How important is weather as a predictor for the demand for ski lift passes in the alpine skiing industry?

Abstract

This thesis aims to examine and interpret the importance of weather variables as predictors for demand in the Norwegian alpine skiing industry. A specific skiing facility has provided a unique data set, containing their daily sales data from the winter seasons of 2014/2015 to 2019/2020. The sales data is used in combination with simulated weather forecast data to develop linear regression forecast models. The predictive performance of the models is compared statistically to analyse the importance of weather variables for predictive accuracy. The main findings show that the importance of temperature, snow depth and precipitation for predictive purposes is low. Seasonal variables, such as day of the week and public holidays, appears to be of greater importance as predictors of demand. The authors find no statistically significant improvement in the predictive ability of models with weather variables compared to models without.