Multi-Scale Modeling of Riverine Ecosystems and Responses of Fish Populations in the Context of Global Climate Change and Predictive Uncertainty
Speaker: Christopher K. Wikle, Professor, Department of Statistics, University of Missouri, Columbia, MO
Christopher K. Wikle, Professor, Department of Statistics, University of Missouri, Columbia, MO
Climate change operates over a broad range of spatial and temporal scales. Understanding its effects on ecosystems requires multi-scale models. For understanding effects on fish populations of riverine ecosystems, climate predicted by course-resolution Global Climate Models (GCMs) must be downscaled to Regional Climate Models (RCMs) to watersheds to river hydrology to population response. An additional challenge is quantifying sources of uncertainty given the highly nonlinear nature of interactions between climate variables and community level processes. This talk presents a modeling approach for understanding and accomodating uncertainty by applying multi-scale climate models and hierarchical Bayesian modeling frameworks to Midwest fish population dynamics and by linking models for system components together by formal rules of probability. The proposed hierarchical modeling approach will account for sources of uncertainty in forecasts of community or population response. The goal is to evaluate the potential distributional changes in an ecological system, given distributional changes implied by a series of linked climate and system models under various emissions/use scenarios. This understanding will aid evaluation of management options for coping with global change.