Abstract: The traffic line in the alpine region is deeply affected by drifting snow disasters. The weather along the entire traffic line is very important on predicting the snowdrift disaster, but there is still a lack of research in this area. In the present study, the mesoscale model WRF (the Weather Research and Forecasting Model) is used to predicate the weather in the winter of the Karamay-Tacheng Railway in Xinjiang from 2018 to 2019. This simulation is mainly two-level nested with a horizontal resolution of 10 km-2 km and adopts different microphysics and planetary boundary layer physics to obtain different distribution probability of wind velocity and dominant wind direction along the railway line. The results show that the WRF simulation results can reflect different probability distribution of wind speeds in different sections along the railway. Combined with the angle between the dominant wind direction and the traffic line, it can provide a more accurate basis for predicting the probability of drifting snow disasters in different sections and a reference for the design of traffic lines and sections form.