Combinatorial Analysis of North Sea Historical Data

Principal Investigator:
Michael A. Langston, Department of Computer Science, University of Tennessee
Abstract:
The "ecosystem approach" to marine management is attracting increasing attention. Unfortunately, there is consensus neither to the precise meaning of this term nor to how useful ecosystem assessments can be effected in practice given the huge amount of heterogeneous data that is required. Understanding interactions between biology and environment are of extremely importance, as is the need to identify gateways critical for energy flow through the system. Preliminary combinatorial analyses has revealed several such gateways. These can be used as benchmarks to assess the state of the ecosystem as a whole.
This investigation employs new algorithmic techniques to elucidate and interpret complex relationships among quantifiable variables of significance to the North Sea ecosystem. Relevant variables encompass a huge variety of biotic and abiotic factors. Novel combinatorial tools and graph algorithms are used to uncover temporal and spatial relationships on a large scale. High performance parallel implementations are synthesized to extract variable sets common to multiple relationships, to determine putative inflection points, and to identify possible regime changes and other patterns of potential interest. A long-term goal is to establish data dependencies upon which we can build integrated ecosystem assessments that lead to conclusions about the impact of anthropogenic and other agents upon the North Sea.