Extreme disaster risk evaluation model based on comprehensive weighting method

The increasing frequency of deadly earthquakes, devastating rainstorms, raging storms, and other extreme natural disasters is putting the whole society at great financial risk. In this paper, the evaluation criteria of natural disaster risk were divided into 9 indicators in three aspects: Perniciousness, Exposure, and Defense. The AHP, EWM, and CRITIC were then combined to calculate the mixed weights of those indicators. The TOPSIS model was utilized to calculate the risk factors, which were subsequently divided into four levels using Fuzzy Cluster. Finally, Logistic Regression and ARIMA models were employed to predict the trend in risk in the coming years. The model was applied to California and Japan, revealing that the risk of natural disasters in these regions may decrease between 2024 and 2027.