More than 50 years of operational research in Forestry

In this blog post, I will refer to a set of methods that have been fundamental in the development of forest management worldwide in recent decades. I am referring to operations research (OR) techniques (initially focused on optimisation methods) necessary to model complex typical forestry problems that encompass different inputs, conditions, and dimensions. Obviously, they are not techniques typical of the forestry arena, although it can be said that, given the problems that forest managers face, the forestry field constitutes a desirable proving ground for methodological and algorithmic developments of management science.

From a personal point of view, my research career cannot be conceived without using these techniques (and, indeed, without the permanent support of my Dissertation Supervisor, Prof. Carlos Romero). In short, the reason for bringing this up has been a paper published a few weeks ago by three prestigious researchers (Rönnqvist, Martell, and Weintraub), and, to facilitate exposition, I will call it “PAPER1”. These authors analyse how the use of these techniques has evolved over the past decades. They completed this task without using bibliometric milestones, but their extensive experience, which, in my opinion, is highly appreciated. I highlight the fact that none of them is North American, which is a clear indicator of the popularisation of these techniques at a global level. On the other hand, the solidity of this contribution is supported by the trajectory of the authors, who have already published constructive reviews and articles in the past to achieve a relatively complete perspective of the progress of these techniques in forestry. Specifically, and based on my own experience, I would highlight two of them. One, published by Prof. Weintraub and Romero, where they compare the progress of these techniques in the forestry and agricultural fields; the other, where the researchers who signed PAPER1 also appear as co-authors. In this study, the nine authors point out 33 open and key problems in the application of these techniques in the forestry field. Related to this, it is appreciated that, in PAPER 1, they provide an assessment of what has happened in recent years regarding these open questions.

Looking into the contents of this paper (PAPER1) very briefly, note that it focuses on optimising the forest value chain management (timber), from the harvest scheduling problem to aspects much closer to the reality of the forest industry. In addition, very didactic sections are provided, dedicated to the history of OR in forestry or to the classic models of forest management, which have been the germ of popularising these techniques to solve real cases in many nations. Unfortunately, Spanish forests are not included in this set of countries. From a historical perspective, and as I have commented above, there are clear relationships between forestry and these techniques. Thus, the authors cite OR pioneers such as Egon Balas, a recognised expert in techniques such as integer programming. He developed seminal models (the late 1950s) on forest product transportation in his native Romania before emigrating to the USA. It is in this spirit that comes to my mind the name of Jared Cohon, author of a pioneering book on multi-objective programming and Dean of Forestry at Yale for five years.

From a forest management perspective, any feasible method used must pay attention to at least three basic prisms. They are, in alphabetical order: economic, spatial, and temporal. The conjunction of these three components is resolved in the models that use OR techniques, from the most basic ones, such as linear programming to the most complicated models that combine various OR methods. Note that the complexity is growing not only due to the increase in the area to be managed, the size of the industry, or the road network that must be designed and through which the daily vehicle routing must be optimized, but also because the number of constraints tends to grow exponentially. These three components underlie the paper according to the type of planning considered. However, OR techniques can also be differentiated between those that address deterministic problems and those that incorporate uncertainty in their analysis. It is easy to think that uncertainty is crucial if we associate it with disturbances in forest systems (wildfires are the most evident case). A large section is also dedicated to the techniques used in this kind of problems.

A final chapter describes the challenges associated with using these techniques in the future. Thus, summarised aspects such as climate change, carbon sequestration, and biodiversity conservation are cited. At this point, the focus could be broadened by assuming the challenge of fundamentally addressing a single provision ecosystem service (wood) to try to incorporate into the analysis other that the authors affirm. However, without any doubt, another ESs are candidates (cultural or regulation ones). It is necessary to remember that, regarding provision ES, there are also models for different NTFPs, which are paramount in the Iberian Peninsula (mushrooms, pine nuts, cork, etc.). Extension towards these new ES is sometimes inaccessible due to the difficulty of finding reliable models to start the modelling phase. However, it is a beautiful challenge, and, at this point, it should be remembered, on the one hand, that forest modelling benefits from techniques such as GIS and remote sensing, etc., that help to refine and improve the initial models, as well as disseminate the results obtained. On the other hand, specific procedures address these multi-objective problems, such as multi-criteria decision-making techniques (MCDM).

Although we could talk about other challenges, I would like to finish with one in particular that is becoming increasingly important: the integration of the preferences of different stakeholders in forest management. Although, in this case, forest ownership is of considerable importance, it is increasingly claimed that the participants related, for various reasons, to the use and enjoyment of forest systems can, in some way, contribute to modify the forest management plans with their preferences. I am referring, in general, to a set of techniques called group decision-making techniques (GDM), which incorporate various procedures, and not all of them could be classified within the umbrella of OR. However, a trend that seems to be consolidating is that as the problems become more complex (size, objectives, restrictions, ecosystem services, uncertainty, incorporation of the preferential weights of the stakeholders, etc.), the hybridisation of different techniques, such as MCDM and GDM, is seen in the literature. Finally, all that remains for me is to thank the authors for such a brilliant contribution, recommend reading it, and wish that in the coming decades the conjunction of OR techniques and forest systems and what surrounds them will be as fruitful as up to now.

Leave a Comment

Your email address will not be published. Required fields are marked *


Other Post