Typhoon damage in coastal areas is a complex disaster characterized by heavy rain, strong winds, and turbulent waves. The strength of these strong winds and waves is determined by the intensity of the typhoon. In this study, a variable representing strong winds and waves was developed. A damage prediction model using rainfall, wind, and the developed variable was then proposed and thresholds for each variable were calculated across cities, counties, and districts located in coastal areas. Two damage prediction models were compared based on variables. Standard normal variables related to strong winds and rainfall by duration were calculated using Empirical Copula. The first logistic regression model was constructed using these variables. The second model used a multiple logistic regression model, considering strong winds and rainfall by duration. Although there are more cities, counties, and districts where the multiple logistic regression model exhibited better accuracy, there have been several instances where the typhoon damage induction threshold suggested by this model was abnormal. On the other hand, a relatively stable critical value was yielded using standard normal variables by Empirical Copula. This is because standard normal variables are calculated based on the probability according to the order of rainfall and strong wind magnitude. Hence, the critical value falls within the minimum and maximum range of observed values. According to the research results, it is possible to predict whether typhoon damage will occur in cities, counties, and districts located in coastal areas using the typhoon information provided by the Korea Meteorological Administration during a typhoon event, which can aid in preparing for typhoon disasters.