Analysis of traffic accident time series in the Karlovy Vary Region and prediction of their development

Road traffic accidents represent a long-term social issue influenced by various temporal, regional and behavioural factors. The aim of this study was to identify the trend and seasonal components of road traffic accidents in the Karlovy Vary Region in the period 2015–2024, to evaluate their development and to verify the possibilities of predicting future accident rates using a selected time series model. The study was based on secondary data on road traffic accidents processed through content analysis. Descriptive statistics and time series methods were applied to evaluate accident development, while the ARIMA model was used for short-term prediction. The results showed that the development of road traffic accidents in the Karlovy Vary Region cannot be described by a clear long-term increasing or decreasing trend, but rather by significant variability over time and recurring seasonal fluctuations. Furthermore, men were consistently more frequently involved in traffic accidents than women, although the development dynamics of both groups were similar. The predictive model achieved an acceptable level of accuracy for short-term forecasting and indicated a continuation of the existing development pattern without major structural changes. The main limitation of the study is the use of aggregated monthly data, which does not allow for a detailed assessment of accident causes or severity. Future research could focus on more detailed accident characteristics, regional comparisons or the application of alternative predictive models.