Research
The central theme of our research is the development of novel statistical approaches to treat a diverse array of emerging marine ecological data sets, with the aim of addressing specific outstanding research questions in marine ecology. The systems under consideration are typically non-Gaussian and evolve in space and time according to nonlinear dynamics (often described by stochastic differential equations). To sample these systems on the appropriate space and time scales, observation technology increasingly relies on remote sensing, telecommunications, and automated in-situ sensors based on optics, electromagnetics and acoustics. Marine environmental data are thus complex data structures characterized by measurements on many different types of biological, physical and chemical variables, large data volumes, and complicated spatial and temporal dependence structure.
In order to address the important ecological questions, our methodology must consider a diversity of data types (categorical, binary, continuous), and the nonstandard error structures inherent in sampling the dynamically evolving nonlinear ocean environment. Formulation and analysis of these models will require both close interaction with marine scientists and due consideration to the interplay between dynamic systems and statistical estimation problems. Towards this end, we have established close working relationships with university researchers, as well as government scientists to ensure access to a broad spectrum of ocean ecosystem data.
As noted above, the complex data types currently being acquired by ocean observing programs have already outstripped existing analytical tools - and there are ambitious future plans for greatly expanded ocean data collection. The nature of the required methodology will, of course, depend on the questions asked, which are in turn shaped by the data available. The approach taken is collaborative and interdisciplinary. It involves both statisticians and marine scientists, and support research by graduate students and post-doctoral fellows. The central focus of the project is to answer important ecological questions by developing and applying novel statistical methods.
Some sub-projects are outlined below. For several of these there is already work in progress, while for others the research is still in the beginning stages. All have significant potential for addressing outstanding issues in marine ecology.