Estimating Transport Footprint along Highways at Local Levels: A Combination of Network Analysis and Kriging Methods

ABSTRACT

Transport impacts on the natural environment have been increasingly quantified by ecological footprint analysis. Existing studies estimate transport footprint at local levels when traffic data are available. This study introduces an approach to estimate and visualize the transport footprint along highways where traffic data are unavailable by using a combination of network analysis and Kriging methods. The use of this approach is demonstrated in Houghton County, Michigan, USA. The estimated transport footprint along highways provides useful information for environmental and transportation planners to understand traffic impacts on the natural environment at local levels and identify hotspots/corridors of the transport footprint.

Key Words:

• highways
• Kriging
• network analysis
• transport footprint

ACKNOWLEDGEMENTS

The authors are indebted to Jun Zhu at University of Wisconsin-Madison for her suggestions on the methodology proposed in this study. Appreciation is extended to William Lam and Agachai Sumalee, the co-Editors, and two anonymous referees for their many helpful comments.

Notes

¹Estimating transport footprint in more recent years would provide more timely updates on the impact of highway traffic on the environment because some of the transport footprint components—vehicle fleet composition, fuel efficiency, and traffic counts—could change over time. This study is conducted based on 2000 data as they are easily available, and this study focuses on developing a method for estimating transport footprint rather than on the estimation itself.

²Prior studies have respectively suggested that, for one liter of gasoline, the ecological footprint is 3.3 m² (Statistics Canada 1996), or for one liter of diesel, 3.9 m² (Girouard, Walsh, and Becker 1999) and 3.6 m² (Skatteudvalg 1999).

³Using the average percentage of 45% likely underestimates the transport footprint, considering the long winter and heavy snow in Houghton County. A higher road maintenance percentage, when becoming available, should be used in future research.

⁴The local forest productivity adjustment factor of 0.9 is based on personal communication with Dr. Kurt Pregitzer (now at the University of Idaho but, at the time of this study, at Michigan Technological University), who has studied the ecosystem and carbon balance in Houghton County. The adjustment factor of 0.9 can be applied to the period of 1990–2010, during which time period forest coverage and composition are relatively constant in Houghton County and its surrounding areas.

⁵The vehicle types include passenger cars, motorcycles, other 2-axle/4-tire vehicles, single-unit 2-axle trucks with 6 or more tires, combination trucks, and buses. The proportion of each vehicle type registered as of 2000 and their fuel efficiency for the year 2000 are used to estimate the average fleet fuel efficiency. The proportion of vehicle types and fuel efficiency is multiplied for each vehicle type, and the products are then summed up. See Tables 3 and 4 of the article by Chi and Stone (2005) for details.

⁶Chi and Stone (2005) included physical highway areas in the calculation of transport footprint. This study excludes it, as the main goal is to develop an approach for estimating transport footprint along highways where traffic data are not available. In addition, according to Chi and Stone (2005), the physical transport footprint is negligible in the total transport footprint.