Search

In recent decades, meltwater runoff has accelerated to become the dominant mechanism for mass loss in the Greenland ice sheet1,2,3. In Greenland’s high- elevation interior, porous snow and firn accumulate; these can absorb surface meltwater and inhibit runoff4, but this buffering effect is limited if enough water refreezes near the surface to restrict percolation5,6. However, the influence of refreezing on runoff from Greenland remains largely unquantified. Here we use firn cores, radar observations and regional climate models to show that recent increases in meltwater have resulted in the formation of metres-thick, low-permeability ‘ice slabs’ that have expanded the Greenland ice sheet’s total runoff area by 26 ± 3 per cent since 2001. Although runoff from the top of ice slabs has added less than one millimetre to global sea-level rise so far, this contribution will grow substantially as ice slabs expand inland in a warming climate. Runoff over ice slabs is set to contribute 7 to 33 millimetres and 17 to 74 millimetres to global sea-level rise by 2100 under moderate- and high-emissions scenarios, respectively—approximately double the estimated runoff from Greenland’s high-elevation interior, as predicted by surface mass balance models without ice slabs. Ice slabs will have an important role in enhancing surface meltwater feedback processes, fundamentally altering the ice sheet’s present and future hydrology.

Liquid water stored on the surface of ice sheets and glaciers impacts surface mass balance, ice dynamics, and heat transport. Multispectral remote sensing can be used to detect supraglacial lakes and estimate their depth and area. In this study, we use in situ spectral and bathymetric data to assess lake depth retrieval using the recently launched Landsat 8 Operational Land Imager (OLI). We also extend our analysis to other multispectral sensors to evaluate their performance with similar methods. Digital elevation models derived from WorldView stereo imagery (pre-lake filling and post-drainage) are used to validate spectrally derived depths, combined with a lake edge determination from imagery. The optimal supraglacial lake depth retrieval is a physically based single-band model applied to two OLI bands independently (red and panchromatic) that are then averaged together. When OLI-and WorldView-derived depths are differenced, they yield a mean and standard deviation of 0.0±1.6 m. This method is then applied to OLI data for the Sermeq Kujalleq (Jakobshavn Isbræ) region of Greenland to study the spatial and intra-seasonal variability of supraglacial lakes during summer 2014. We also give coefficients for estimating supraglacial lake depth using a similar method with other multispectral sensors.

Liquid water stored on the surface of ice sheets and glaciers impacts surface mass balance, ice dynamics, and heat transport. Supraglacial lakes, therefore, play a significant role in ice sheet behavior as a surface expression of the glacial hydrological system. Multispectral remote sensing can be used to detect supraglacial lakes and estimate their depth and area. In this study, we use in situ spectral and bathymetric data to assess lake depth retrieval using the recently launched Landsat 8 Operational Land Imager (OLI). We also extend our analysis to other multispectral sensors to evaluate their performance with similar methods. Digital elevation models derived from WorldView stereo imagery (pre-lake filling and post-drainage) are used to validate spectrally derived depths, combined with a lake edge determination from imagery. The optimal supraglacial lake depth retrieval is a physically based single-band model applied to two OLI bands independently (red and panchromatic) that are then averaged together. When OLI- and WorldView-derived depths are differenced, they yield a mean and standard deviation of 0.0 ± 1.6 m. This method is then applied to OLI data for the Sermeq Kujalleq (Jakobshavn Isbrae) region of Greenland to study the spatial and intra-seasonal variability of supraglacial lakes during summer 2014. We also give coefficients for estimating supraglacial lake depth using a similar method as OLI with other multispectral sensors.