Estimating Pavement Temperature Using Machine Learning Models Based on Climatic Data

Ö. Terzi, T. Baykal, Şebnem Karahançer, Ekinhan Erişkin, Serdal Terzi

2025

Data confidence 70 medium

How this score is calculated

It is built from checkable properties of the record itself: whether the article has a DOI, how many independent sources confirmed its data, whether the required fields are filled in, and how recent the latest confirmation is. Disagreement between sources and a missing identifier lower it. A value of 80 or above means the record was published without further checking; 60 to 79 means it is published at reduced confidence, with a manual check advisable. A record verified by an editor by hand is never pushed below 90 by a recalculation. This scores our data about the article, not the article itself.

Abstract

This study focuses on estimating pavement temperature using machine learning models based on various climatic parameters. Climatic parameters, including air temperature, precipitation, wind speed, relative humidity, specific humidity, surface pressure, dew point temperature, and wet-bulb temperature, are obtained from meteorological databases. These parameters directly affect pavement temperature and are used as input parameters for the machine learning models. Random Forests, Extra Tree, Gradient Boosting, XGBoosting, Light Gradient Boosting Machine, CatBoost, and Hist Gradient Boosting algorithms are employed to model this relationship. Model performance was evaluated using R² and RMSE metrics, with results indicating that R² values for the testing set ranged between 0.90 and 0.93, demonstrating the effectiveness of the applied algorithms. The study confirms the capability of machine learning algorithms to estimate pavement temperature effectively, highlighting LightGBM and CatBoost as particularly promising approaches. These findings offer valuable insights for future research on model selection and development in pavement temperature prediction.