thermalFOAM: A numerical model for coastal permafrost erosion validated using physical experiments

Arctic coastlines underlain by ice-rich permafrost are retreating at accelerating rates due to the compounding effects of rising air and ocean temperatures, longer ice-free seasons, and increasing storm activity. Unlike temperate coasts, erosion of unconsolidated, ice-rich permafrost bluffs is governed by a thermomechanical process in which wave-driven heat transfer induces thawing, and hydrodynamic forcing removes the sediment matrix. Despite its significance for coastal hazard prediction, infrastructure resilience, and climate feedback, this coupled process remains poorly represented in existing models, largely because of limited experimental data, logistical challenges associated with long-term field monitoring, and analytical formulations that rely on simplified or weakly constrained parameterizations, particularly for convective heat transfer at the turbulent water-permafrost interface. Here, we present thermalFOAM, a physics-based numerical framework implemented in OpenFOAM, for simulating the ablative erosion of permafrost bluffs under wave forcing. The solver resolves transient heat conduction with phase change through an enthalpy-porosity formulation, incorporates temperature-dependent thermophysical properties, and drives dynamic mesh evolution through a calibrated erosion law. A key innovation is a wave-aware Robin boundary condition that enables spatially and temporally varying thermal forcing based on instantaneous water surface elevation, allowing the model to capture intermittent wetting that governs heat transfer in the swash and surf zones. Validation against laboratory datasets demonstrated that thermalFOAM successfully reproduced the observed niche geometries and retreat rates across the tested parameter space. This integrated framework bridges laboratory-scale process understanding and field-scale prediction, offering an open-source tool for assessing Arctic coastal dynamics under future climate scenarios.