Estimating vehicle miles traveled (VMT) in urban areas using regression kriging

Abstract

The recent increase in demand for performance-driven and outcome-based transportation planning makes accurate and reliable performance measures essential. Vehicle miles traveled (VMT), the total miles traveled by all vehicles on roadways, has been utilized widely as a proxy for traffic impact assessment, vehicle emissions, gasoline consumption, and crashes. Accordingly, a number of studies estimate VMT using diverse data sources. This study estimates VMT in the urban area of Bucheon, South Korea, by predicting the annual average daily traffic for unmeasured locations using spatial interpolation techniques (i.e., regression kriging and linear regression). The predictive performance of this method is compared with that of the existing Highway Performance Monitoring System (HPMS) method. The results show that regression kriging could provide more accurate VMT estimates than the HPMS method and linear regression, especially with a small sample size. additional control variables. Pathomsiri et al. [16] develop an econometric model for estimating VMT in multivehicle households and apply it to the 2001 National Household Transportation Survey. Erlbaum [17] applied the fuel-sales-based method to estimate VMT. White [18] and Greene [19] used odometer reading data to estimate annual VMT by driver. However, the previously mentioned methods, based on household activity surveys, fuel sales, and odometer reading data, are too resource-intensive and costly to perform regularly. Also, these methods reflect personal travel and do not reflect the total number of vehicles on the road.Traffic-count-based VMT estimation methods are currently the most commonly performed and preferred method because they are based on actual data for vehicle movement [20]. The predictive accuracy of traffic-count-based methods depends on the quality and coverage of traffic count data, given that the length of all road network sections is known. Thus, it can be said that if traffic counts were available for all roads in a network, the VMT estimate would be the most accurate measure of vehicle movement. Traffic counts are, however, only available at segments of road networks where there is a count station because of the relatively high cost of the stations. For urban areas, traffic count data are collected less frequently than in other areas because of the complexity of measurement [21]. For this reason, the Highway Performance Monitoring System (HPMS) method of VMT estimation, a representative traffic-count-based method performed in the USA, simply extrapolates the VMT of a sample section into other sections, so long as the other sections are in the same strata of traffic volume group and road functional system [20]. Despite the advantages of being a relatively simple and quick procedure, the HPMS method has been criticized by its accuracy and sample size demand needed to achieve the required precision level. Moreover, stratification of sampling by traffic volume group requires knowledge of the traffic count information on all roads [14]. A number of rec...

Vehicle Kilometers Traveled (Vkt): Metodologia Para O Cálculo E Suas Aplicações Nos Sistemas De Transporte

“…Os modelos baseados no tráfego não circulante utilizam dados dinâmicos que são fornecidos por diferentes agências governamentais (Kim et al, 2016).…”

The Effect of COVID-19 Pandemic on Emitted PM2.5 in Urban Road Networks: Using Loop Data and Kriging Method for Passenger Cars in the Central Part of the City of Lodz

“…But, they have a great potential to reduce data collection cost and time and to assist analysts to conduct their studies by providing representative flows for the networks (Pinto et al, 2020) when real values are not available. Regarding the advantages of Kriging methods over the other predictive ones for spatial analysis and their acceptable results for modeling street networks (Kim et al, 2016;Klatko et al, 2017;Pinto et al, 2020;Shukla et al, 2020), this model is described in the next part of the text.…”

Factors Affecting Road Traffic: Identifying Drivers of Annual Average Daily Traffic Using Least Absolute Shrinkage and Selection Operator Regression

“…For example, in Eom et al ( 30 ) and Wang and Kockelman ( 31 ), Kriging interpolation is applied using the mean square prediction error (MSPE) and absolute percentage error (APE) to evaluate the spatial models developed. Similarly, in Selby and Kockelman ( 19 ), Kriging interpolation and geographic weighted regression (GWR) are also applied and the models are assessed with the APE, while in Kimet et al ( 32 ), regression Kriging interpolation is applied and the model outputs are assessed by using the mean absolute percentage error (MAPE) evaluation metric.…”