Calibration of Unsteady HEC-RAS Model for Complex Overflowing Channels – CVFED Sacramento River Model nearthe Flood Relief Structures

Speaker: Chakri Malakpet, PE, CH2M HILL, 2485 Natomas Park Dr, Sacramento, CA 95833. PH: 916 286 0435

Email:

Co-author: Alan Turner, PE, CFM, CH2M HILL, 9193 South Jamaica Street, Englewood, CO 80112. PH: 720 286 0844

Email:

Co-author: Mark Glidden, PE, CH2M HILL, 9193 South Jamaica Street, Englewood, CO 80112. PH: 720 286 5153

Email:

The California Department of Water Resources, as part of the CVFED program, askedCH2M HILL todevelop a comprehensiveunsteady HEC-RAS model of the Sacramento River system upstream of theFremont Weir. This is part of an extensive flood control systemconsisting of a complex network of levees, weirs, flood relief structures and bypassesto protect the Central Valley and City of Sacramento from flood flows.

To represent the complex system hydraulics, an unsteady HEC-RAS model was calibrated to observed flow, stage data and surveyed high watermarks at key locations in the system. The complexity of the riverine model nearthe flood relief structures between the Hamilton City and Butte City gagespresented numerous challenges related to calibration. Between Hamilton City and Butte City gages, flood flows from the Sacramento River, through flood relief structures (3B’s, M&T and Goose Lake), enter into the Butte Sink. These flood flowsproduce similar water surface profilesin the Butte sink and the Sacramento River, creating dynamic flow exchange between the channels.

The challenge for calibration was that changes to flow and water surface in the Sacramento River impacted the overflow into the Butte Sink, altering the flow interactionand water surface in the channels. It became necessary to refine parameters to address differences between computed and recorded values for both flow and stage. To calibrate the model, modifications were performed to spatially varied Manning’s n values and lateral weir coefficients,in conjunction with addition of levee cards and ineffective flow areas in the main channel overbanksto capture complex natural hydraulic functions. This paper presents the methodology of calibration, sensitivity of the system to system parameters and performance of the model for different observed events. The process used for the calibration and lessons learned will be discussed to benefit other modelers.