Historical evaluation and projection of precipitation phase changes in the cold season over the Tibetan Plateau based on CMIP6 multimodels

The solid precipitation in the cold season over the TP is important for the recharge of snow cover, glaciers, rivers, etc. During the past decades, the TP has experienced enhanced warming in the cold season, which could transfer more snowfall to rainfall and influence the land-atmosphere hydrological cycle of the Asian water tower. However, most CMIP6 models underestimate the cold season temperature and overestimate the total precipi-tation, and the performance of the precipitation phase in the CMIP6 simulations remains unclear. This study investigates the precipitation phase bias in the cold season by comparing simulations from CMIP6 to observa-tions. The results suggest that the snowfall-to-precipitation ratio (SPR) from CMIP6 was significantly under-estimated, and the SPR at the southern edge of the TP was underestimated by approximately 30%. However, the decreasing trend of SPR is significantly overestimated. Comparing the influence of the surface temperature and total precipitation, the correlation coefficient between the mean SPR bias and the simulated temperature bias is-0.701 with a 99% confidence level, which is much greater than that with total precipitation bias. The mean SPR in the seven best models is much greater than the ensemble mean of the original 23 CMIP6 models over the TP. Moreover, the SPR varies significantly under different future scenarios. Under the most severe scenario, the decrease in SPR exceeds-21.45%, substantially influencing the water balance over the TP, while the decline is only-9.79% under the weak scenario. Therefore, the warming caused by high emissions should not be ignored when projecting the precipitation phase, and the significant effect of surface temperature on the precipitation phase requires additional consideration when investigating the hydrological cycle imbalance over the TP.