An event-based analysis of weather-related fire brigade operations

Extreme weather conditions like wind gusts and heavy precipitation frequently lead to impacts like treefall or flooded roads. This in turn leads to a large number of emergency calls to fire brigades, which they have to deal with in addition to their routine operations. If suitable warning information about approaching hazardous weather is available in time, there is potential for preparatory actions like calling in additional staff or preparing equipment and vehicles. However, in order to properly prepare for an event, it is important to understand how a particular fire department is affected by different types of weather conditions.

Therefore, an event detection algorithm is applied to operation data provided by different fire brigades in Southern Germany. The algorithm is applied to time series of hourly numbers of weather-related emergency calls to detect time periods with high numbers of operations. Each of the identified time periods, which we refer to as events, is characterized by different parameters, like total number of operations, maximum hourly number of operations, duration and affected area. Furthermore, the meteorological conditions during the event are characterized using convective cell tracks, tracks of large-scale windstorms, radar-based precipitation amounts and wind gusts from reanalysis data. The 100 events with the highest number of operations are selected and their characteristics are compiled into an event catalog. To facilitate the analysis of the event catalog, an interactive dashboard is developed to visualize the mutual interactions between the different event characteristics.

It is shown that thunderstorms are responsible for most of the events, followed by windstorms and continuous rainfall. In general, events related to thunderstorms are short and intense and cover a smaller area, whereas events related to windstorms generally affect a larger area. Events related to continuous rainfall are shown to have the longest durations, leading to fire brigade operations for time periods of up to several days.

The event catalog is helpful to analyze past events, but can also be useful in the context of impact forecasting. In case of an approaching meteorological hazard, it allows to visualize the potential impacts on fire departments referring to previous events and their characteristics. Furthermore, the event catalog builds the foundation to the development models for the prediction of weather-related fire brigade operations in future work.