Integrating forest-stand simulations with paleoecological records to examine long-term forest dynamics

Abstract

Computer models permit us to study the future dynamics of present-day forests. Before applying a simulation model to predict long-term behavior of a forest, the model output should be tested against appropriate forest data. A forest-stand simulation model (FORET) and one of its variants (FORWET) have been tested with time-parallel simulations in which climate variance was inferred directly from pollen chronologies (internal consistency) and from such independent sources as other pollen chronologies, regional periglacial geomorphology, and global climate model output (authenticity). Comparison of simulation output translated into inferred pollen, with actual pollen chronologies, confirmed the reliability and adequacy of the model for certain predictive tasks, e.g., estimating response of forest biomass storage to expected CO/sub 2/-induced warming. 82 references, 5 figures, 3 tables.