Hao Yang , Kuang Xiao , Xing Xiang , Xing Wang , Xi Wang , Yunsong Du , Guangming Shi , Xin Zheng , Hongli Tao , Huanbo Wang , Fumo Yang
{"title":"基于多层感知器的高时空分辨率道路CO2排放预测","authors":"Hao Yang , Kuang Xiao , Xing Xiang , Xing Wang , Xi Wang , Yunsong Du , Guangming Shi , Xin Zheng , Hongli Tao , Huanbo Wang , Fumo Yang","doi":"10.1016/j.aeaoa.2025.100368","DOIUrl":null,"url":null,"abstract":"<div><div>On-road carbon emissions represent a significant portion of transportation emissions in China and are a critical focus for future carbon reduction efforts. High spatio-temporal resolution emission inventories are vital for facilitating dynamic carbon reduction in cities. This study employs the Multilayer Perceptron (MLP) model to simulate variations in road traffic volume at the segment level and predict on-road CO<sub>2</sub> emissions with high spatio-temporal resolution. The results demonstrate that this method can effectively reproduce the spatio-temporal distribution of on-road traffic, with R<sup>2</sup> exceeding 0.6 for most road types and overall RMSE of 88 vehicles/h, respectively. Applied in Chengdu's Jinniu District, southwestern China, results show CO<sub>2</sub> emissions peak during morning (7–9 a.m.) and evening (16–18 p.m.) commutes, concentrated on main roads. Morning peaks are lower but grow faster than evening peaks. CO<sub>2</sub> emissions significantly increase on holidays and weekends with moderate temperatures and no or light rain. These insights support urban dynamic carbon reduction planning.</div></div>","PeriodicalId":37150,"journal":{"name":"Atmospheric Environment: X","volume":"27 ","pages":"Article 100368"},"PeriodicalIF":3.4000,"publicationDate":"2025-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Prediction of on-road CO2 emissions with high spatio-temporal resolution implementing multilayer perceptron\",\"authors\":\"Hao Yang , Kuang Xiao , Xing Xiang , Xing Wang , Xi Wang , Yunsong Du , Guangming Shi , Xin Zheng , Hongli Tao , Huanbo Wang , Fumo Yang\",\"doi\":\"10.1016/j.aeaoa.2025.100368\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"<div><div>On-road carbon emissions represent a significant portion of transportation emissions in China and are a critical focus for future carbon reduction efforts. High spatio-temporal resolution emission inventories are vital for facilitating dynamic carbon reduction in cities. This study employs the Multilayer Perceptron (MLP) model to simulate variations in road traffic volume at the segment level and predict on-road CO<sub>2</sub> emissions with high spatio-temporal resolution. The results demonstrate that this method can effectively reproduce the spatio-temporal distribution of on-road traffic, with R<sup>2</sup> exceeding 0.6 for most road types and overall RMSE of 88 vehicles/h, respectively. Applied in Chengdu's Jinniu District, southwestern China, results show CO<sub>2</sub> emissions peak during morning (7–9 a.m.) and evening (16–18 p.m.) commutes, concentrated on main roads. Morning peaks are lower but grow faster than evening peaks. CO<sub>2</sub> emissions significantly increase on holidays and weekends with moderate temperatures and no or light rain. These insights support urban dynamic carbon reduction planning.</div></div>\",\"PeriodicalId\":37150,\"journal\":{\"name\":\"Atmospheric Environment: X\",\"volume\":\"27 \",\"pages\":\"Article 100368\"},\"PeriodicalIF\":3.4000,\"publicationDate\":\"2025-08-01\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Atmospheric Environment: X\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://www.sciencedirect.com/science/article/pii/S2590162125000589\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q2\",\"JCRName\":\"ENVIRONMENTAL SCIENCES\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Atmospheric Environment: X","FirstCategoryId":"1085","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S2590162125000589","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q2","JCRName":"ENVIRONMENTAL SCIENCES","Score":null,"Total":0}
Prediction of on-road CO2 emissions with high spatio-temporal resolution implementing multilayer perceptron
On-road carbon emissions represent a significant portion of transportation emissions in China and are a critical focus for future carbon reduction efforts. High spatio-temporal resolution emission inventories are vital for facilitating dynamic carbon reduction in cities. This study employs the Multilayer Perceptron (MLP) model to simulate variations in road traffic volume at the segment level and predict on-road CO2 emissions with high spatio-temporal resolution. The results demonstrate that this method can effectively reproduce the spatio-temporal distribution of on-road traffic, with R2 exceeding 0.6 for most road types and overall RMSE of 88 vehicles/h, respectively. Applied in Chengdu's Jinniu District, southwestern China, results show CO2 emissions peak during morning (7–9 a.m.) and evening (16–18 p.m.) commutes, concentrated on main roads. Morning peaks are lower but grow faster than evening peaks. CO2 emissions significantly increase on holidays and weekends with moderate temperatures and no or light rain. These insights support urban dynamic carbon reduction planning.