Automatic detection of forest management units to optimally coordinate planning and operations in forest enterprises.

Mountain forests provide not only wood as a raw material but also numerous ecosystem services, such as protection against natural hazards, recreation and carbon sequestration, and they are important hosts for biodiversity. To manage these forests efficiently and in a target-oriented manner, both forest management planning and efficient harvesting operations are required. However, in most cases these two aspects are handled independently from each other. To link planning with forest operations, it is essential to divide forests into smaller areas with characteristics that are as homogeneous as possible, so-called forest management units (FMUs). The goal is that each FMU has self-contained fine access (e.g. skid roads, cable roads), and that the FMUs can be managed and planned independently. The aim of this study was to develop a spatial optimisation model that automatically identifies FMUs. The optimisation has three goals: [I] FMUs should be as compact as possible (spatially contiguous as the best case); [II] forest management should be technically and operationally coordinated within an FMU; and [III] FMUs should be as homogeneous as possible, for example regarding site properties, ecosystem service provided, and administrative affiliation. Results showed that our presented spatial optimisation model is a capable method for automatically identifying FMUs. The approach used to set up the model based on a p-median problem formulation (mixed integer linear programming) led to clearly comprehensible solutions that can be achieved in a reasonable computation time. Three solving strategies for successful computation implementation are described. Although the raw results must be reviewed by experts, they facilitate the planning process. More scenarios can be evaluated compared with the classical manual planning approach, ultimately leading to higher-quality solutions.