Single-tree level optimization model for cable-based forest operations planning in steep terrain

A common challenge in planning forest operations with cable-based systems in steep terrain is to maintain the expected harvesting efficiency such that the revenues exceed logging costs. One of the difficulties in realizing a positive economic marginal return is the spatial distribution of the trees selected for harvesting. Indeed, selection scheme, generally must comply with strict silvicultural constraints, also limiting the volume of timber to be harvested. Using spatial optimization at single-tree level can address the operational problem maximizing the profit under given constraints. In this study, a decision-support optimization model based on integer programming (IP) was implemented to select trees to be harvested, according to economical and spatial objectives. To create a base to apply the optimization approach, field data and StanForD 2010 data from a Mounty 5000 cable yarder were used to develop efficiency and productivity models using multiple linear regression method. Tree parameters, as spatial input scenario, describing a specific forest stand, were derived by means of high-resolution airborne laser scanning at individual tree level. Trees were further rated by financial value and then connected with the productivity models to run the optimization process. Our preliminary results show that an improving in key-performance indicators of forest operations, such as the economic marginal return, could be achieved by pre-selecting the most suitable trees to cable-based harvesting system, while still respecting the timber volume reported in the harvest plan.