Glaciers in high mountain regions are retreating rapidly due to climate change, exposing subglacial environments sculpted over millennia. Overdeepenings are a common part of the post-glacial landscape, and their development impacts landscape evolution, proglacial sediment dynamics, glacial lake formation, and associated hazards like Glacial Lake Outburst Floods (GLOFs). While Geographic Information System (GIS)-based approaches integrating Digital Elevation Models (DEMs) and ice thickness models have become critical for estimating overdeepening locations and characteristics, there has been little assessment of the extent to which the results are dependent on the input data that are used. To address this, the study first evaluated the relative influence of DEMs and ice thickness models on overdeepening predictions in the Central Himalayas. It compared results from five ice thickness models: Farinotti’s Ensemble, Huss and Farinotti, GlabTop2, OGGM, and Millan; paired with four DEMs: Copernicus, ASTER GDEM, SRTM, and ALOS PALSAR. The analysis revealed that ice thickness models exert a greater influence than DEMs on overdeepening predictions. The combined framework was then applied to other regions, including the Peruvian Andes, Alaska, the European Alps, and New Zealand’s Southern Alps, to further assess overdeepening variability and its dependency on the choice of ice thickness dataset. While no consistent trends emerged across regions and models, certain ice thickness models, such as OGGM and GlabTop2, were found to overestimate overdeepening parameters in many instances. Despite some commonalities, the maximum overlap in overdeepening predictions between any two models was 53.6%. To address the uncertainty in using a single ice thickness dataset to predict overdeepening locations, a potential framework that integrates multiple datasets is therefore proposed. In the absence of robust validation data, this approach could be used to provide confidence bounds on the estimation of overdeepening size and volume, helping to communicate uncertainties to stakeholders. More generally, the findings highlighted the variability in overdeepening parameters, emphasizing the need to account for this range when using overdeepenings as a foundation for further analysis. Establishing a clear purpose before estimating overdeepenings is recommended, as it can guide the selection of the most appropriate ice thickness model. This study provides insights into the consistency among ice thickness models and underscores the importance of multi-model approaches for robust and reliable assessments.
