Wave-induced erosion at ice-shelf fronts under irregular and breaking-wave conditions

Wave–ice interactions are widely recognised as one of several mechanisms contributing to ice-shelf front retreat. However, their role at small spatial and temporal scales remains difficult to quantify, particularly under breaking-wave conditions. Here, we investigate wave-induced erosion at the ice–ocean interface by combining satellite observations, laboratory experiments, and simplified, scale-aware modelling.

Ice-front retreat during the austral summer of 2024 is analysed using Sentinel-1 SAR imagery. Multiple coastline-tracking methods are applied to quantify the spatial and temporal variability of the ice front. The observations reveal periodic collapse events that tend to be temporally synchronised with elevated wave activity. This points to a strong link between wave forcing and short-term ice-front instability.

To interpret these observations, we extend classical wave-induced melting formulations by introducing a first-order, breaking-aware modelling procedure. Wave shoaling and depth-limited breaking are accounted for by tracking the evolution of wave height with water depth and by using the near-breaking horizontal particle velocity as the effective velocity scale driving heat transfer at the ice–water interface. This simple approach captures the enhancement of wave-induced erosion associated with irregular and incipiently breaking waves, while remaining computationally efficient.

The formulation is first evaluated using small-scale laboratory experiments conducted at the University of Caen, where harmonic, non-breaking waves interact with an ice cube. In this controlled experiment, measured erosion rates agree well with theoretical predictions, confirming the validity of classical approaches when wave breaking is absent. When applied to field conditions at Fimbulisen, however, the breaking-aware formulation substantially increases predicted erosion rates relative to classical theory but still systematically underestimates observed retreat. The remaining discrepancy points to unresolved turbulence processes associated with fully developed breaking and motivates the need for more advanced theoretical and experimental treatment of wave-breaking-induced mixing at the ice–ocean interface.