Using Hydrologic Data to Forecast Floods

By Charles W. Boning and Eugene A. Stallings

Floods are dangerous, life-threatening, and destructive. Accounts of floods and flood disasters as well as droughts and famine appear in numerous biblical writings. The Romans used dams and diversions in attempting to reduce potential floods and to manage water resources. Concerns about floods have existed in the United States since the first settlers came to the New World. Early efforts to reduce flood-related deaths and damages, however, were primarily devoted to flood control measures such as levees, dams, and storage reservoirs. Flood-warning systems would not be feasible until suitable communication systems were developed.

Early flood-warning systems for people living along streams undoubtedly involved personal travel and verbal exchange of information. Communication systems such as the telephone greatly improved timeliness of flood warnings. Federal, State, and (or) local water-management agencies began to remotely access data from USGS gaging stations in the 1930's by using an instrument called `Telemark.' When accessed by telephone, the Telemark transmitted river stage by a series of beeps or rings. Very high frequency radio also began to be used in the 1930's to obtain river-stage data. By the 1950's, river stage at hundreds of USGS gages throughout the United States could be accessed to forecast floods, provide flood warnings, or for water-management purposes including drought.

The continued evolution of communication and stage-sensing equipment has further improved data access for flood-warning and flood-forecasting purposes. The principal devices currently in use in USGS gages for obtaining near realtime data for flood warning or other water-management purposes are those that transmit data via satellite to a receiving station where the data are then relayed to equipped with conventional radio and telephone systems, which contribute to the timeliness of data critical to warning or water-management purposes.

The initial use of remotely accessed data to forecast floods was often limited to the interpretation of correlative relations that existed between different streams or stream locations, which required a great deal of intuitive expertise and knowledge on the part of the river or weather forecaster. The rapid development of more sophisticated computer systems since the 1970's, however, has permitted large amounts of data to be incorporated into computer-simulation models for more accurate and timely forecasting. Although data from USGS gages are used for a wide variety of water-management purposes, a most critical use of such data is to provide warnings or forecasts of floods.

The National Weather Service (NWS) of the National Oceanic and Atmospheric Administration, is the Federal agency in charge of weather forecasts and warnings for the Nation. Many people, however, are not aware that the NWS is also charged by law with the responsibility to issue forecasts and warnings of floods. The mission of the NWS's Hydrologic Services program is to save lives and decrease property damage by issuing timely flood warnings and river-stage forecasts. Although many cities, counties, or other local flood-management agencies are involved in the operation of local flood-warning networks, the NWS, through its nationwide hydrologic-forecasting mandate, is the principal agency that uses nonstructural methods to decrease flood damage. In addition, NWS reviews hydrologic and meteorological data during periods of deficient rainfall and snowfall to make predictions of water-supply and to issue drought advisories

Flooding along major rivers that is caused by rainfall takes many hours and even weeks to develop. Floods caused by snowmelt runoff may take months to develop. Flash floods occur when intense precipitation falls during a brief time span on smaller rivers; the time between the onset of intense precipitation and the cresting of the river is hours instead of days. More than 10,000 precipitation and streamflow stations, including more than 3,000 streamflow gages operated by the USGS, provide hydrologic data to NWS offices across the country for use in the flood-forecasting program. Hydrologic data collection at stream-gaging stations is telemetered through the Geostationary Operational Environmental Satellite (GOES) to regional NWS River Forecast Centers (RFC). The data are incorporated into hydrologic models that are used in making hydrologic forecasts and advisories. The data provided by the USGS streamflow gages are essential in creating timely and more accurate hydrologic forecasts and advisories. Most of these data are archived over a long period of time. As a result, a large base of hydrometeorological data is available for research and development activities and for statistical analysis.

The NWS has 13 regional RFC's that serve as the principal mechanism for developing warnings against floods. Each RFC is responsible for warnings within at least one major river system. The staff of each RFC prepares river and flood forecasts and warnings, and provides guidance to a network of 52 National Weather Service Forecast Offices (WSFO). Each WSFO redistributes this information to users in cities, counties, and States (including Puerto Rico) within its area of responsibility.

A new era of river and flood forecasting is beginning. Expanded use of telemetry at the USGS streamflow stations and refinement of telemetry equipment continue to improve the timeliness and reliability of data that are transmitted for forecasting purposes. The NWS is beginning a major modernization of its forecasting system--the Next Generation Weather Radar--a doppler-based system that is replacing conventional radar. Automated surface observing systems are replacing manual weather observations, and advanced telecommunications systems are improving the integration and distribution of data. Modernization of the NWS forecasting systems will result in earlier and better warnings of severe thunderstorms, tornadoes, winter storms and hurricanes, general flooding, flash floods, and water supply.

Scientific understanding of the Earth's atmosphere and technology for data communication continue to improve, but the need for hydrometeorological information and data for weather and river forecasting will not diminish. A principal basis for constructing, verifying, and executing streamflow simulation models for river and flood forecasting purposes is the raw hydrologic data from USGS streamflow stations. Without such data, the NWS mandate to provide river flood forecasts for the protection of life and property would be greatly limited.