Benchmark Problem 1:  description

Manning n=0.000:

Manning n=0.010:

Manning n=0.015:
Gages, Cliffs

Gages, Cliffs vs data

Gages, Cliffs
Gages, Cliffs-2

Gages, Cliffs-2 vs data

Vorticity&Currents (QuickTime mov)

Vorticity&Currents (QuickTime mov)

Vorticity&Currents (QuickTime mov)


  • The constant flow (u=0.115 m/s) state is used to define incoming characteristics on the left and right boundaries, and to initialise the domain. Manning friction coefficient is uniform over the domain.
  • At the beginning of the simulations, the flow is symmetric about the centerline. The higher the friction, the longer it takes for the Karman vortex street to develop.
  • The variation range and vorticity values are matched well, better with Cliffs-2. The major mis-fit is in the mean u-velocity on the centerline (gage 1): modeled mean velocity is smaller than observed (similar result in Figs. 6, 8 in Lloyd&Stansby II)
  • Cliffs differs from NOAA's MOST in land-water interface, which is not essensial for this particular benchmark. Cliffs-2 (to be released later) differs from Cliffs (and MOST) in the difference scheme over varing topography. Cliffs-2 yields somewhat more regular vortex pattern at low friction, and indentical (with Cliffs) results at higher friction.
  • Benchmark Problem 2

    Benchmark Problem 3

    Benchmark Problem 4

    Benchmark Problem 5