Abstract: In this paper, we use the daily minimum temperature data of Dayi, Xiling Snow Mountain Ski Resort and the surrounding area from October to February of each year from 2005 to 2017, temperature and precipitation data from ECMWF.Through statistical analysis of the difference between the average minimum temperature of regional automatic station and Dayi station.Based on the forecast value of the lowest temperature of Dayi station and the difference of the lowest temperature one by one, the correction method of the difference of the lowest temperature one by one is established.we replaced the temperature forecast of the ski site with the grid temperature forecast data closest to the ski resort, then selected the lowest temperature, and used the temperature forecast change to establish a model temperature correction method for the lowest temperature forecast. We established a integrated correction method of the minimum temperature based on the weighted average of the minimum temperature difference correction and the model temperature correction. Based on the above three methods, combined with the results of intelligent grid precipitation prediction and the minimum temperature threshold for snowfall.Three revised prediction methods for the lowest snowfall temperature in Xiling Snow Mountain were established. By examining the above three lowest temperature prediction methods, it was found that the lowest temperature integrated correction method had the highest forecast accuracy.