Setting forest road maintenance and upgrade priorities based on environmental effects and expert judgment

The prioritization of road maintenance projects is an important forest engineering task subject to limited budgets and competing investment needs. Large investments are made each year to maintain and upgrade forest road networks to meet economic and environmental goals. Many models and guidelines are available for single-criterion analysis of forest roads, but guidelines for multi-criterion analysis are rare. Additionally, even single-criterion approaches often rely on expert judgment to inform models of user preferences and priorities. These preferences are used to make tradeoffs between alternatives that contain data that are physical and biological, quantitative and qualitative, and measured on many different scales. The Analytic Hierarchy Process (AHP) has the potential to provide a consistent approach to the ranking of forest road investments based on multiple criteria. AHP provides a consistent, quantifiable approach to problems involving multi-criterion analysis, but it has not been applied to road management. The road investment problem differs from traditional AHP applications in that a large number of alternatives are compared at one time. AHP methodology is discussed, including the foundations, assumptions, and potential for use in prioritizing forest road investments to meet economic and environmental goals. Using AHP, the problem of scheduling maintenance and upgrade of forest roads is presented as a hierarchy and pairwise comparisons are elicited from decision makers to determine the relative importance of road characteristics to management goals. Issues regarding the use of the AHP to prioritize forest road investments are discussed. Three metrics are proposed to assist in evaluating the quality of a solution and to revise preferences when necessary. Various models using mathematical programming are used to incorporate the results of an AHP analysis into the scheduling of forest road maintenance and upgrade activities involving non-monetary benefits. Solutions to two non-linear integer programming formulations are found using a threshold accepting heuristic.