Research on Avalanche Meteorological Hazard Assessment Based on Multi-source Data and Multi-factor

The accurate assessment and early warning of avalanche disasters are crucial for disaster prevention and mitigation in mountainous areas during winter and spring. This study systematically developed a meteorological risk assessment framework for avalanches in the Tianshan region of Xinjiang, integrating historical avalanche cases with meteorological data and research literature. The framework comprises four key components: topographic factors (disaster-prone environments), pre-avalanche snow conditions, meteorological conditions prior to the event, and weather conditions during the avalanche period. It includes seven evaluation factors: pre-avalanche snow depth, altitude, slope gradient, prior temperature data, prior cumulative snowfall, and daily snowfall amount, new snow accumulation depth on the day of the event. On this basis, the paper first normalizes each disaster factor by the method of graded value assignment, then calculates the hazard index of the environment and the hazard index of meteorological factor respectively by the method of equal weight sum, and then obtains the comprehensive meteorological hazard index of avalanche by the algorithm of multiplication, and finally obtains the quantitative grading of avalanche meteorological hazard index and the evaluation result of avalanche meteorological hazard index. The model is applied to calculate the spatial distribution of avalanche risk in the Tianshan area of Xinjiang in February 2024. The results show that the actual avalanche occurrence area is consistent with the high risk area calculated by the model. This study provides a preliminary quantitative method and technical support for future avalanche risk assessment and early warning. In the future, it will further integrate the disaster-prone environment and underlying surface elements, optimize the normalized grading threshold and factor weight distribution, and attempt to conduct multi-scenario experiments to enhance the model's comprehensive predictive capability and applicability.