A database is comprised of multiple tables that contain records and fields. Fields describe the type of information stored, and records are the items in the database. For example, if a residential telephone book were a database table, the three fields might be defined: name, address, and telephone number. Information on each person in the telephone book would be stored in one record. Storing phone book entries in a computerized database table (as opposed to a printed book) allows you to quickly access the information by any field (whose telephone number is 555-3212?), print custom reports (print all people in the phone book with the last name Smith), and perform calculations (what percentage of people in the phone book live in zip code 65401?). Database tables are typically conceptualized as a spreadsheet in which the columns correspond to fields and the rows to records.
A water resources database table can contain hundreds of thousands of records and may require dozens of fields to fully characterize the data. With large databases it is important to break the data into smaller, more easily managed tables and then link them. For example, if 100,000 water quality readings are taken at a station located at "Medlock Bridge Road-SR 141-USGS Gage 02335000", it would be wasteful to repeat this long descriptive name for each of those 100,000 records. Instead, a station identifier is used (such as "CR0130") and a second file is created to list station identifiers and station names as well as other station descriptions. This is what it means for a database to be relational: multiple tables are used to efficiently store the data and relationships are formed to link the tables.
Multiple tables in a relational database are linked to each using key fields which are short identifiers pointing to other tables and sorted using index fields. If the United States were a database and each citizen were a record, your key field would probably be your social security number. That one nine-digit number uniquely identifies you and links you to numerous relational database tables (such as credit history, tax records, retirement income, etc.).
In summary then, environmental database design requires that we define the categories of data we want to collect (tables), the types of information within those categories (fields), and the relationships between the multiple tables (key and index fields). The two key design goals for the WRDB were 1) provide tables and fields to completely and accurately describe field and laboratory environmental data collected and/or used by EPD and 2) keep the design as simple as possible to minimize the effort needed to maintain the database; keeping it as small and fast as possible. The table structures and relationships were developed after reviewing many other laboratory and technical environmental database systems based upon a careful review of EPD’s needs. We believe the resulting system strikes a good balance between the two stated objectives.
See also Working and Master Table Overview.