Route Specific Analysis and Prediction of Road Traffic Accidents in Gombe State Using ARIMA Models

Abstract

Over the years, road traffic accidents have become a significant concern in both developed and developing nations. Thousands of lives are lost annually, in addition to substantial property damage. In Nigeria, numerous studies have examined road traffic accident data, yet few have focused on the dynamics and contributing factors affecting accident severity in the North East region. This study adopts an autoregressive integrated moving average (ARIMA) approach, building upon the Box-Jenkins methodology, to examine the trends and forecasts of road traffic accidents in Gombe State. A predictive model was developed using ten years of FRSC data for four major routes: Gombe–Alkaleri, Gombe–Darazo, Gombe–Kaltungo, and Gombe–Bajoga. Results show that the Gombe–Alkaleri route has the highest accident rate, highlighting the urgent need for targeted safety interventions. The analysis identified ARIMA (1,0,2) as the optimal model for the Gombe–Darazo and Gombe–Kaltungo routes, while ARIMA (2,0,2) and ARIMA (5,1,1) provided the best fit for the Gombe–Bajoga and Gombe–Alkaleri routes, respectively. Contributing factors to accident occurrence include poor road conditions, speeding violations, and dangerous driving practices. The fitted models demonstrate robust predictive accuracy and provide evidence-based insights into future accident patterns.