Climate-driven evolution on thermal behavior of composite bridges in high-altitude cold regions

The thermal design of bridges, particularly in high-altitude cold regions, is facing severe challenges due to climate change, which tends to substantially increase the amplitude of temperature variation. This study aims to quantitatively assess the long-term evolution of uniform thermal actions on a steel–concrete composite bridge in a high-altitude cold region under future climate scenarios, and to evaluate whether current design code provisions provide adequate coverage against projected thermal extremes. This study focuses on a steel–concrete composite cable-stayed bridge located in the Huangnan Prefecture of Qinghai Province, China. Daily meteorological forcing sequences comprising maximum/minimum air temperature, mean wind speed and total solar radiation are constructed for the bridge site covering the period of 2001-2100. Considering the prohibitive computational cost of century-scale, day-by-day finite-element heat-conduction analyses, the effective temperature is adopted as the engineering representation of the uniform thermal action. To enable long-term projection of the evolution of uniform thermal actions, an effective temperature-meteorological parameter regression formula is employed to efficiently estimate the daily maximum effective temperature ( T e,max , the daily maximum effective temperature of the girder) and the daily minimum effective temperature ( T e,min , the minimum structural temperature below a reference state) of the bridge structure. The resulting effective temperature series are subjected to trend analysis and extreme value statistics, the projected values are benchmarked against the design limits prescribed in current bridge design codes. In particular, TaiESM1 is selected for detailed analysis as an illustrative high-temperature case because it yields the largest sample maximum of daily T e,max among the ten GCMs under SSP5-8.5; it is not treated as a general upper bound or a typical ensemble member. The extreme T e,max reaches 43.21 °C under the Shared Socioeconomic Pathway SSP5-8.5 (a high-emission scenario) scenario and the 39 °C code limit is exceeded in 17 out of 100 years. In contrast, the multi-model ensemble median at the end of the century is approximately 38 °C (SSP5-8.5), indicating that TaiESM1 represents the upper tail of the ensemble distribution. Under SSP2-4.5 scenario, the extreme T e,max is 39.63 °C with only one year of exceeded limit. The extreme T e,min is −30.25 °C under both scenarios, remaining within the −32 °C code limit. The results demonstrate that high-emission scenarios will significantly elevate the high-temperature thermal actions on bridges and increase the risk of code exceedance, while the low-temperature threshold at this bridge site remains adequately covered by current code limits. The findings provide a case-based reference for assessing climate-related thermal actions in steel–concrete composite bridges exposed to high-altitude cold-region conditions.