The purpose of this research is to address outstanding questions in marine ecology through the development of novel statistical approaches for the analysis of complex spatial/temporal data. An initial workshop on the project theme took place August 17-19, 2005 at Dalhousie University. It brought together approximately 70 researchers in statistics and marine ecology. Outstanding research questions were reviewed and statistical methods and modeling approaches with the potential to address such questions were identified. As a result of this workshop, in April 2006 NPCDS agreed to fund the project.
This research, led by Michael Dowd, Joanna Flemming, Chris Field, and Rick Routledge, is concerned with developing and applying novel statistical approaches in order to analyze complex marine ecological data. The primary motivation is that advances in observing technologies have far outpaced our ability to analyze these new and complex data types. This has resulted in a research "bottleneck" wherein marine scientists await development of the statistical and modeling tools needed to fully exploit these new data. A central aim of this work is to help eliminate this bottleneck.
The marine ecological data considered here are highly diverse and includes: marine mammal tracking data along with observations of their diet (fatty acids); satellite imagery of plankton and primary production; in situ observations of marine biogeochemical variables from autonomous moving platforms; marine gene sequence data; and fisheries data, both in the form of time series and other survey data, as well as newer types such as video imaging from submersibles and satellites. These data are characterized by a complex temporal and spatial dependence structure, and are not generally amenable to analysis by traditional (linear) methods. Consequently, our goals are to develop new data analysis and statistical modeling techniques to allow researchers to better utilize these data to address outstanding questions in marine ecology.