"The purpose of the GNIS data base development was to promote geographic feature name standardization and to serve as an official repository of domestic geographic names information. The use of GNIS information varies, and has a wide range of applications. The broad categories of use include emergency preparedness, local transportation planning, regional planning, product marketing, site selection and analysis, genealogical research, and toponymic problem solving requiring the use and analysis of geographic names."
"Please acknowledge USGS. The Geographic Names Information System was developed by the U.S. Geological Survey in cooperation with the U.S. Board on Geographic Names."
"A random sample of 10% of the entries in the system were visually verified against the compilation source data (large-scale USGS topographic maps) to ensure an accuracy rate of at least 95%."
"Locative references (geographic coordinates, topographic map, and county) are cross-checked for logical consistency."
"Accuracy of these digital data is based upon the use of source graphics which are compiled to meet National Map Accuracy Standards. Comparison to the graphic source is used as control to assess digital positional accuracy."
"GNIS was originally developed on a mainframe computer under General Information Processing System (GIPSY), a text processor developed by the University of Oklahoma. The system was comprised of three databases: the National Geographic Names Database (NGNDB), the Topographic Map Names Database (TMNDB), and the Reference Database (RDB). The largest of the three databases, the NGNDB, contained the feature name records for nearly 2 million geographic features. The NGNDB was organized by State, with separate and distinct files being maintained for each State or territory. The TMNDB was an inventory of the USGS topographic maps used in compiling Phase I of the NGNDB, and the RDB contained annotated bibliographies of the sources used during Phase II compilation.
Data compilation for the GNIS NGNDB database began in 1968 when geographic names data for Massachusetts were collected and digitally archived. In 1976, geographic names data for Kansas and Colorado were collected to determine the feasibility of compilation on a national scale. After analysis and a favorable evaluation of this pilot project, geographic names data for the remaining States and territories were compiled from 1978 to 1982. The initial compilation, known as Phase I, is now complete, and the system contains entries for named features shown on USGS topographic maps with the exception of roads and highways. Large-scale maps were used during Phase I (i.e., the majority of the names were compiled from 1:24,000-scale, 7.5-minute USGS topographic maps). The 1:62,500-scale maps or the 1:250,000-scale maps were used when no 1:24,000-scale maps existed.
After Phase I was complete, the geographic names records were compared against the BGN's accumulated records. Some variant names and other BGN related data were added at this time. In 1993, an effort to redesign GNIS using Oracle was initiated. The TMNDB was the first database to be converted to Oracle and was renamed the Geographic Cell Names Database (GCNDB). The GCNDB has moved from an inventory of published maps to a list of approved names for spatially referenced USGS National Mapping Division (NMD) products. Following the conversion of the TMNDB, the NGNDB was converted to Oracle. Individual State and territory files were merged into one Oracle table. The RDB was incorporated into Oracle as a separate table. The Oracle version of GNIS became the official database management system in 1995. The GCNDB serves the GNIS, NMD data production systems and NMD product generation systems. Phase 1 entailed the collection of feature names from USGS large-scale topographic maps, U.S. Forest Service maps, Office of Coast Survey charts, Federal Aviation Administration files, Federal Communications Commission files, and files of the Army Corps of Engineers. These data were compared to the records of the U.S. Board on Geographic Names. Accuracy of these digital data was based upon the use of source graphics which are compiled to meet National Map Accuracy Standards. Comparison to the graphic source was used as control to assess digital positional accuracy.
During this phase II, geographic names and associated information are collected from official State and local maps and documents. Also, a wide variety of other approved sources, including historical documents, are used to complete the data collection. The Geographic Names Information System is resident on a Sun Enterprise 4000 computer. It uses Sun Solaris Version 2.6 as its operating system. The system contains approximately 1.5 million records, averaging 500 bytes in length."