Why we need statistical models of paleodata to predict the future

A utilitarian goal of paleoecology is "to understand the past to plan for the future." Predictive models of the future are increasingly quantitative forecast models, and these models need information on the slow and/or infrequent processes studied in paleoecology. However, paleodata cannot be incorporated into forecast models based on the qualitative expert judgement of paleoecologists. Forecasts need calibrated data with uncertainty. The PalEON project calibrates paleodata and assimilates it into models. We show that current models overestimate the rate of change in slow carbon pools. In particular, they miss the special role that late-successional species play in accumulating aboveground carbon.