Background/Question/Methods Climatic extreme events are expected to occur more frequently in the future and substantially affect ecosystem production and carbon sequestration. Quantitative knowledge of such effects is extremely limited due to a few reasons. First, a limited number of modeling studies have been conducted to examine ecosystem responses to climate extreme events. Most of the modeling studies have been used to explore ecosystem responses to interannual variability in precipitation. Not many have been designed to simulate likelihood of unprecedented climate extremes and/or record-breaking events in droughts. Second, modeling results are rarely compared with experimental results despite many field experiments that have been conducted in recent years. These studies that have compared modeling results with observations in field all show substantial mismatches. The underlying mechanisms underlying such mismatches are not fully understood. This knowledge gap hinders scientific advances in predicting ecosystem dynamics in the future warmed world with climate extremes.
In this talk, we suggest a few steps towards improving model prediction of ecosystem responses to extreme drought events. First, we need to design model structure with multiple soil layers to simulate moisture dynamics along soil profiles. If a model is used to predict long-term long responses to extreme climate, the model needs to simulate dynamic vegetation with evolving traits. Second, it is essential to use data for calibrating model parameters not only related to plant production and other carbon cycle processes but also soil moisture dynamics using statistically rigorous methods, such as data assimilation. Third, plants and microbes are likely adjusting their processes of their survivorship and growth via adaptation and acclimation in response to long-term changes in climate extreme events. It is challenging yet essential to develop model structure and/or parameterization to simulate acclimation and adaptation. We will use examples to illustrate these points.