Surface meltwater impacts Greenland Ice Sheet mass balance indirectly by reducing albedo and promoting hydrofracture. However, fully understanding both processes requires accurate mapping of small-scale features such as ponds, channels, and moulins that govern meltwater formation and drainage. Here we investigate surface water dynamics at high spatial (∼1 m) and temporal resolution by applying deep learning to high-resolution imagery from the SkySat constellation. We develop a U-Net model that robustly classifies surface meltwater with higher accuracy than a conventional thresholding approach. Our mapping reveals that small water features (<0.015 km2) account for a substantial fraction of surface water area in the western Greenland Ice Sheet ablation zone, especially during May (67%) and August (38%). However, we find that seasonal variability in surface water area is primarily driven by the filling and draining of 12 large supraglacial lakes. The high spatial resolution of the SkySat imagery reveals that much of this variability can be attributed to the development of narrow supraglacial channels that facilitate the drainage of upstream lakes into a single downstream lake. When the downstream lake rapidly drains, we observe synchronous lake drainage across our study site between 13–16 June. This cascading drainage event explains how lakes drain even when they are situated in compressive ice flow regimes and provides an alternative mechanism for synchronous lake drainages typically attributed to transmission of stress perturbations. Our study demonstrates that deep learning applied to high-resolution satellite imagery can provide valuable insights into supraglacial hydrology.

Plain Language Summary
Meltwater produced by the Greenland Ice Sheet is a major contributor to sea-level rise. But, before it flows into the ocean, meltwater can further increase ice sheet mass loss by (a) reducing surface albedo, which increases solar energy absorption, and (b) weakening glaciers and ice shelves, which enhances iceberg production. Understanding both processes requires detailed maps of small water features, such as ponds, streams, and moulins that are difficult to detect with standard satellite imagery. In this study, we applied a deep learning model (U-Net) to map surface meltwater in high-resolution SkySat images. We found that small water features account for a substantial fraction of surface water area, especially in May and August. However, most seasonal variation in meltwater coverage is driven by the filling and draining of 12 large supraglacial lakes. The high-resolution SkySat images reveal that many of these lakes drain simultaneously through a previously undocumented mechanism that begins when several upstream lakes drain into a single downstream lake via narrow supraglacial channels. When the downstream lake drains rapidly, it triggers drainage of the entire connected system. This work shows that combining deep learning with high-resolution satellite images can enhance our understanding of ice sheet meltwater dynamics.
... Read more

Greenland Ice Sheet (GrIS) mass loss has accelerated in recent decades, primarily due to enhanced surface melt. Identifying the causal dependencies of surface melt remains challenging with conventional correlations. Using the (Formula presented.) causal discovery algorithm applied to CESM2 large-ensemble simulations and evaluated against two high-resolution regional climate models, we identify significant contemporaneous positive links from melt to net shortwave radiation (reflecting melt–albedo feedback) and from sensible and latent heat fluxes to melt. These results highlight shortwave radiation and turbulent heating as dominant drivers of GrIS summer melt anomalies over the ablation zone at monthly timescales. Compared with correlations, (Formula presented.) isolates fewer but more physically interpretable dependencies. By the end of the century (SSP3-7.0), these links persist but the turbulent heat-related ones become undirected, indicating reduced statistical identifiability and possible stronger instantaneous surface–atmosphere coupling in a warmer climate. ... Read more

Antarctic atmospheric rivers (ARs) are a form of extreme weather that transport heat and moisture from the Southern Hemisphere subtropics and/or mid-latitudes to the Antarctic continent. Present-day AR events generally have a positive influence on the Antarctic ice-sheet mass balance by producing heavy snowfall, yet they also cause melt of sea ice and coastal ice sheet areas, as well as ice shelf destabilization. In this Review, we explore the atmospheric dynamics and impacts of Antarctic ARs over their life cycle to better understand their net contributions to ice-sheet mass balance. ARs occur in high-amplitude pressure couplets, and those strong enough to reach the Antarctic are often formed within Rossby waves initiated by tropical convection. Antarctic ARs are rare events (~3 days per year per location) but have been responsible for 50–70% of extreme snowfall events in East Antarctica since the 1980s. However, they can also trigger extensive surface melting events, such as the final ice shelf collapse of Larsen A in 1995 and Larsen B in 2002. Climate change will likely cause stronger ARs as anthropogenic warming increases atmospheric water vapour. Future research must determine how these climate change impacts will alter the relationship among Antarctic ARs, net ice-sheet mass balance and future sea-level rise. ... Read more

Correction to: Nature Reviews Earth & Environmenthttps://doi.org/10.1038/s43017-024-00638-7, published online 11 February 2025. In the version of the article initially published, the colour bar in Fig. 3 was labelled “Annual accumulated snowfall (kg m²)” and has now been amended to “Annual accumulated snowmelt (kg m²)” in the HTML and PDF versions of the article. ... Read more

The simulation of ice sheet-climate interactions, such as surface mass balance fluxes, is sensitive to model grid resolution. Here we simulate the multi-century evolution of the Greenland Ice Sheet (GrIS) and its interaction with the climate using the Community Earth System Model version 2.2 (CESM2.2) including an interactive GrIS component (the Community Ice Sheet Model v2.1 [CISM2.1]) under an idealized warming scenario (atmospheric (Formula presented.) increases by 1% (Formula presented.) until quadrupling the pre-industrial level and then is held fixed). A variable-resolution (VR) grid with 1/ (Formula presented.) regional refinement over the broader Arctic and (Formula presented.) resolution elsewhere is applied to the atmosphere and land components, and the results are compared with conventional (Formula presented.) lat-lon grid simulations to investigate the impact of grid refinement. Compared with the (Formula presented.) runs, the VR run features a slower rate of surface melt, especially over the western and northern GrIS, where the ice surface slopes gently toward the periphery. This difference pattern originates primarily from higher snow albedo and, thus, weaker albedo feedback in the VR run. The VR grid better captures the CISM ice sheet topography by reducing elevation discrepancies between CAM and CISM and is, therefore, less reliant on the downscaling algorithm, which is known to underestimate albedo gradients. The sea level rise contribution from the GrIS in the VR run is 53 mm by year 150 and 831 mm by year 350, approximately 40% and 20% less than that of the (Formula presented.) runs, respectively. ... Read more

Surface mass balance (SMB) represents the net effect of all processes that add or remove mass from the surface of an ice sheet. For the Antarctic Ice Sheet (AIS), snowfall is the primary SMB contributor, delivering ~2300 Gt annually (van Wessem et al. 2018; Agosta et al. 2019; Mottram et al. 2021). Sublimation represents the largest negative term in the AIS SMB, given that most surface melt refreezes within the firn (Mottram et al. 2021). ... Read more