Approaches for Addressing Spatial Connectivity of Final Harvests Within Forest Harvest Scheduling Algorithms

Abstract

Harvest scheduling, or the scheduling of management activities within a forest for a given period of time, is an important aspect of forest planning. Often, harvest scheduling results in a tactical plan that allows forest managers the ability to understand where to go, and what to do, at different points in time. In the development of a harvest schedule, an objective is optimized and constraints are satisfied. As examples, an objective may be to maximize wood produced or revenue obtained over time, or to minimize environmental impact over time. Examples of constraints include restrictions on the flow of wood produced over time, the condition of the standing inventory (uncut forests), the amounts of areas of different management activities, and the location and timing of specific management activities. In many cases, these mathematical problems are formulated either with exact methods (linear or mixed-integer programming) or heuristic methods (simulated annealing, tabu search, genetic algorithms, etc.). This work describes the manner in which the connectivity of final harvests is assessed and controlled in both types of approaches. This work also explores how the control of activities differs between cases (a) when the focus is on controlling the timing of the final harvest of adjacent pairs of forest management units, and (b) when the focus is on controlling how large a collective area might become when multiple adjacent forest management units are scheduled for a final harvest within a given time window.