Higher Antarctic ice sheet accumulation and surface melt rates revealed at 2 km resolution

Abstract

Antarctic ice sheet (AIS) mass loss is predominantly driven by increased solid ice discharge, but its variability is governed by surface processes. Snowfall fluctuations control the surface mass balance (SMB) of the grounded AIS, while meltwater ponding can trigger ice shelf collapse potentially accelerating discharge. Surface processes are essential to quantify AIS mass change, but remain poorly represented in climate models typically running at 25-100 km resolution. Here we present SMB and surface melt products statistically downscaled to 2 km resolution for the contemporary climate (1979-2021) and low, moderate and high-end warming scenarios until 2100. We show that statistical downscaling modestly enhances contemporary SMB (3%), which is sufficient to reconcile modelled and satellite mass change. Furthermore, melt strongly increases (46%), notably near the grounding line, in better agreement with in-situ and satellite records. The melt increase persists by 2100 in all warming scenarios, revealing higher surface melt rates than previously estimated.

Major Modes of Climate Variability Dominate Nonlinear Antarctic Ice‐Sheet Elevation Changes 2002–2020

“…This could be because the SMB models are unable to reproduce the precipitation in this region, especially in ∼2007 at Thwaites Glacier, but this would require a highly localized signal as this event does not occur at PIG. This is not implausible given existing SMB model limitations in lowaltitude coastal regions (Kappelsberger et al, 2023;Noël et al, 2023). The misfit could be caused by errors in background altimeter models, however we note we obtain nearly identical results using the alternative data set of Schröder et al (2019) and misfit also exists in SMB-corrected GRACE fields (King et al, 2023b).…”

Physically‐Informed Super‐Resolution Downscaling of Antarctic Surface Melt

“…The SUPREME melt product is compared with another downscaled RACMO product besides RACMO 5.5 km, referred to as RACMO 2 km, as developed by Noël et al (2023a). RACMO 2 km was obtained through a statistical-downscaling approach applied to RACMO 27 km, resulting in a spatial resolution of 2 km.…”

“…A frequently employed technique is statistical downscaling, a method that enhances the resolution of RCM variables, typically by leveraging their linear correlation with elevation. Recently, statistical downscaling has been employed to generate high-resolution surface mass balance (SMB) variables for Antarctica (Gallée et al, 2011), including snowfall (Ghilain et al, 2022) and surface melt (Noël et al, 2023a). Similarly, over Greenland, SMB components have been subject to statistical downscaling in studies by Hanna et al (2005), Hanna et al (2008Hanna et al ( , 2011, Franco et al (2012), Noël et al (2016), and Tedesco et al (2023).…”

Distinct Central and Eastern Pacific El Niño Influence on Antarctic Surface Mass Balance

“…This ocean-driven mass loss is counteracted by mass gain in some regions of Antarctica, driven by processes underlying surface mass balance (SMB), that is, the sum of accumulation and ablation (melting) processes at the ice sheet surface (Agosta et al, 2013;Mottram et al, 2021). Snow accumulation, the major control on SMB in Antarctica, has been increasing over this period, although with significant spatial and temporal variability (Hanna et al, 2020;Noël et al, 2023). Understanding and projecting the balance between mass loss due to basal melting and calving and mass gain due to snow accumulation is essential for understanding the Antarctic contribution to global sea level rise.…”