详情

原文传递 Comparative Assessment of Geospatial and Statistical Methods to Estimate Local Road Annual Average Daily Traffic

题名：	Comparative Assessment of Geospatial and Statistical Methods to Estimate Local Road Annual Average Daily Traffic
正文语种：	eng
作者：	Mathew, Sonu;Pulugurtha, Srinivas S.
作者单位：	Univ North Carolina Charlotte Infrastruct Design Environm & Sustainabil Ctr 9201 Univ City Blvd Charlotte NC 28223 USA;Univ North Carolina Charlotte Infrastruct Design Environm & Sustainabil Ctr 9201 Univ City Blvd Charlotte NC 28223 USA\|Univ North Carolina Charlotte Dept Civil & Environm Engn 9201 Univ City Blvd Charlotte NC 28223 USA
关键词：	Annual average daily traffic (AADT);Local road;Geographic information system (GIS);Geographically weighted regression (GWR);Kriging
摘要：	Collecting traffic data and/or estimating and reporting annual average daily traffic (AADT) is important for planning, designing, building, and maintaining the road infrastructure. However, AADT is not available for most local functionally classified roads (referred to as local roads in this paper), which comprise a major proportion of the roads in the United States. The AADT of a local road depends on geospatial data such as road density, socioeconomic and demographic characteristics, and proximity to the nearest nonlocal road. The suitability of these explanatory variables for modeling local road AADT has not been widely explored, nor have methodological approaches been comprehensively compared in the past. Therefore, the focus of this research is on exploring geospatial and statistical methods and conducting a comparative assessment to estimate local road AADT. The AADT based on traffic counts collected at 12,899 stations on local roads in North Carolina during 2014, 2015, and 2016 was considered in model development and validation. The road, socioeconomic, and demographic characteristics based on the data gathered from the North Carolina Department of Transportation (NCDOT) for 2015 were considered as the explanatory variables. Five different modeling methods were examined and compared to estimate AADT on local road links. They include traditional ordinary least squares (OLS) regression, geographically weighted regression (GWR), and geospatial interpolation methods such as kriging, inverse distance weighting (IDW), and natural neighbor interpolation. The model development and validation results showed that the GWR model performed better compared with the other considered geospatial and statistical methods. The GWR model can better capture the effect of geospatial variations in the data, by geographic location, when estimating local road AADT. Local road AADT estimates help practitioners in planning and prioritizing road infrastructure projects for future improvements and air quality estimates, in addition to Highway Safety Improvement Program (HSIP) and Highway Performance Monitoring System (HPMS) reporting.
出版年：	2021
期刊名称：	Journal of Transportation Engineering
卷：	147
期：	7