An accurate prediction of global sea-level rise requires that the cause of recent and intensifying glacier acceleration along the Antarctic ice sheet (AIS) coastal margins is understood. Ice shelf thinning and abrupt ice shelf collapse have been linked to surface melt that is prevalent across coastal Antarctica. Primarily, surface melt is determined by the surface energy balance (SEB) which is regulated locally by clouds. Clouds regulate the amount of radiation received by the surface, with competing cloud warming and cloud cooling effects.
With the CloudSat-CALIPSO satellites, cloud observations have become available for a large spatial grid, providing an AIS wide observational based dataset. In combination with the regional climate model RACMO2, a state-of-the art hybrid dataset is constructed with both high temporal and spatial resolution as well as good cloud representation. Simulations with the snow model SNOWPACK are performed to investigate the response of the ice shelves to seasonal cloud forcing, comparing an all-sky scenario to a clear-sky scenario. In the clear-sky scenario only cloud radiative effects are been removed, with the aid of neural networks. Results show that clouds have a warming effect for each season of the year with an average 16.5 W/m2 and its minimum in summer. However, daytime cloud cooling effects are shown to have more impact on meltwater production. As a result, clouds reduce melt by 23.9 ± 10.3 Gt/yr which could impact ice shelf instability. On the other hand, the cloud radiative warming more directly increases sublimation mass loss by 34.7 ± 15.6 Gt/yr. The results express the need for accurate future cloud regime predictions when predicting future AIS contributions to sea level rise.

As of yet, there is no consensus on the role of the cloud radiative effect (CRE) on the Greenland Ice Sheet (GrIS). This study focuses on the seasonal and temporal variability of the CRE, to better understand the response of the firn. To do so, we combine satellite observations, climate-model output, and a snow model. We separate short-term and long-term impacts.
The results show a positive CRE for all seasons, with an annual short-term CRE of 24.7 $Wm^{-2}$, which is largest in fall. The long-term response of the GrIS to the CRE is positive and dominant in summer ablation areas, decreasing the albedo and enhancing melt-water runoff.
This long-term effect stresses the influence of the firn conditions on its response to CRE, and highlights the need to include a snow model to study GrIS cloud radiation. The (lack off) long-term component of the CRE explains the conflict in previous studies.