Modeling Pavement Deterioration on Nepal’s National Highways: Integrating Rainfall Factor in a Hazard Analysis

Pavement deterioration is influenced by various factors with degradation rates varying widely depending on the type of pavement, its use, and the environment in which it is located. In Nepal, where the climate varies from alpine to subtropical monsoon, understanding pavement degradation is essential for effective road asset management. This study employs a Markov deterioration hazard model to predict pavement deterioration for the national highways managed by Nepal’s Department of Roads. The model uses Surface Distress Index data from 2021 to 2022, with traffic and cumulative monsoon rainfall as explanatory variables. Monsoon rainfall data from meteorological stations were interpolated using Inverse Distance Weighted and Empirical Bayesian Kriging 3D methods for comparative analysis. To compare the accuracy of interpolated values from the IDW and EBK3D methods, error metrics such as Mean Absolute Error (MAE), Root Mean Squared Error (RMSE), and Mean Bias Error (MBE) were employed. Lower values for MAE, RMSE, and MBE indicate that EBK3D, which accounts for spatial correlation in three dimensions, outperforms IDW in terms of interpolation accuracy. The monsoon rainfall interpolated values using the EBK3D method were then used as an explanatory variable in the Markov deterioration hazard model. The Bayesian estimation method was applied to estimate the unknown parameters. The study demonstrates the potential of integrating the Markov deterioration hazard model with monsoon rainfall as an environmental factor to enhance pavement deterioration modeling. This model can be adapted for regions with a similar monsoon climate and pavement types making it a practical framework for supporting decision-makers in strategic road maintenance planning.