[To appear in Proceedings of the ASCE-EWRI World Water & Environmental Resources Congress, Salt Lake City, Utah, June 27 – July 1, 2004] Comparison of Estimates of Uncertainty of Discharge at U.S. Geological Survey Index-Velocity Gages on the Chicago Sanitary and Ship Canal, Illinois Thomas M. Over 1 , James J. Duncker 1 , and Juan A. Gonzalez-Castro 2 1 U.S. Geological Survey, Illinois District, 221 N. Broadway, Urbana, IL 61801; PH (217) 344-0037; FAX (217) 344-0082; email: tmover@usgs.gov or jduncker@usgs.gov 2 Hydrology and Hydraulics Division, South Florida Water Management District, 3301 Gun Club Road, West Palm Beach, FL 33406; email: jgonzal@sfwmd.gov Abstract Estimates of uncertainty of discharge at time scales from 5 minutes to 1 year were obtained for two index-velocity gages on the Chicago Sanitary and Ship Canal (CSSC), Ill., instrumented with acoustic velocity meters (AVMs). The velocity measurements obtained from the AVMs are corrected to a mean channel velocity by use of an index-velocity rating (IVR). The IVR is a regression-derived relation between the AVM velocity estimates and those obtained using acoustic Doppler current profilers (ADCPs). The uncertainty estimation method is based on the first- order variance method, but the AVM velocity error is estimated from an empirical perspective, using the statistics of the IVR regression. It is not clear whether to include the standard error of the IVR regression ( ) in the discharge uncertainty. At the 5-minute time scale when is included, has the dominant contribution to the discharge uncertainty, and the discharge uncertainty (expressed as the standard deviation of the discharge estimate) is about 5 m 2 ε σ 2 ε σ 2 ε σ 3 /s at one gage and 8 m 3 /s at the other, independent of discharge. When is not included, the discharge uncertainty at the 5-minute time scale is much smaller (about 0.5 m 2 ε σ 3 /s) and depends more strongly on discharge. For time scales 1 day or greater and when is not included, the uncertainty of the IVR parameters dominates the discharge uncertainty, and the value of the discharge uncertainty is about 0.4 m 2 ε σ 3 /s for one gage and 0.5 m 3 /s for the other gage. INTRODUCTION The total uncertainty of discharge measurements traditionally has been estimated as the square root of the summation of the squares of the total uncertainties from different sources (e.g., Carter and Anderson, 1963; Simpson and Oltman, 1993). In the International Organization for Standardization (ISO) 6416 standard (ISO, 1992),