortwave radiation by melt processes that darken the surface of the ice, notably the exposure of dust layers, pooling of surface meltwater, increased content of interstitial water, and growth induced by liquid meltwater of ice algal assemblages that are pigmented that inhabit the surface of bare ice (Greuell, 2016; Wientjes et al., 2011; Stibal et al., 2017; Tedstone et al., 2017; Williamson et al., 2018; Ryan et al., 2018). It was argued that that these bare ice processes have contributed substantially to a reduction in albedo and associated increase in melt that has been observed across the ablation zone of Greenland between 2000 and 2011 (van den Broeke et al., 2017; Stibal et al., 2017; Tedstone et al., 2017), in spite of operating over a relatively small area of the ice sheet. Ryan et al. collectively term this category of physical and biological melt-albedo processes that darken bare ice the “the bare ice feedback”. The extent of bare ice through the annual migration of the summer snowline is also increased by a seasonal increase in downwards shortwave radiation. Glacial bare ice is exposed which enhances further the adsorption of shortwave radiation by the ice sheet, as sufficient energy is received at the surface to melt completely the winter snowpack that has accumulated. Ryan et al. term this melt albedo feedback the “snowline-albedo feedback.” Ryan et al. say snowlines have received surprisingly little focus in Greenland beyond the pioneering facies work in the 1960s and the 1990s (Benson, 1962; Fahnestock et al., 1993; Partington, 1998), though the importance of this process has been recognised for a long time in alpine glacier settings (Smith et al., 1997; Klein & Isacks, 1999; Rabatel et al., 2012). The importance of the snowline-albedo feedback in the amplification of melt, and its efficacy relative to the bare ice-albedo feedback has, therefore, not yet been evaluated.

In this paper Ryan et al. assess the importance of the snowline-albedo feedback and the influence it has on the meltwater production of the Greenland Ice Sheet using new remotely sensed products of a bare ice presence. They derived this product from daily Moderate resolution Imaging Spectroradiometer (MODIS) satellite imagery acquired by NASA’s Terra satellite and validated by use of Landsat 5,7 and 8 satellite imagery and in situ field observations. They used their product to map the variability of the snowline across Greenland from 2001-2017, as well as to evaluate the impact on the total net shortwave radiation relative to processes that darken bare ice and firn/snow. Then they combined their snowline dataset with the surface topography to investigate how the strength of the snowline-albedo feedback changes as snowlines rise to higher elevations under a warming climate. As a result of their heavy use for the prediction into the future of the Greenland Ice Sheet melting and contribution of runoff to global sea level rise, they assessed whether regional climate models (RCMs) accurately determine snowline elevations.

Discussion

Ryan et al. suggest that in principle, the current generation of physical RCMs should be capable of capturing the current and future strength of the snowline-albedo feedback, as they couple sophisticated multilayer snow models with realistic ice sheet topography (van den Broeke et al., 2016; Steger et al., 2017; Fettweis et al., 2017; Noël et al., 2018; Noël et al., 2015). Ryan et al. found, however, that 2 RCMs that are commonly used to forecast meltwater runoff from Greenland [MAR3.9 (Fettweis et al., 2017) and Regional Atmospheric Climate Model (RACMO) 2.3p2 (Noël et al., 2018) do not capture maximum snowline elevations and the extent of bare ice accurately. Compared to the remotely sensed bare ice presence metrics of Ryan et al. found that, on average, RACMO2.3p2 overestimates by 13% during the 2001-2017 study period. They also found that these discrepancies are correlated significantly with total summer runoff. It is suggested by this that MAR3.9 and RACMO2.3p2 do not sufficiently capture the role of the snowline-albedo feedback during extreme melt years (e.g., 2010, 2012, and 2016). The failure of RCM to predict accurately snowline elevation and the extent of bare ice during high melt years, given that bare ice exposure is a primary control on the production of meltwater and that extreme melt events are projected to increase in the future (Vizcaino et al., 2014), raises uncertainty in 21^st century forecasts of future runoff from Greenland contributions to global sea level rise. Also, uncertainty in bare ice extent adds an additional challenge for modelling the spatiotemporal variability of bare ice albedo and its impact on runoff in the future (Tedstone et al., 2017; Ryan et al., 2018).

It is indicated by the findings of Ryan et al. that future projections of ice sheet runoff by the current generation of semi-empirical models (e.g., positive degree day or temperature index) must also be treated with caution. It is assumed by these models, which are typically calibrated to RCM-modelled runoff, that the strength of melt-albedo feedbacks that were observed in the past remains constant into the future (Franco et al., 2013; Mengel et al., 2016; Mengel et al., 2018). Since the snowline-albedo feedback is not captured accurately by RCMs and there is an increase nonlinearly in exposure of bare ice as snowlines rise, it appears this assumption is not justified. Ryan et al. suggest that they should account for the increasing strength of melt-albedo feedbacks, such as is induced by snowline migration, if semi-empirical models are to be used for forecasts of the future contributions to sea level rise from Greenland.

The importance of the snowline in Greenland in the amplification of ice sheet melt was quantified for the first time by this study by Ryan et al. They found that substantial seasonal and interannual variation was exhibited by snowlines and they are the dominant control on the absorption of shortwave energy in the production of meltwater in ablation zones.

Hydrological and biological processes are secondary to the extent of bare ice exposure that is associated with fluctuations of the snowline, though they also darken bare ice and influence the absorption of shortwave energy and the generation of runoff. Snowlines will rise to higher elevations in a warmer climate and amplify melt seasonally even more than they do at present due to the hypsometry of the ice sheet, if they are not offset by enhanced accumulation of snow in winter. Rising snowlines will reduce the albedo of the ice sheet as well as the capacity of the ice sheet to retain meltwater, as bare ice has a much lower porosity than snow and firn. Also, the exposure of bare ice will increase the rate of turbulent heat transfer from the atmosphere to the surface of the ice sheet, as bare ice has substantially higher surface roughness than snow. It is therefore critical for the accuracy of projecting future runoff contribution from the Greenland Ice Sheet to global sea level rise, the representation of snowlines and bare ice exposure in climate models.

Sources & Further reading

Ryan, J. C., et al. (2019). "Greenland Ice Sheet surface melt amplified by snowline migration and bare ice exposure." Science Advances 5(3): eaav3738.