Prediction of On-Road CO2 Emission in Urban Area Using State-of-The-Art Ensemble Machine Learning Model

Transportation emissions, including carbon dioxide (CO2), nitrogen oxides (NOx), carbon monoxide (CO), and particulate matter (PM), are a significant contributor to air pollution and greenhouse gas emissions. In urbanizing cities, poor air quality due to transportation emissions, especially CO2 from vehicles, is a considerable challenge. CO2 is a greenhouse gas contributing to climate change and global temperature rise. Therefore, it is critical to figure out CO2 emissions from transportation. Travel activity data must be collected throughout time and space for emissions connected to traffic to develop temporal CO2 emission prediction models. However, emission prediction research from on-road vehicle fleets is limited, particularly in developing countries such as Malaysia. Therefore, a new modeling approach using ensemble machine learning techniques is proposed in this study, which combines lab-tested and fieldwork data collection to estimate CO2 emissions at a finer geographical scale. The approach aims to uncover dynamic vehicle behavior and CO2 predictions in Klang Valley urban city by identifying patterns in vehicle activities in road networks.