Geospatial Statistical Analysis Seminar

June 2 - 6, 2003

Venues:

Mornings: The Coastal Institute on Narragansett Bay Auditorium, URI Bay Campus
Afternoons: The Policy Simulation Lab, Coastal Institute in Kingston

Objective:

This seminar series is designed to provide training for advanced environmental science students in geospatial statistical methods, by providing both overview and hands-on experience with fundamental and state-of-the-science analyses. These methods are used to display and interpret patterns of data on the landscape, explore and interpret relationships between different types of data in a spatial context, and account for spatial patterns in the context of other types of analyses. Training will be provided by experts brought together for this seminar, and include participation by research scientists applying these techniques to current environmental issues.

Benefit:

There is increasing recognition of the value of analyzing environmental information within a spatial context, and the importance of understanding relationships between environmental variables and biological and ecological responses. A variety of statistical approaches are being developed to quantify such relationships, and address environmental issues including, e.g., interpretation and prediction of spatial and temporal patterns of wildlife distribution, assessment of species-habitat relationships, and diagnosis of causes of ecological impairment at the watershed, waterbody and watershed scale. Environmental science students taking this course will learn what types of analytical methods are available, the data requirements for these methods and the degree of certainty associated with the outputs. This seminar provides a unique opportunity to become familiar with a wide spectrum of geospatial techniques, and the experts who are developing and applying them.

Approach:

This seminar was developed as an intensive introduction to some of the analytical methods that are being developed and applied to environmental issues. This 5-day seminar series (2 – 6 June 2003) will be taught by internationally-recognized academics and professionals, who will provide lecture (morning series) and hands-on training (afternoon series). The lecture series will be open to about 100 on-site participants. The on-site lecture and hands-on sessions will be offered at the University of Rhode Island main campus and Bay Campus, which is located next to the US EPA’s Office of Research and Development National Health and Environmental Effects Laboratory (NHEERL) Atlantic Ecology Division (Narragansett, RI). The hands-on session will be limited to about 40 participants, who will have access to appropriate computer technology in the Policy Simulation Lab in the Department of Environmental and Natural Resource Economics. Workstations will be configured to support some of the most frequently used spatial analysis tools, including ArcView, ArcGIS, FragStats, SAS, etc.. The morning sessions will outline the theory behind each of the techniques, the advantages or shortcomings of each technique, and examples of when and how this technique could be applied. Topics that will be covered include use of presence/absence data, necessary steps to looking for patterns in point data, continuous data, a new alternative to kriging, and lattice/graph theory. The afternoon sessions will allow practicing different techniques using available analysis packages.

Participants:

The audience for the seminar consists of graduate and advanced undergraduate students, faculty, and research personnel from the EPA Environmental Sciences Research and Training Opportunities Program (RTOP) Consortium (University of Maryland Eastern Shore, University of Massachusetts, University of Rhode Island, Delaware State University, University of Southern Maine, and Rutgers University), staff from the US EPA NHEERL-Atlantic Ecology Division.