Ecological processes such as bird migration are complex, difficult to observe, and occur at the scale of continents, making it impossible for humans to grasp their broad-scale patterns directly. However, novel data sources—such as large sensor networks and millions of bird observations reported by human “citizen scientists"—are providing new opportunities to understand ecological phenomena at very large scales. The ability to fit models, test hypotheses, make predictions, and reason about human impacts on biological processes at this scale promise to revolutionize ecological science and environmental policy.
In this talk, I will present novel algorithmic approaches to overcome challenges throughout the “pipeline” from low-level data interpretation to model fitting to high-level decision-making in large-scale ecological science, including: (1) biological interpretation of NEXRAD weather radar, (2) probabilistic modeling of bird migration using citizen science data and (3) optimizing land purchases to support the recovery of endangered species. I will highlight contributions from this work that extend well beyond ecology, including a very general optimization framework for maximizing the spread of a cascading process in a network, and a formalism called Collective Graphical Models for efficiently reasoning about probabilistic models of large populations of individuals when only aggregate data is available.