Initialization of coupled thermo-hydro-mechanical models of permafrost terrain using the frozen-ground-fem package

Permafrost degradation under climate change poses growing risks to infrastructure and environmental stability in cold regions. Processes such as frost heave, thaw settlement, and thermokarst are intensifying due to rising temperatures, changing precipitation, and evolving snow and surface hydrology. While many thermal models capture heat transfer in permafrost, they often overlook the coupled thermal, hydraulic, and mechanical (THM) interactions that are crucial for predicting thaw-induced deformation and consolidation. A critical yet underreported aspect of permafrost modelling is the “spin-up” phase: an initialization procedure that applies cyclic boundary forcing to stabilize subsurface conditions before transient simulations. Prior studies, including Ross et al. (2022) , demonstrated that thermal spin-up alone can require over 10,000 cycles for permafrost ground, but neglected hydromechanical effects such as void ratio evolution and stress redistribution. In this study, we extend Ross et al.'s approach by implementing a coupled THM spin-up strategy using the open-source Python 3 package frozen-ground-fem. This one-dimensional, large-strain finite element framework solves for both temperature and void ratio, incorporating phase change, cryosuction, nonlinear constitutive laws, and residual stress initialization. Examples using Ross et al.'s input data illustrate how coupled initialization improves stability and accuracy of permafrost predictions. Our results demonstrate that joint thermal and hydromechanical spin-up significantly alters equilibrium profiles and enhances simulation reliability. The frozen-ground-fem model is freely available on GitHub, supporting open and reproducible development of climate-resilient permafrost infrastructure models. • Thermo-hydro-mechanical spin-up using the open-source frozen-ground-fem package. • Coupled temperature and void ratio profiles establish initial ground conditions. • Sub-Arctic and Arctic sites show distinct thermal and mechanical ground conditions. • Reproducible initialization method for robust THM modelling of permafrost terrain. • Coupled spin-up enhances model stability for long-term permafrost simulations.