Analysis Of Spatiotemporal Characteristics And Driving Factors Of Taxi Carbon Emissions Based On Trajectory Data

Currently,carbon emissions from the transportation industry in China rank second only to the industrial sector,and their proportion is rapidly increasing.Taxis,as a vital component of the urban transportation system,partially fulfill residents’ travel needs.With their extensive coverage,long operating hours,and ability to reflect the state of the road network,taxis are a significant concern for industry management departments regarding carbon emissions.However,research on calculating taxi carbon emissions using trajectory data and analyzing their spatial and temporal distributions is currently limited.Furthermore,there is a lack of indepth research on the driving mechanisms behind taxi carbon emissions.This article aims to address these gaps by calculating taxi carbon emissions using GPS data,analyzing their spatial and temporal characteristics and identifying hotspots.Additionally,it investigates the driving mechanisms behind taxi carbon emissions.The findings of this study have significant implications for public transportation route planning and the development of emission reduction policies by industry management departments.The study utilizes multiple sources of data,including taxi trajectory data,point of interest,road network,and community location data.Building upon existing research on the spatial heterogeneity and driving mechanisms of taxi carbon emissions,this article utilizes March 2019 GPS trajectory data of taxis in Xi’an as the basis of analysis.To ensure data validity,the data undergoes preprocessing.The study then examines the characteristics of taxi operating speeds and calculates the carbon emissions of taxis in Xi’an using a motor vehicle emission model.Subsequently,the study defines the research period,identifies taxi carbon emission hotspots using the W-DBSCAN clustering algorithm and kernel density estimation method,divides traffic zones,and conducts regression analysis of taxi carbon emissions using the OLS,GWR,and MGWR models.The study also explores the impact of non-collinear point of interest（POI）indicators on taxi carbon emissions in traffic zones.A multi-scale geographical weighted regression model is established to quantitatively describe the spatial heterogeneity of urban facilities’ impact on taxi carbon emissions and analyze the driving mechanisms behind carbon emissions.The study reveals the following findings:1.The average speed of taxis in Xi’an is approximately 21 km/h per day,and the average speed remains consistent throughout the week.2.The daily average CO₂ emissions of each taxi in Xi’an amount to 32.97 kg,and the annual CO,NO_x,CO₂,and CH₄ emissions from taxis in Xi’an are 281.82 t,30.92 t,156,140.73 t,and 3.00 t,respectively.3.Taxi carbon emissions exhibit significant spatial and temporal distribution,with distinct high-emission periods and regions.During weekdays,the second peak of emissions is mainly concentrated on the city’s main roads and expressways,while during non-workdays,the second peak is primarily concentrated near transportation hubs,university campuses,and shopping malls,among others.4.The non-collinear POI indicators contribute to CO₂ emissions in the following order: government office density,financial institution density,accommodation service density,shopping place density,education institution density,medical density,scenic spot density,and public facility density.Among these indicators,government office density has the strongest impact on CO₂ emissions and plays a leading role.