With global warming and intensified rainfall, the heat and moisture transfer processes within frozen subgrades beneath asphalt pavements have become increasingly complex, posing risks to highway stability in cold regions. This study developed a multi-physics coupled indoor simulation system based on a typical asphalt highway structure on the Qinghai-Tibet Plateau to examine subgrade responses under solar radiation, wind, and rainfall. Results showed that rainfall shifted the dominant depth of moisture migration from 7 cm to 12 cm, with moisture at 2 cm and 7 cm increasing rapidly by 2.67 % and 1.58 %, respectively. A nonlinear decrease-increase pattern was observed at 12 cm due to the capillary barrier effect. Evaporative latent heat significantly suppressed surface warming, reducing the temperature rise at 2 cm by 59.2 %, and delayed heat transfer to deeper layers (reductions of 54.5 %-64.7 %). A cumulative heat flux prediction model, incorporating solar radiation, evaporation, convection, and surface wetting, showed high accuracy (R-2 = 0.981 and 0.952; relative errors: 4.1 % and 9.6 %). Sensitivity analysis identified the surface wetting rate coefficient (beta) and evaporation attenuation coefficient (gamma) as dominant factors (F > 6.0). These findings improve understanding of rainfall-induced thermal effects and offer guidance for climate-resilient road infrastructure in permafrost regions.

来源平台：CASE STUDIES IN THERMAL ENGINEERING