Effects of temporal aggregation in integrated strategic/tactical and strategic forest planning

Integrated approaches to forest planning lead to large models and there is a subsequent need for a reduction of the number of constraints and variables. One way of achieving this is to aggregate data, either spatially or temporally. In this paper an integrated forest planning mixed integer model that takes into account both long-term strategic and shorter-term tactical forest management decisions is utilised. The study analyses the consequences of temporal aggregation in the strategic part of the integrated model, whereas the tactical part is modelled by a fixed set of years divided into seasons. For reference, analyses are also made using a pure strategic linear programming model. Cases using both equal and variable strategic period lengths are presented, and two case study areas are used. Results indicate that integrated plans, as well as strategic plans, are not particularly affected by the number of equal length strategic periods when more than five periods, i.e. less than 20 year period length, are used. When modelling strategic and integrated problems using variable-length periods, care should be taken to ensure that harvest operations late in the planning horizon get enough timing options to be adequately described. (C) 2006 Elsevier B.V. All rights reserved.