Multiple Regression Models of Snow Damage Prediction According to the Snow Damage Vulnerability Groups

Abstract

The frequency of natural disaster has been increased due to the climate change, and the heavy snow has been also increased during the winter. Therefore, many researches have been conducted to estimate snow damage vulnerability. In this study, multiple regression models for snow damage prediction were developed using snow damage vulnerability which was calculated in the previous research. The snow damage vulnerability were categrized into 5 groups. The developed models were even applied in the area where never had a snow damage in the history. Three multiple regression models were developed according to the snow damage vulnerabilities. The input data for the model were snow vulnerability index, snow depth, the exceedance percentage of snow design criteria, daily relative humidity, daily minimum temperature, and daily maximum temperature. As a result, normalized root mean square error(NRMSE) was low enough to apply the models to estimate the snow damage. The developed models could be applied to estimate the snow damage even in the areas where the snow damage have not been occurred. Keywords: Snow Damage, Snow Damage Estimation, Multiple Regression Analysis, Snow Damage Vulnerability Groups