Spatio-Temporal Forecasting of Traffic Accidents Using Prophet Models with Statistical Residual Validation
Preprint
- 15 January 2026
- preprint
- Published by MDPI AG in Preprints.org
Abstract
This study develops a spatio-temporal forecasting artifact for road traffic accidents in Ecuador, addressing a critical limitation in existing predictive approaches that rely predominantly on point error metrics without validating the statistical assumptions underlying forecast uncertainty. Motivated by pronounced territorial heterogeneity in accident incidence and the need for reliable decision-support tools, the research proposes a multiregional modeling framework that integrates statistical residual validation to enhance the robustness of road safety planning. Using a dataset of 27{,}648 monthly observations covering all 24 provinces from 2014 to 2025, the study applies the Prophet model within a Design Science Research paradigm and a CRISP-DM implementation cycle. Separate provincial models are estimated with a 24-month forecasting horizon, and methodological rigor is ensured through systematic residual diagnostics using the Shapiro--Wilk test for normality and the Ljung--Box test for temporal independence. Empirical results indicate that the Prophet-based artifact outperforms a naïve seasonal benchmark in 70.8\% of the provinces, demonstrating excellent predictive accuracy in structurally stable regions such as Tungurahua (MAPE = 10.9\%). At the same time, the framework enables the identification of critical emerging risks in provinces such as Santo Domingo and Cotopaxi, where projected increases exceed 49\% despite acceptable point forecasts. The findings confirm that point accuracy alone does not guarantee the validity of confidence intervals and that residual validation is essential for trustworthy uncertainty quantification. Overall, the proposed approach provides a robust foundation for a predictive surveillance system capable of supporting differentiated, evidence-based road safety policies in territorially heterogeneous contexts.Keywords
All Related Versions
- Version 1, 2026-01-15, Preprints
- Published version: Computation, 14 (4), 78.
This publication has 50 references indexed in Scilit:
- Integrating deep learning and statistical models for traffic prediction and accident prevention in smart citiesCluster Computing, 2025
- Hierarchical spatial network models for road accident risk assessmentAnnals of Operations Research, 2024
- From sky to road: Incorporating the satellite imagery into analysis of freight truck-related crash factorsAccident Analysis & Prevention, 2024
- Effectiveness of road safety interventions: An evidence and gap mapCampbell Systematic Reviews, 2024
- Introduction: A Broad Perspective on the Concepts of Urban Dynamics, Environment, and Health2023
- Machine learning and automatic ARIMA/Prophet models-based forecasting of COVID-19: methodology, evaluation, and case study in SAARC countriesStochastic Environmental Research and Risk Assessment, 2022
- A Comparison Between Seasonal Autoregressive Integrated Moving Average (SARIMA) and Exponential Smoothing (ES) Based on Time Series Model for Forecasting Road AccidentsArabian Journal for Science and Engineering, 2021
- Financial Time Series Forecasting Using Prophet2021
- Design Science Research Methodology2021
- A performance analysis of prediction intervals for count time seriesJournal of Forecasting, 2020