Improve Predictions of Future Land-ice Loss and Impacts on Sea Level
Land ice loss—especially from northern hemisphere glaciers and the Greenland ice sheet—now exceeds thermal expansion in its contribution to rising sea level. While the loss of glacier mass has continued for the past few decades with a slight increase in recent years, the rate of mass loss from the Greenland ice sheet has dramatically increased in the past decade and continues to increase. These rapid changes are the result of increased discharge from grounded ice into the ocean and from increased ice melting, which more than outweigh increases in surface accumulation. In light of these observational facts, it is unsettling that neither quantitative prediction of future land ice loss nor credible estimation of an upper bound of future sea level are possible (IPCC, 2007). Correcting this situation requires a predictive understanding of the processes responsible for land ice loss.
Greenland contains enough ice to raise sea level an average of 6.5 meters. Glaciers and ice caps, occurring mostly in the Arctic, could contribute an additional 0.35 meters. Roughly one third of people live at or near the coast and will be directly affected by rising sea level. The direct impacts of a one-meter sea level rise include the displacement of over 100 million people, loss of nearly one trillion dollars in global gross domestic product (GDP), and the flooding of 2.2 million km2 (Anthoff et al., 2006). Effective mitigation or adaptation strategies to respond to higher future sea level require credible and accurate projections of future land ice loss.
Objectives (5-year time frame)
1. Determine the impact of ocean waters on tidewater and outlet glaciers
Observations of the spatial patterns and magnitudes of land-ice loss indicate the dominant role that ocean heat has in forcing increased ice discharge. Process studies that include circulation of the water near the ice, rapid melting of floating glacier tongues, calving at the glacier terminus, and the glacier's response (terminus position and changing elevation and velocity field) as these changes propagate inland are at an early stage.
1.1. Collect bathymetric data proximal and sub-glacially on a number (5-10) of large and/or recently responsive tidewater glaciers, including ice sheet outlet glaciers.
1.2. Develop, adapt, and deploy oceanographic instrumentation to monitor water properties in the vicinity of active tidewater glaciers. Link oceanographic measurements with simultaneous measurements of ice flow and calving.
1.3. Monitor glacier elevation and velocity to quantify the strength and extent of glacier response to oceanic changes.
1.4. Expand modeling efforts to simulate the intense interactive processes at play between ocean and ice in narrow fjords.
1.5. Through a combination of existing observations and new models, link oceanographic circulation on continental shelves (extending into the fjord environments) with oceanographic conditions in the deeper ocean and atmospheric patterns.
2. Determine the processes controlling the intra-annual and inter-annual variability of land ice discharge
It has recently been discovered that meltwater formed at the surface of an ice sheet can cause a large and sudden increase in ice flow speed. More generally, the dynamic interaction of subglacial water flow with the overlying ice leads to multiple processes that either increase ice flow through lowered basal friction, or decrease ice flow through enlarged subglacial channels that lower effective pressures. A new functional relationship between the forcing effect of surface meltwater and ice flow must be determined based on extensive field observations to properly incorporate this effect in predictive models of future ice sheet behavior. Basal resistance can vary for non-hydrologic reasons such as deformation and sedimentation of subglacial material, yet the impact on glacier flow speed is very poorly understood. For the narrower outlet glaciers, or those with extremely low basal shear stresses, the lateral resistance of slower moving ice adjacent to the glacier can dominate the glacier's flow speed.
2.1. Collect and analyze new data on surface melt fluxes; the propagation and development of surface, englacial, and subglacial hydrologic networks; and the corresponding three-dimensional ice motions over multiple seasons and years. Projections of surface melt can then be extended to estimates of changing ice discharge.
2.2. Explore the subglacial environment, for example with advanced ice-penetrating radar systems, to assess resistive stresses that modulate ice flow rates.
2.3. Monitor deformation and thermal conditions within glacier marginal areas over seasonal and interannual time scales to quantify the sensitivity of outlet glacier discharge to changes in the glacier margins.
3. Improve predictions of pan-arctic surface precipitation and methods to accurately downscale precipitation patterns to the glacier basin scale
Precipitation and melting are major components of determining the overall growth or loss of land ice. Nearly half of the ice loss experienced by the Greenland ice sheet during the past 50 years is attributed to changes in its surface mass balance (sum of all accumulation effects minus all ablative effects). Meteorological modeling of precipitation patterns over the large ice sheet are robust in the relatively broad, featureless interior (for example, with Regional Atmospheric Climate Model (RACMO), but are more difficult at the mountainous coast and very poor in the alpine regions occupied by much of the remaining arctic land ice. Global circulation models provide the best predictions of future precipitation magnitude and distribution but lack spatial detail. Downscaling general circulation model/global climate model (GCM) output, or even output from regional climate models, such as RACMO, to account for the influences of local orography is poorly developed and must be improved. This is a relatively specialized area of climate modeling, but requires attention before it becomes the limiting uncertainty in projections of future land ice loss.
3.1. Meteorological modeling of alpine environments should be expanded to include the various components of glacier accumulation and ablation.
3.2. Test areas where densely sampled data in both space and time exist should be incorporated into meteorological models at both the local scale (0.1–1 km) and meso-scale (10–100km) to investigate accurate downscaling strategies.
4. Quantify the regional pattern of relative sea-‐level change driven by the predicted pattern of land ice loss
The magnitude of future sea level is usually stated as a globally averaged value. Regional changes in sea level can vary up to many tens of percent from the global mean, depending on how additional water from lost land ice and thermal expansion of the upper mixed layer of the ocean are distributed by ocean currents and the changes in the gravity field resulting from changes in mass redistribution (Mitrovica et al., 2001). Because much of the expected land ice loss is sourced in various locations distributed across the Arctic, the variability of sea level change is expected to be particularly large in the Arctic.
4.1. Global gravity models should be employed to explore the possible patterns of isostatic (i.e., local) sea level change. Recent observations of ice mass loss can be extrapolated to future decades. Along with predictions of possible future land ice loss, these future patterns provide a rich sample space within which ranges of possible and likely sea level change across all arctic coastlines.
4.2. The ever-increasing set of observations of local sea level can be compared with predictions based on the observed pattern of land ice loss to improve the veracity of gravity models.
Science Steering Committee Contact: Robert Bindschadler (robert [dot] a [dot] bindschadler [at] nasa [dot] gov)
Version: October 2012
U.S. CLIVAR International Workshop, June 2013: Understanding the Response of Greenland's Marine Terminating Glaciers to Oceanic and Atmospheric Forcing. A draft report summarizing the workshop is available here.