A process-based framework for validating forest landscape modeling outcomes

Abstract

Forest landscape models (FLMs) simulate forest dynamics by integrating stand- and landscape-scale processes. Thus, evaluating FLMs simulations necessitates including both processes. Thus far, stand-scale processes were evaluated in some FLMs, whereas landscape-scale processes were rarely evaluated. This study presents a framework that evaluates both stand- and landscape-scale processes. For the stand-scale processes, we proposed using stand density management diagrams to evaluate the simulated stand development trajectories that encapsulate the interplay of tree growth, competition, and mortality. For the landscape-scale processes, we evaluated seed dispersal, the basic spatial process driving forest landscape dynamics and not evaluated previously, through comparing simulated tree species colonization pattern against tree age distribution data from inventory data. We demonstrated the applicability of the framework to a 300-year historical forest landscape reconstructed using LANDIS. Given the common features, the framework is applicable to other FLMs or terrestrial ecosystem models operating at large scales.