Peter Garrahan and Bob Bowen
The database is designed as a hierarchy. The least changing variables are in higher order tables (e.g., cruise id, year, month, etc.), while variables that change the most are in the lower order tables (e.g., time of collection, net number, taxon collected, etc.) There are five levels (master tables) within the database: CRUISES, STATIONS, TOWS, NETS, and RAW_COUNTS. The information in the relational database is contained only once as opposed to being repeated as header information. Unique identification numbers are given to each table entry and these links are then used to track relationships between and among tables. For example, all stations from a particular cruise will have its Master Cruise Data identifier (MCD_ID) and subsequent data that link to those stations will also link to that cruise. Searches and selections of data can include information from any or all tables, as necessary.Table names, variable names and descriptions are given in the following tables. Most column variable names and instrument names were taken from the U.S. GLOBEC Georges Bank data thesaurus; those that were not, follow the GLOBEC data protocols. The taxonomic code variable (TAXCOD) is from the National Oceanographic Data Center's Taxonomic List, v. 8. Taxonomic information is built into these ten-digit codes as they reflect the systematic nomenclature.
Data derived from the RAW_COUNTS (numbers per cubic meter and numbers per square meter of water column) are set up as views. These views are recreated as they are accessed and show the most recent data. Provisions were made in the database to also include zooplankton data from Process cruises as well as data from other instruments (CTD data or fluorometry).
During the session, a poster displayed the database design and a laptop accessed the GLOBEC Zooplankton Homepage at http://globec.gso.uri.edu.