GAM Modelling of Daily Number of Traffic Accidents as a Function of Meteorological Variables in the Czech Republic
Year: 2025
Type: article
Abstract: Meteorological conditions exert a considerable influence on traffic patterns. This paper examines the influence of meteorological variables on the daily number oftraffic accidents requiring fire brigade intervention. The influence of meteorological variables, including maximum temperature, wind speed, air pressure, precipitation,snow cover and sunshine, was examined. A Generalized Additive Model for variables with a Poisson distribution was employed for modelling purposes, as this allows for therepresentation of non-linear dependencies. The analysis demonstrates that the lowest incidence of accidents occurs at temperatures approximating 10 C. The average dailynumber of accidents increases with windy weather, the minimum number of accidents occurs at zero precipitation, and the accident rate rises with higher levels of sunshine.In the Czech Republic, the period of greatest risk in terms of road traffic accidents is the summer and winter months. The findings may have several practical applications,for example, in the improvement of meteorological warnings in traffic.
Authors: Petra Kolísková, Jiří Neubauer
Institution: University of Defence
Language: English
Topic: Traffic and Road Safety
Subfield: Safety, Risk, Reliability and Quality
Field: Engineering
Domain: Physical Sciences
Open Access status: diamond
OpenAlex ID: W4412371466