The long-term objective of this application is to develop and test a set of tools and procedures for spatio-temporal analysis of environmental health data in a GIS framework. Current commercial geographic information systems have very little capability to handle aspects of geographic information. Temporal data is critical to understanding environmental health, especially for effects with long latency periods.
Specific aims are to define appropriate data models and query languages, as well as logical operations, for handling geospatial lifelines, which ideally are continuous time-stamped records of people's positions in geographic space. Methods for visualization, generalization, and handling missing data all will be developed and tested. These methods will allow researchers to roll back residential locations of people who are cases of some health problem, to discover whether they were more strongly clustered at times in the past than they are today. This could counteract the tendency for human migration and mobility to break up and thus mask geographical clusters of cases for conditions that have long latency periods. While the core concepts of this project will involve collections of individual-level data, we also will examine how aggregate data can be used to address some of the same issues. The project also will investigate implications of the availability of such analysis, database, and reasoning procedures for personal and data privacy. Existing methods for interpolating and visualizing three-dimensional data will be adapted to deal with representations of environmental measures as functions of two spatial and one temporal dimension. We will then develop methods to efficiently intersect individual geospatial lifelines with such fields representing environmental hazards or toxins, to determine which individuals are likely to have been exposed, and to estimate both maximum and cumulative exposure levels. The project also will determine methods for efficiently finding geospatial lifelines most similar to a given one, so that geospatial life history can be controlled for in case-control studies.