Author(s): Audrey L. Ishii; James E. Wilder
Linked Author(s):
Keywords: No Keywords
Abstract: Simulation of one-dimensional, unsteady, subcritical flow in open channels requires the specification of upstream and downstream boundary conditions. The selection of data used to specify the boundary conditions in combination with the selection of data used within the river reach to check the model calibration can affect the assessment of the best value for the roughness coefficient, Manning's n. A test of the sensitivity of the calibration of a one-dimensional, unsteady-flow model to boundary-condition data selection was performed utilizing three different combinations of boundary conditions for a 23.0-kilometer-long reach of the Fox River in northeastern Illinois, USA. It was found that the use of more than one combination of boundary-condition data increases the likelihood that the calibration process will identify the best overall roughness coefficient because using different types of boundary data adds information to the internal calibration check. In particular, river-stage results are most sensitive to Manning's n when discharge is used as one boundary condition, and discharge results are most sensitive to Manning's n when stage data are used for both upstream and downstream boundary conditions. The simulated results using the different boundary conditions were in close agreement with each other for only one value of Manning's n, and these results also matched the collected data best.
Year: 1993