Distribution and seasonal evolution of supraglacial lakes on Shackleton Ice Shelf, East Antarctica

Abstract

Abstract. Supraglacial lakes (SGLs) enhance surface melting and can flex and fracture ice shelves when they grow and subsequently drain, potentially leading to ice shelf disintegration. However, the seasonal evolution of SGLs and their influence on ice shelf stability in East Antarctica remains poorly understood, despite some potentially vulnerable ice shelves having high densities of SGLs. Using optical satellite imagery, air temperature data from climate reanalysis products and surface melt predicted by a regional climate model, we present the first long-term record (2000–2020) of seasonal SGL evolution on Shackleton Ice Shelf, which is Antarctica’s northernmost remaining ice shelf and buttresses Denman Glacier, a major outlet of the East Antarctic Ice Sheet. In a typical melt season, we find hundreds of SGLs with a mean area of 0.02 km2, a mean depth of 0.96 m, and a mean total meltwater volume of 7.45 x 106 m3. At their most extensive, SGLs cover a cumulative area of 50.7 km2 and are clustered near to the grounding line, where densities approach 0.27 km2 per km2. Here, SGL development is linked to an albedo-lowering feedback associated with katabatic winds, together with the presence of blue ice and exposed rock. Although below average seasonal (December-January-February, DJF) temperatures are associated with below average peaks in total SGL area and volume, warmer seasonal temperatures do not necessarily result in higher SGL areas and volumes. Rather, peaks in total SGL area and volume show a much closer correspondence with short-lived high magnitude snowmelt events. We therefore suggest seasonal lake evolution on this ice shelf is instead more sensitive to snowmelt intensity associated with katabatic wind-driven melting. Our analysis provides important constraints on the boundary conditions of supraglacial hydrology models and numerical simulations of ice shelf stability.

Supraglacial lake bathymetry automatically derived from ICESat-2 constraining lake depth estimates from multi-source satellite imagery

“…While this initial study focused on lakes in grounded ice in Greenland, Watta can potentially be applied to Antarctic melt lakes as well. While mass loss in Antarctica over the next 100 years is generally thought to be dominated by the basal melt under ice shelves (Schlegel, 2018), emerging research has focused on the potential importance of surface hydrology over Antarctica (Arthur, 2020 high elevations (Stokes, 2019) and are likely to become more prevalent on firn-depleted ice shelves in future warming scenarios, which could potentially trigger their collapse and consequently lead to accelerated sea level rise (Lai, 2020).…”

“…Past remote-sensing work has derived lake volumes from high-resolution (~1m) Worldview imagery using a physical optical depth approach as well as an empirical method using in-situ estimates (Moussavi, 2016;Pope, 2016). Both the physically-based and empirically-based methods are limited to supraglacial lakes which contain minimal particulate matter (Arthur et al, 2020), and by the depth of the lake, assuming that the reflection depletion in imagery is limited at great depths, implying a physical limit to the ability to calculate depth (Box and Ski, 2007). Additional work has applied a similar physically-based approach using Sentinel-2 from Copernicus as well as a combination of LandSat and Sentinel-2 imagery (Moussavi, 2020).…”

The triggers of the disaggregation of Voyeykov Ice Shelf (2007), Wilkes Land, East Antarctica, and its subsequent evolution

“…The Normalized Difference Water Index (NDWI ice ) adapted for ice was used to extract pixels containing liquid water, which uses the red and blue bands (Yang and Smith, 2013). A threshold value of 0.25 was applied, meaning pixels with NDWI >0.25 were assumed to be water-covered, following previous studies (Bell and others, 2017; Banwell and others, 2019; Arthur and others, 2020; Dell and others, 2020; Moussavi and others, 2020). No exposed rock nunataks or outcrops are present in this region and we isolated clouds, cloud shadow, seawater, shaded snow and dark areas (such as crevasses), following Moussavi and others (2020) (Table 2).…”

“…Moussavi and others (2020) found an accuracy of >94% when comparing lake extents derived from the same automated methods with manually digitised polygons. A conservative uncertainty of 1% is therefore assigned to the total lake area for each time step, following Stokes and others (2019) and Arthur and others (2020).…”

“…RACMO2.3 has been shown to realistically simulate the surface mass balance of East Antarctica (van Wessem and others, 2014). At this resolution, RACMO2.3 is known to be unable to fully resolve localised atmospheric processes around ice shelf grounding zones, including katabatic winds and enhanced turbulent mixing, but higher resolution model outputs are not yet available for this region of East Antarctica (Van Wessem and others, 2018; Arthur and others, 2020).…”