Designing environmental databases for statistical analyses

The Environmental Monitoring and Assessment program (EMAP) collects data th at are used to statistically assess the environmental condition of large ge ographic regions. These data are then posted on the EMAP web site so that a nyone can use them. Databases used for the statistical analyses, "analytica l" databases, differ in design from the "general-use" databases used by a s econdary audience. Their scope is usually restricted in time, in geographic extent: and in type and content of data. often being limited to a single s cientific discipline. Their structure may be more horizontal than vertical, so that statistical programs can import the data easily. Their design is s trongly influenced by the nature of the scientific analysis because the goa l is to create a good computing environment for that analysis. We illustrat e these aspects of design with an analytical database for estuaries in the U.S. mid-Atlantic region.