Van tellen naar voorspellen: Sturen op risico’s met een voorspellend wiskundig model op basis van historische brandweerdata

In this paper, we present a mathematical model for the prediction of chimney fires using data collected by the Twente safety region over the years 2004 up to 2015. The explanatory variables used are the number of inhabitants per areal unit of 500 by 500 metres, the daily mean temperature and an indicator variable for October. Spatially and temporally correlated noise is added to be able to capture latent factors such as human behaviour. The model is validated using the data over the years 2016 and 2017. The model is implemented as a dashboard on a public website and yields as output a hazard map for chimney fires in Twente based on the weather forecast for the coming six days. This way, the dashboard can be used for prevention and planning purposes.