Analysis of the rainfall variability and the correlation with traffic accidents: A case study in Sydney, New South Wales

Extreme weather has been one of the massive concerns for traffic management in recent years. Many studies identified that the reason for an increase in traffic crashes is related to weather conditions, especially raining and the evolving trends in meteorological data in Australia. This research aimed to explore the variability of rainfall data over time and the association between the number of crashes (weather-related conditions) and rainfall data. The Annual Daily Maximum Rainfall (ADMR) over 40 years and Maximum monthly Rainfall (MR) over 5 years (2018–2022) data was collected from two stations near the location where traffic accidents mostly occurred in Sydney. The number of monthly accidents caused by raining (NOA) was also collected from 2018 to 2022 in Sydney. Firstly, two non-parametric tests including Mann-Whitney and Kruskal-Wallis were performed to check the hypothesis analysis. Next, correlations, regression and also multivariate tests were applied to investigate the correlation between these variables. It was found that the ADMR data in the two stations did not vary over 40 years from 1984 in Sydney. Nevertheless, there was a strong relationship between the number of monthly crashes and the maximum monthly rainfall data. As a result, it it relatively feasible to use the MR data to predict NOA data in certain years. Policymakers and researchers may use these tests for safety projections and extreme rainfall predictions in any location across Sydney.