Moving Beyond the Incorrect but Useful Paradigm: Re-envisioning Forests in Earth System Models

Abstract:

Earth system models represent forests, their interactions with the atmosphere, and changes in community composition and ecosystem structure. Together, these processes simulate biosphere feedbacks in a changing climate. The approaches used to simulate forests in Earth system models grossly simplify the complexities of forests. A central feature of forests is their size structure and the associated vertical profile of tree heights. The land surface components of Earth system models abstract this complex vertical structure into a single layer of leaves that exchange energy, mass, and momentum with the atmosphere. This approach has been recognized as "incorrect but useful," in contrast with multilayer canopy models, which are thought to be physically correct but unnecessarily complex. The utility of big-leaf vs. multilayer canopy models reflects a 40-year-old debate about how to model forest-atmosphere exchanges. Similarly, Earth system models have adopted biogeochemical models, which abstract forests into carbon pools and associated transfers among pools, as the paradigm to model the terrestrial carbon cycle. These models, too, ignore the size-structure of forests and reflect a 100-year-old debate in ecology about the nature of forest succession. Another class of models, individual-based forest models (so-called gap models), represent forest succession as an assemblage of individual trees competing for resources in a vertically-structured canopy. Here, we review the science of forests in Earth system models, the history of big-leaf and biogeochemical models, and show how the community is moving in directions that embrace fine-scale forest dynamics and the central paradigm of vertically-structured canopies. Early forest gap model development advanced a theory of forest dynamics that continues to be relevant today. We provide examples from our own work with multilayer canopy models and also show how individual-based gap models can be used to inform the transition from biogeochemical models to ecosystem demography models. As the Earth system community re-envisions how to simulate forests, the fundamental insights and knowledge of forests pioneered in gap modeling are coming to the forefront of global models.