Abstract

This thesis focuses on how uncertainty in forest data affects the outcome of management planning and decision making. In a review article (paper I), previous research aiming to evaluate forestry data are described. The methodology used and the results presented were discussed. A general conclusion was that previous studies concerned highly simplified planning situations, leaving some doubts concerning their real world applicability. In papers II and III, two quantitative approaches of data evaluation were applied to data from sample plot imputations. In paper II, a cost plus loss analysis of using the data for forest management decision making is presented. The usefulness of the data in forestry scenario analysis were scrutinized in paper III as errors in the predictions of future forest states, harvest levels and net income flow. In both papers it was concluded that improvements in the methodology of assessing data would be required. In paper IV, an advance of the cost-plus-loss methodology was suggested by developing a simulation system that aims to capture the hierarchical structure and iterative nature of forestry planning. The simulation system included the tactical and operational levels of a continuous planning process at a specific corporate forest owner. It was characterized by annual re-planning with the option to reassess data of selected stands prior to operational planning. The planning simulation system was used in paper V for evaluation of current practice data. It was concluded that high decision losses occurred as a result of errors in the studied data. The introduction of holding level wood-flow considerations and incitements to cluster harvest activities reduced decision losses compared to stand wise planning without such considerations.