Connecting large-scale atmospheric circulation with Greenland's surface mass balance variability by impact-weighted self-organizing maps

Large-scale atmospheric circulation exerts a dominant control on the surface mass balance (SMB) of the Greenland Ice Sheet, yet circulation classifications are often optimized for atmospheric variability rather than for surface impacts. Here, we present an impact-oriented classification approach that emphasizes those regions of large-scale atmospheric circulation that are most relevant for Greenland’s SMB. Daily summer (June-August) 500 hPa geopotential height fields over a North Atlantic-Arctic domain encompassing Greenland are classified using self-organizing maps (SOMs). Prior to classification, the geopotential height fields are weighted based on their correlation with Greenland-wide SMB derived from a regional climate model (Modèle Atmosphérique Régional), such that regions exhibiting a strong linkage to SMB variability influence the circulation classification more. The weighting is derived from correlation patterns between geopotential height anomalies and Greenland-wide SMB anomalies, with a scaling factor systematically varied and selected to maximize both the separation of SMB characteristics across circulation regimes and the distinctness of the associated geopotential height composites. The resulting classification yields a set of circulation types that closely relate to differences in Greenland-wide SMB. Compared to unweighted SOM classifications, the impact-weighted approach enhances the separation of SMB responses across circulation regimes. By further analyzing the evolution of circulation regimes and their impact on Greenland’s SMB over time, we aim to improve understanding of changes in large-scale drivers relevant for the Greenland Ice Sheet mass loss.