Abstract: Based on geological disaster data and meteorological station observation data of Ganzi Prefecture from 2022 to 2023, the Pearson correlation coefficient was used to analyze the relationship between various meteorological factors and geological disasters. Short-duration rainfall and cumulative rainfall were selected as key meteorological indicators. The ID3 decision tree algorithm was employed to construct a geological disaster prediction model for Ganzi Prefecture, and its performance was evaluated with ROC curves. The results indicate that the model performs well in predicting geological disasters in Ganzi Prefecture and demonstrates general applicability in high-altitude and mountainous disaster-prone areas. This study provides valuable insights for enhancing meteorological disaster prevention and mitigation capabilities, thereby helping to reduce the loss of life and property.