Improving dynamic treatment unit forest planning with cellular automata heuristics

High-resolution, wall-to-wall forest information enables precision-driven decision-making in forest management planning. In a case study we compare planning approaches using such information for solving long-term forest planning problems. The two first approaches are based on dynamic treatment unit (DTU) planning with high-resolution cells (12.5x12.5 m2) or segments (0.27 ha on average), respectively, solved with a cellular automata heuristics. The third approach is a traditional stand-based approach using stands (5.2 ha on average) and linear programming to solve the planning problem. Fixed costs associated with cutting operations are quantified as each treatment unit is charged with an entry cost of 10 000 SEK. The entry costs are included in the DTU approach while in the stand approach entry costs are applied accordingly in a post-optimization routine. In large, the analyses are based on open-access tools and data provided by Swedish authorities. The traditional stand approach produced plans with 5.2-2.7% lower net present value compared to DTU planning. Most of the differences were caused by greater suboptimal losses in plans produced with the stand approach, but entry costs were also lower in DTU plans. While forestry was less profitable, treatment units were more spatially compact with stands, especially compared to cell-based plans. Therefore, we reason that a combination of modelling of direct costs and use of spatial proxy variables, such as common border length, may be advisable in DTU planning to achieve compact and realistic treatment units. Finally, the results indicate that high-resolution data and DTU planning may better utilize forests’ potential of economic production, compared to the traditional stand approach.

show abstract

The multi-faceted Swedish Heureka forest decision support system: context, functionality, design, and 10 years experiences of its use

Lämås

Sängstuvall

Öhman

et al. 2023

Front. For. Glob. Change

View full text Add to dashboard Cite

For several decades, computerized forest decision support systems (DSS) have helped managers and decision makers to analyze different management options and supported the search for preferred management alternatives. In Sweden, a country rich in forests and with a long tradition in intensive forest management, such systems have been developed and available since the 1970s. Changes in societal as well as in forest owners’ preferences and objectives in the 1990s led to a need for forest DSS handling broader perspectives compared to precedent single-objective timber-oriented systems. In Sweden, this led to the initiation of a research programme in the beginning of the 2000s aiming at developing a versatile and multi-objective forest DSS, resulting in the first version of the Heureka forest DSS released in 2009. The system handles several forest values, such as timber and biofuel production, carbon sequestration, dead wood dynamics, habitat for species, recreation and susceptibility to forest damages (spruce bark beetle, wind-throw and root rot). It contains a suite of software for different problem settings and geographical scales and uses simulation as well as optimization techniques. Three software handle projections of the forest using a common core of growth and yield models for simulating forest dynamics. A fourth software, built for multi-criteria decision analysis and including a web-version, enables also group decision making and participatory planning. For more than 10 years, the Heureka system has been used in teaching, environmental analysis, research and as decision support in practical forestry. For example, several research groups using the system for analyses in different problem areas have so far published more than 80 scientific papers. The system is used for nation-wide forest impact analysis for policy support and all large and many medium-sized forest owners use it for their long-term forest planning, meaning that it directly influences forest management decisions and activities on more than 50% of the Swedish forest area. Besides presenting the present system and its use, we also discuss lessons learned and potential future development.

show abstract

Improving dynamic treatment unit forest planning with cellular automata heuristics

Cited by 3 publications

References 22 publications

Performance Analysis of ConvLSTM, FlamMap, and CA Algorithms to Predict Wildfire Spread in Golestan National Park, NE Iran

Performance Analysis of ConvLSTM, FlamMap, and CA Algorithms to Predict Wildfire Spread in Golestan National Park, NE Iran

Dynamic treatment units in forest planning improves economic performance over stand-based planning

The multi-faceted Swedish Heureka forest decision support system: context, functionality, design, and 10 years experiences of its use

Contact Info

Product

Resources

About