Projecting Ice-Affected Streamflow by Extended Kalman Filtering

• Main Authors: Holtschlag, David J., Parker, Charles T., Grewal, Mohinder S.
• Other Authors: GEOLOGICAL SURVEY LANSING MI WATER RESOURCES DIV
• Format: Text
• Language: English
• Published: 1997
• Subjects: Hydrology; Limnology and Potamology; Snow; Ice and Permafrost; *ICE FORMATION; *STREAMFLOW ANALYSIS; ATMOSPHERIC TEMPERATURE; KALMAN FILTERING; ICE MECHANICS; CLIMATE; OPEN WATER; ICE FORECASTING; RIVER CURRENTS; PLATTE RIVER; SAINT JOHN RIVER; Ice; permafrost
• Online Access: http://www.dtic.mil/docs/citations/ADA335228; http://oai.dtic.mil/oai/oai?&verb=getRecord&metadataPrefix=html&identifier=ADA335228

An extended Kalman filter was developed to automate the real-time projection of ice-affected streamflow, based on routine measurements of stage and air temperature and the relation between stage and flow during open-water conditions. The form accommodates three dynamic modes of ice effects: sudden formation-ablation, stable ice conditions, and final elimination. The filter was applied to historical data from two long-term stream-flow-gaging stations. They were stable and parameters converged for both stations, producing estimates that were highly correlated with and linearly related to published streamflow values in a log-transformed metric. At St. John River at Dickey, Maine, logarithms of projected streamflow values were within 8% of the logarithms of published values 87.2% of the time and within 15% of published values 96.6% of the time during periods of ice effects. At Platte River at North Bend, Nebraska, logarithms of projected streamflow values were within 8% of the logarithms of published daily values 90.7% of the time and within 15%, 97.7% of the time during ice-affected conditions. This extended Kalman filter allows estimation of ice-affected streamflow at other gaging stations by adjusting filter parameters to site-specific conditions.