Close Window

Hydrologic Research Center: "Improvement of hydrologic model forecasts using a real time updating technique"

Final Report

The Sacramento Soil Moisture Accounting Model is used by US NWS River Forecast Centers (RFCs) to generate streamflow estimates for a wide variety of hydrologic products. These products range from flow forecasts in major river basins to headwater basin forecasts. The Office of Hydrology has automated the forecast procedure through the development and national implementation of the NWS River Forecast System (NWSRFS) software package.

The ability to adjust Sacramento Model state variables using the observations of discharge is a key element in improving the accuracy of the streamflow forecasts and in providing the capability for predicting forecast uncertainty. Georgakakos, et al. (1988) developed the theory and evaluated applications of the Hydrologic Forecast System (HFS) which combines a non-linear state estimator with the Sacramento Model for realtime updating from discharge observations.

The unique feature of the HFS is that it is based on a new robust design that explicitly associates state estimator parameters with hydrologic model physics and a priori degree-of-belief estimates of input and parameter uncertainty. Objectives of the project were to 1) facilitate the implementation of HFS onto the California-Nevada RFC (CNRFC) computers, 2) to provide training to CNRFC hydrologists in the theory and use of the HFS, and 3) to provide HRC scientists with insight into the issues associated with HFS implementation in California basins from the perspective of an operational hydrologist with experience in the calibration and use of the Sacramento Model.