Modelling and Evaluating the Sensitivity of River Ice Thickness in the Mackenzie River Basin to a Changing Climate

Climate change is altering river-ice regimes across northern basins, with important implications for river hydraulics, infrastructure and flood hazards. In this study, river-ice thickness was simulated at 59 hydrometric stations across the Mackenzie River Basin (MRB) for the period 1980–2024 using a thermodynamically based Stefan ice-growth model driven by daily mean air temperature from the Canadian Surface Reanalysis (CaSR). Cumulative freezing degree-days (CFDD) were derived from 10 km gridded air temperature fields, and site-specific Stefan coefficients (α) were calibrated using measured average ice thickness data from the updated Canadian River Ice Database (CRID). The model reproduced observed river-ice thickness with good accuracy, achieving a median coefficient of determination (R2) of 0.96 across all stations. Basin-wide analysis revealed statistically significant warming in annual and seasonal air temperatures, averaging 0.37 °C decade−1, accompanied by widespread declines in CFDD. These climatic changes translated into widespread reductions in simulated annual maximum river-ice thickness, averaging 1.1 cm decade−1, together with a shift toward earlier peak ice thickness. Sensitivity analyses with uniform air-temperature increases of +1 to +3 °C further indicated average reductions in maximum river-ice thickness of up to 10 cm. Overall, the results demonstrate that the Stefan ice-growth model provides a robust and computationally efficient framework for basin-scale assessment of river-ice thickness and show that river ice across the MRB is already responding to climate warming, with continued thinning expected under projected future warming scenarios.