Robust decision making for forest management under climate change and uncertainty

Abstract: Due to its long-term nature, forest management faces various uncertainties that may negatively impact the forest and its provision of services to nature and humans. The most prominent uncertainty that raises a lot of attention in forest science is climate change. The recent drought years have strongly impacted the forest and increased mortality, leading to forest diebacks in Germany and many other parts of the world. This intensified public awareness regarding forest mortality and raised the need to deal with adaptive forest management. Climate change and many other uncertainties that strongly impact the delivery of ecosystem goods and services from forests can be considered as deep and dynamic over time. Deep uncertainty is characterized by the inability to attach a single probability distribution to a set of possible outcomes. In order to integrate it in decision-making it requires non-probabilistic approaches that explore the decision and uncertainty space under a large set of plausible scenarios of the uncertain factors.
The question how to deal with deeply uncertain situations and how to integrate deep uncertainties in decision making processes, analytically and methodologically, has opened up a whole new field of research that is known as “robust decision making under deep uncertainty” that especially found a rise since the early 2000’s. Originally settled in the field of water resources management, it finds increasing application in other fields of natural resources management under uncertain (climate) change both in science and in practice. The core design principles of these Robust Decision Making (RDM) approaches include i) the use of Exploratory Modelling over large uncertainty and management scenarios to explore the uncertainty and decision space and ii) the dynamic nature of decision making, which bases decisions and actions on so-called observed signposts that signal when and how to adapt. Until today it has found only very limited application in forest management. Few studies have applied similar concepts with simplified methods or have applied Robust Optimization, which is viewed as critical when applied under deep uncertainty.
Against this background, this thesis explores the application of RDM approaches to the field of forest management as a new application domain. What robust decision making approaches exist and how have they been successfully applied to other fields of natural resources management? How could they be adapted to forest management as a new context? In order to design dynamic decision rules that adapt decision to observed signposts: What uncertainties can critically impact the performance of forest management and, based on the critical uncertainties, what could be potential signposts that signal the need to adapt current management? On that basis: How can dynamic adaptive decision rules be formulated and how could they improve the performance of current management and its robustness to climate change and other deep uncertainties?

We started answering these questions by reviewing existing approaches to RDM and their application to natural resources management under climate change uncertainty (among other uncertainties). Next we used Exploratory Modelling, as an essential part of RDM, and a global sensitivity analysis in combination with a beech growth model to identify the relative impact of different sources of uncertainties on the management objectives and to identify potential signposts for adaptation. The analysis was conducted for an even-aged beech stand in South-West Germany, a test plot under classical beech management that is based on stand basal area at different time steps. We generally chose beech management (desired stand basal area) for our analysis as beech is the dominant tree species in Germany and considered as relatively robust to climate change. Based on the results we again used Exploratory Modelling and the growth model, as well as multi-objective optimization to derive adaptive decision rules based on observed signposts. We explored how they could improve robustness and performance in multiple objectives compared to a continuation of current management for different growth regions in Germany when facing climate change and other sources of deep uncertainty. We especially focused on past mortality as a signpost and mortality reduction as an objective, since the recent drought years have shifted the management focus in this direction.
We found a number of RDM that follow a similar pattern. These approaches can easily be mixed and matched, depending on the decision context and analysis goal. With regard to measuring robustness we found that a global satisficing robustness metric is especially suited for forest management, since forest management is often tied to minimum performance standards. The sensitivity analysis revealed that different sources of uncertainties have a different relative impact, depending on the management objectives. For example, climate change showed the most critical impact on carbon sequestration, while it had a negligible impact on the Net Present Value of timber yield. We recommended, next to economic signposts, the use of past stand basal area or volume increment as a promising signpost for adaptation, since it is highly affected by climate change and is regularly tracked during forest inventories. Conversely, multi-objective optimization under uncertainty showed no clear relationship between past basal area increment and decisions regarding stand basal area (the decision lever). The reason might be that stand basal area does not send a strong signal regarding the change in management compared to a lever such as the amount of basal area per hectare removed. Yet, using past mortality as a signpost for an adaptive decision rule led to substantial increases in robustness and overall performance (except for carbon sequestration which was barely affected) for different growth regions in Germany. We observed that the potential for the adaptive decision rule is very region-specific. For the Norther growth regions robustness could not be increased to a high level, while for the central and southern German regions the adaptive decision rule led to total robustness.
In conclusion, the results suggest a high potential for adaptive decision making based on signposts in forest management to achieve a higher robustness to climate change and other uncertainties. Extended research on the efficacy of different signposts and potential integration of multiple signposts into an adaptive decision rule could further improve the performance and robustness of adaptive decision making in forest management

Object Details

• Location: Deutsche Nationalbibliothek Frankfurt am Main
• Extent: Online-Ressource
• Language: Englisch
• Notes: Universität Freiburg, Dissertation, 2020

Classification and Topics

• Keyword: Climatic changes ; Forest management ; Uncertainty ; Decision making

Contributors, Places and Time

• Event: Veröffentlichung
• (where): Freiburg
• (who): Universität
• (when): 2021
• Creator: Radke, Naomi
• Contributor: Hanewinkel, Marc ; Keller, Klaus ; Albert-Ludwigs-Universität Freiburg. Fakultät für Umwelt und Natürliche Ressourcen

Further information

• DOI: 10.6094/UNIFR/175698
• URN: urn:nbn:de:bsz:25-freidok-1756987
• Rights: Der Zugriff auf das Objekt ist unbeschränkt möglich.
• Last update: 15.08.2025, 7:22 AM CEST