Statistical Modeling of Weather Impact on Road Accidents: A Nonlinear Distributed Lag Analysis in Tirana, Albania

This research aims to explore the impact of weather conditions on road accidents in Tirana through an in-depth analysis of meteorological data alongside accident records. This study employs a distributed lag non-linear model (DLNM) for its analysis of whether temperature, rainfall, and cloud coverality affect road safety both immediately after exposure but also at later time points. The results point to a U -shaped correlation between temperature levels and the risk of accidents, with both lower and higher temperatures contributing to increased accident risk. Rainfall, on the contrary, may be associated with a lower likelihood of an accident, although the effect is only very weakly significant: it makes sense that people might drive slightly more carefully during bad weather. The cloud cover analysis shows that the presence of medium-height clouds increases the danger to aircrafts, while high-altitude air masses help reduce it. In various countries around the world, studies have shown similar results, further demonstrating the impact that weather has on road safety as a whole. In light of these findings, the study recommends local-specific efforts to focus on public awareness campaigns and improvements in infrastructure that could help address hazards during extreme weather situations. The study has also discussed its limitations and recommendations for extending the present research such as employing real-time weather information, incorporating more variables.