Abstract for talk, "Environmental change research on vegetation: assimilating the evidence for understanding and prediction" by Jim Clark

Environmental change research involves attempts to understand and predict complex processes at large scales of space and time. There is often substantial information available to the investigator, but it is difficult to assimilate, because it comes from such diverse sources and scales. Examples include experiments and monitoring of individual organisms or small plots to landscape-scale manipulations to regional landscape change, observed from satellites, to paleoecological evidence for long-term change.

I summarize examples of how emerging tools can help scientists assimilate evidence to facilitate understanding and prediction. The research challenges I address include increased aridity in continental interiors, migration capacity in the face of rapid climate change, and factors affecting biodiversity, including climate change, CO2, and altered disturbance regimes. From combinations of experiments, observational evidence, wireless sensor networks, and remote sensing I discuss how Bayesian methods can help us move from data to inference to prediction.