Studying the effect of weather conditions on daily crash counts using a discrete time-series model

Abstract

In previous research, significant effects of weather conditions on car crashes have been found. However, most studies use monthly or yearly data and only few studies are available analyzing the impact of weather conditions on daily car crash counts. Furthermore, the studies that are available on a daily level do not explicitly model the data in a time-series context, hereby ignoring the temporal serial correlation that may be present in the data. In this paper, we introduce an Integer Autoregressive model for modelling count data with time interdependencies. The model is applied to daily car crash data, metereological data and traffic exposure data from the Netherlands aiming at examining the risk impact of weather conditions on the observed counts. The results show that several assumptions related to the effect of weather conditions on crash counts are found to be significant in the data and that if serial temporal correlation is not accounted for in the model, this may produce biased results.

Weather Conditions, Weather Information and Car Crashes

“…In this study, we go one step further and parametrize the influence of various characteristics of weather conditions, thereby enabling a more precise assessment of risky thresholds. The study by Brijs et al [29] comes closest to the study discussed here. In [29] an integer autoregressive Poison regression model is applied to daily crash occurrence in three-differently sized and structured urban municipalities in the Netherlands.…”

“…In [29] an integer autoregressive Poison regression model is applied to daily crash occurrence in three-differently sized and structured urban municipalities in the Netherlands. Yet, Brijs et al [29] use only data of one year, whereas structural differences between the municipalities are captured by generic city dummies, rather than by e.g. variables capturing differences in the main road network and traffic composition.…”

“…variables capturing differences in the main road network and traffic composition. As in this study, Brijs et al [29] use day-of-the-week dummies to capture typical variations (cycles) in traffic patterns.…”

Rainfall impacts on traffic safety: rain-related fatal crashes in Texas

“…Jones et al (2008) identified the following statistically significant factors: the anticipated number of fatalities based on the number and age distribution of inhabitants, street lengths and vehicle counts in the region, the proportion of local roads, numerical mean of vehicles per person, percentage of urban roads, and the geometric configuration of the roads. Brijs et al (2008) examined the significance of using an integer-valued autoregressive method to study the number of daily accidents and found that reoccurring time relationships must be considered in traffic count analysis.…”

Um Modelo Garma Bivariado Com Distribuição Condicional De Poisson

“…No contexto de Epidemiologia pode-se, por exemplo, estar interessado em modelar o número de indivíduos que sofrem de uma determinada doença através do tempo (Zeger, 1988) e (Davis et al, 1999). Ainda existem poucos modelos desenvolvidos na literatura estatística que tratam de séries temporais para dados de contagem multivariados, porém em diversasáreas pode-se ver a presença desse tipo de dados, como medicina (Lambert, 1992), marketing (Brijs et al, 2004), estudo de resultados esportivos (Karlis e Ntzoufras, 2003) e em engenharia (Lambert, 1992). Na literatura estatística/econométrica, existem vários modelos que tratam do ajuste de séries temporais univariadas para dados de contagem (Davis et al, 1999).…”