High-Temporal and Accurate Calving Front Dataset of Petermann Glacier (2016–2023) Using Sentinel-2 and SAM

Monitoring glacier calving fronts is essential for understanding ice dynamics and their response to climate change. The present ice frontdatasets are limited to temporal resolution and precision, making it difficult to efficiently capture the rapid detailed changes in calving processes. In this study, we proposed a novel approach of calving front extraction derived from Sentinel-2 satellite images, integrated with samgeo and geemap modules based on Segment Anything Model (SAM) and Google Earth Engine (GEE) framework. Our approach significantly reduces computational costs and improves efficiency and model generalizability ability, eliminates the requirements of data preprocessing, manual annotation, and model training compared with traditional methods. And we acquire a high-temporal resolution (sub-weekly) and highly accurate datasets of the Petermann Glacier calving front from 2016 to 2023. The accuracy assessment demonstrates an average deviation of just 1.5 pixels (~30m), outperforming the accuracy of CALFIN deep learning-based results (~100m), which exhibit strong adaptability and robustness with tuning points prompts. This novel workflow and dataset provide a precise and efficient solution for monitoring glacier advance and retreat, contributing to improved assessments of ice sheet mass loss and glacier hazards.