Year-Round High-Resolution Sea Ice Freeboard Retrieval Using ICESat-2 ATL03 Photon Data

Abstract. Arctic sea ice freeboard is critical for estimating ice thickness and characterizing surface morphology, yet it remains poorly constrained, especially during melt seasons due to limitations in conventional altimetry products. The ICESat-2 ATL07 product retrieves surface heights at variable along-track segment lengths (10–200 m) and identifies floes and leads using fixed thresholds (photon rate, background rate and the width of height distribution) to support freeboard estimation, which smooths small-scale features and reduces reliability over sea ice surface with complex spatial variations or affected by melting. To address these challenges, we present a year-round, high-resolution (5 m) freeboard retrieval method (HRFM) based directly on ICESat-2 ATL03 photon data. A two-stage denoising strategy is implemented to robustly extract signal photons, while a machine-learning classifier, trained on 25 coincident Sentinel-2 images, discriminates between sea ice, thin ice, and leads across seasons. Identified lead segments provide local sea surface references from which freeboard is estimated. Validation against Airborne Topographic Mapper (ATM) data shows that HRFM reduces the surface-height root-mean-square error (RMSE) for strong beams from 0.12 m (ATL07) to 0.08 m (by 33 %), while weak-beam retrievals achieve comparable accuracy. HRFM better preserves ridge-related heights that are underestimated by ATL07. The classification attains a precision of 0.96 and a recall of 0.95 for lead detection, supporting reliable freeboard estimation. After applying the method at a pan-Arctic scale, the spatial patterns of retrieved freeboard are consistent with the ICESat-2 ATL20 product, but with seasonal mean differences reaching up to 0.04 m. By improving both topographic fidelity and lead detection, HRFM mitigates common limitations of ICESat-2 sea ice products and enables high-resolution freeboard estimates across seasons.