National Weather Service Interactive River Forecasting Using State, Parameter, and Data Modifications

Adams, Thomas; Smith, George; ,

National Weather Service Interactive River Forecasting Using State, Parameter, and Data Modifications

by Thomas Adams, Hydrologic Research Lab, Silver Spring, United States,
George Smith, Hydrologic Research Lab, Silver Spring, United States,

Document Type: Proceeding Paper

Part of: Engineering Hydrology

Abstract:

Interactive flood and river flow forecasting using the National Weather Service Interactive Forecast Program (IFP) is outlined. The IFP is an extension of the National Weather Service River Forecast System (NWSRFS), which has been used routinely in operational settings at NWS River Forecast Centers (RFCs) throughout the U.S., essentially in its existing form, since the mid-1980s. Processing of the many hydrologic models and support routines that comprise NWSRFS have been made on mini- and mainframe computers in batch mode from NWS RFCs as remote job entries. The hydrologic component of NWSRFS has been ported from the mainframe environment to run on scientific workstations in a Unix and X Window System environment, this constitutes the NWSRFS-IFP. The features of the IFP: local processing, increased processing speed, improved visual display and output within a graphical user interface (GUI), and ease of use should significantly improve the forecaster's ability to visualize river conditions and detect errors in model simulations. The IFP will allow forecasters to quickly make state and parameter changes and modify input data to improve agreement between observed and simulated flows, and, consequently, improve forecasts. Individual models are calibrated for each basin within the RFC drainage system area of responsibility during an initial calibration and NWSRFS system definition phase. Subsequently, during forecast periods, forecasters make state and parameter adjustments to models and modify input data in response to random errors manifested in simulation results that are due to imperfect model calibrations, errors resulting from model simplifications of complex physical processes, and space-time averaging and estimation of rainfall and other environmental variables.

Return to search