An assessment of forest sector modeling approaches: conceptual differences and quantitative comparison

Abstract

Forest sector models are widely used for policy and economic analyses. Basic assumptions vary considerably between models, but little attention has been paid to the impacts these differences may have on model results. Norway provides a great opportunity to fill this void as it has two forest sector models currently in use based on different modeling assumptions, but sharing the same data source. In one model, agents are assumed to be myopic in the meaning that they rely only on the current market conditions. Harvest behavior is in this model based on econometric relations and exogenous forest growth rates. The second model approach employs extensive forest data and assumes that agents have perfect information over the model time horizon. We discuss the differences between the two modeling approaches and compare quantitatively the model results of two case studies. However, both types of models are rather sensitive to changes in assumptions and data. The strengths and weaknesses of the two model approaches and their appropriateness for responding to the study questions should be considered when choosing modeling methodology. Using both models in parallel removes the uncertainty caused by the foresight assumption/optimization routine and thus provides in total more information.

Behavioral models of climate change adaptation and mitigation in land‐based sectors

“…Between these extremes, a range of approaches are adopted, including discussion of potential uncertainties, 78,79 the use of a number of social and environmental scenarios, 64,82 analyses that build on unstructured collections of runs, 58,74,75 structured (e.g., Monte Carlo) approaches, 36,37,56,68,87 and mathematically rigorous explorations of parameter space . Once again, the diversity of approaches limits the scope for general conclusions or cross‐model comparisons to be established 109 . This problem is exacerbated by a tendency for the least behaviorally rich models to make the most confident predictions, overlooking the uncertainties (or even inaccuracies) inherent in their basic assumptions 110 .…”

Combining backcasting with forest sector projection models to provide paths into the future bio-economy

“…This may have influenced the shape of the harvest curve in the S3 and subsequent scenarios, with much larger leaps in the second half of the time horizon. Whether perfect foresight models or dynamic-recursive models that assume myopic agents are preferable in market models is a topic of discussion (Heide et al 2004;Sjølie, Latta, Adams, et al 2011;Sjølie et al 2015); in any case it is important for modelers and users of results to be aware of the features and effects of the model in use.…”

“…Trømborg and Solberg (2010) found that energy price increments improved the production level in the sawmill industry and reduced the output of pulp and especially particle boards. Sjølie et al (2015) compared impacts in two forest sector models of Norway, NorFor and NTM III, of changes in sawnwood demand and installation of a biodiesel plant. More available sawmill residues caused by sawnwood demand shifts led to higher stationary bioenergy production.…”

“…NorFor has been extensively applied for policy and economic analyses in the Norwegian forest sector (Sjølie et al 2013a(Sjølie et al , 2013b(Sjølie et al , 2013c(Sjølie et al , 2015. Model structure is described in detail in Sjølie, Latta, Gobakken, et al (2011) and data in Trømborg and Sjølie (2011); only a brief description is thus provided here.…”

Forest sector modeling

“…The last paper (Sjølie et al 2015) is a conceptual and quantitative comparison of two forest sector models that share most of the data, but differ in the optimization routine and timber supply representation. While the first model uses recursive-dynamic optimization with timber supply derived from econometric relations, the second model is based on inter-temporal optimization with timber supply formulated by data from forest inventories coupled with a set of management options.…”