Search

Because Antarctic surface melt is mostly driven by local processes, its simulation necessitates high‐resolution regional climate models (RCMs). However, the current horizontal resolution of RCMs (≈25–30 km) is inadequate for capturing small‐scale melt processes. To address this limitation, we present SUPREME (SUPer‐REsolution‐based Melt Estimation over Antarctica), a deep learning method to downscale surface melt to 5.5 km resolution using a physically‐informed super‐resolution model. The physical information integrated into the model originates from observations tied to surface melt, specifically remote sensing‐derived albedo and elevation. These remote sensing data, in addition to a Regional Atmospheric Climate Model (RACMO) run at 27 km resolution, account for the diverse drivers of surface melt across Antarctica, facilitating effective generalization beyond the training region of the Antarctic Peninsula. A comparison of SUPREME with a dynamically downscaled RACMO run at 5.5 km over the Antarctic Peninsula shows high accuracy, with average yearly RMSE and bias of 5.5 mm w.e. yr−1 and 4.5 mm w.e. yr−1, respectively. Validation at five automatic weather stations reveals SUPREME's marked improvement with substantially lower average RMSE (81 mm w.e.) compared to RACMO 27 km (129 mm w.e.). Beyond the training region, SUPREME aligns more closely with remote sensing products associated with surface melt than super‐resolution models lacking physical constraints. While further validation of SUPREME is needed, our study highlights the potential of super‐resolution techniques with physical constraints for high‐resolution surface melt monitoring in Antarctica, providing insights into the impacts of localized melting on processes affecting ice shelf integrity such as hydrofracturing. Plain Language Summary: To improve surface melt monitoring in Antarctica, high‐resolution climate models are essential. Existing models, like the Regional Atmospheric Climate Model (RACMO), do not have a fine enough spatial resolution to capture small‐scale melt processes. To overcome this, we introduce SUPREME (SUPer‐REsolution‐based Melt Estimation over Antarctica), a method that refines surface melt data to a higher resolution of 5.5 km using an advanced super‐resolution model. We enhance this model with physical information derived from observations directly related to surface melt, specifically using albedo and elevation data from remote sensing. By incorporating these observations, along with RACMO data at 27 km resolution, we account for the various triggers of surface melt across Antarctica. SUPREME accurately predicts high‐resolution surface melt beyond the training region of the Antarctic Peninsula, showing promising results compared to existing melt observations. Further validation is needed, but this approach, combining super‐resolution techniques and remote sensing data, holds potential for accurate surface melt monitoring in Antarctica. This may advance our understanding of the impacts of localized features on processes affecting ice shelf integrity such as meltwater‐induced hydrofracturing. Key Points: Our method downscales Antarctic surface melt from a regional climate model, employing a physically‐informed super‐resolution architectureThe super‐resolution model relies on physical information derived from remote sensing data, specifically surface albedo and elevationIncorporating physical information boosts model generalization, enabling accurate high‐resolution surface melt beyond the training region [ABSTRACT FROM AUTHOR]

Regional climate models (RCMs) have a high computational cost due to their higher spatial resolution compared to global climate models (GCMs). Therefore, various downscaling approaches have been developed as a surrogate for the dynamical downscaling of GCMs. This study assesses the potential of using a cost‐efficient machine learning alternative to dynamical downscaling by using the example case study of emulating surface mass balance (SMB) over the Antarctic Peninsula. More specifically, we determine the impact of the training framework by comparing two training scenarios: (a) a perfect and (b) an imperfect model framework. In the perfect model framework, the RCM‐emulator learns only the downscaling function; therefore, it was trained with upscaled RCM (UPRCM) features at GCM resolution. This emulator accurately reproduced SMB when evaluated on UPRCM, but its predictions on GCM data conserved RCM‐GCM inconsistencies and led to underestimation. In the imperfect model framework, the RCM‐emulator was trained with GCM features and downscaled the GCM while exposed to RCM‐GCM inconsistencies. This emulator predicted SMB close to the truth, showing it learned the underlying inconsistencies and dynamics. Our results suggest that a deep learning RCM‐emulator can learn the proper GCM to RCM downscaling function while working directly with GCM data. Furthermore, the RCM‐emulator presents a significant computational gain compared to an RCM simulation. We conclude that machine learning emulators can be applied to produce fast and fine‐scaled predictions of RCM simulations from GCM data. Plain Language Summary: Over the last century, climate scientists have tried to deepen their understanding of the behavior of climate processes through two types of computer climate simulations: global (GCMs) and regional (RCMs) climate models. GCMs cover the whole planet but do not contain fine spatial details, whereas RCMs provide highly detailed information but cover small areas and come at a high additional computational cost. Therefore, we imitated regional models from global models using machine learning to facilitate their faster development. To test our machine learning framework, we focused on the Antarctic Peninsula and aimed to reproduce the surface mass balance (SMB) of ice formation and loss. We trained our model to learn the relationship between a group of low‐resolution images of climate variables and a high‐resolution image from SMB images in the same region. Our results show that the machine learning model is fast and could recreate regional images of ice sheet processes from global data almost identical to existing on‐site observations. This is a good start for further usage of machine learning emulators. In conclusion, we can make fast and detailed reproductions of SMB processes at regional scales from globally accessible climate data using machine learning. Key Points: We developed a computationally fast machine learning emulator to downscale a global climate model (GCM) to regional resolutionThe emulator reproduces regional high‐resolution surface mass balance predictions over the Antarctic Peninsula from a GCMThe imperfect model framework outperforms the perfect model framework in the application success of the deep learning emulator [ABSTRACT FROM AUTHOR]

While the influence of surface melt on Antarctic ice shelf stability can be large, the duration and affected area of melt events are often small. Therefore, melt events are difficult to capture with remote sensing, as satellite sensors always face the trade-off between spatial and temporal resolution. To overcome this limitation, we developed UMelt: a surface melt record for all Antarctic ice shelves with a high spatial (500 m) and high temporal (12 h) resolution for the period 2016–2021. Our approach is based on a deep learning model, specifically a U-Net, which was developed in Google Earth Engine. The U-Net combines microwave remote sensing observations from three sources: Sentinel-1, Special Sensor Microwave Imager/Sounder (SSMIS), and Advanced Scatterometer (ASCAT). The U-Net was trained on the Shackleton Ice Shelf for melt seasons 2017–2021, using the fine-scale melt patterns of Sentinel-1 as reference data and SSMIS, ASCAT, a digital elevation model, and multi-year Sentinel-1 melt fraction as predictors. The trained U-Net performed well on the Shackelton Ice Shelf for test melt season 2016–2017 (accuracy: 91.3%; F1-score: 86.9%), and the Larsen C Ice Shelf, which was not considered during training (accuracy: 91.0%; F1-score: 89.3%). Using the trained U-Net model, we have successfully developed the UMelt record. UMelt allows Antarctic-wide surface melt to be detected at a small scale while preserving a high temporal resolution, which could lead to new insights into the response of ice shelves to a changing atmospheric forcing. • Satellite resolution trade-offs hamper accurate surface melt detection in Antarctica. • We present UMelt: An Antarctic-wide, high-resolution surface melt record. • UMelt was developed by merging multi-source remote sensing data using deep learning. • UMelt detects intricate surface melt events that were difficult to capture before. [ABSTRACT FROM AUTHOR]