A Dynamic Forest Management Model to Understand Interactions Between Human Systems and Forests
Studies have found that understanding forest management is critical in understanding the interaction between the carbon cycle and the integrated human-Earth system. This makes effectively representing forest management decisions such as planting and harvesting important. In the current literature, there are three broad categories of models which represent forest management decisions. The first are forest sector models which represent extensive detail in the forest sector but are limited in representation of other sectors or other land uses. On the other hand, there are multisector models (MSD models) where both the forest land type actively competes with other types of land use for investment, and the land use or land sector itself in these models is coupled with other sectors such as energy, water, and climate. However, the representation of forest change in these models is often at a very aggregate level with no representation of optimal age-based rotation for forest harvests. Finally, there are land-focused models that operate at a much finer scale, which represent linkages between land sectors at a very fine scale. However, the effects of energy and other sectors of the economy on the forestry sector are not accounted for unless coupled with larger MSD models. In the current literature, there are no global integrated human-Earth system models that endogenously determine rotation ages or other forestry decisions that are responsive to other sectors. To fill this gap in the literature, we implement a novel dynamic forest harvest model in a global state-of-the-art multi-sector dynamics model, namely the Global Change Analysis Model (GCAM).
Under a Baseline scenario, we have found that there will be a small loss of forest cover globally by the end of the century, largely driven by demand for food crops. In a scenario where global energy and terrestrial system net carbon is consistent with a 2.6 W/m^2 radiative forcing in 2100 and the value of terrestrial carbon is considered in land use decisions, afforestation and maintenance of older forests emerge as key scenario components. This will result in increased forest cover driven largely by higher plantation cover by 2050. However, this 2.6 scenario does still see a decline in forest cover later in the century, with saturating demand for wood products and increasing demands for bioenergy. Similarly, the 2.6 scenario does see increasing roundwood prices as land dedicated to bioenergy increases, driven by corresponding increased demands in the energy sectors. This, in turn, has the effect of increasing the average rotation age of forests globally. This result highlights the need to consider the multisector dynamics outside the forestry or even the land sector in forest management decisions.
We implement an approach that explicitly tracks forest age and generates rotation ages for forest harvest that are responsive to changes in wood prices. Furthermore, the forest sector in GCAM competes for investment with other land use types in the future years based on harvest profit. The GCAM forestry module also includes the capability to model an economically efficient consideration of terrestrial carbon, which, depending on the scenario, will incentivize land uses that lead to increased carbon, such as afforestation. One major advantage of the new dynamic forest model in GCAM is that the forest rotation age or optimal age of harvest is determined by the wood price and is not parameterized or pre-calibrated. Correspondingly, the model will also be able to explore the impact on the optimal rotation age in scenarios where terrestrial carbon is considered.
Our research highlights the importance of integrating forest management decisions with other land-use sectors to effectively project and track changes in terrestrial carbon. The dynamic model allows for a nuanced understanding of how forest management can affect net terrestrial carbon and the broader implications for land use and energy sectors.