Modeling surface trapped river plumes: A sensitivity study

Proceedings of the International Conference on Estuarine and Coastal Modeling



  • The Publications Warehouse does not have links to digital versions of this publication at this time
  • Download citation as: RIS | Dublin Core


To better understand the requirements for realistic regional simulation of river plumes in the Gulf of Maine, we test the sensitivity of the Blumberg-Mellor hydrodynamic model to choice of advection scheme, grid resolution, and wind, using idealized geometry and forcing. The test case discharges 1500 m3/s of fresh water into a uniform 32 psu ocean along a straight shelf at 43?? north. The water depth is 15 m at the coast and increases linearly to 190 m at a distance 100 km offshore. Constant discharge runs are conducted in the presence of ambient alongshore current and with and without periodic alongshore wind forcing. Advection methods tested are CENTRAL, UPWIND, the standard Smolarkiewicz MPDATA and a recursive MPDATA scheme. For the no-wind runs, the UPWIND advection scheme performs poorly for grid resolutions typically used in regional simulations (grid spacing of 1-2 km, comparable to or slightly less than the internal Rossby radius, and vertical resolution of 10% of the water column), damping out much of the plume structure. The CENTRAL difference scheme also has problems when wind forcing is neglected, and generates too much structure, shedding eddies of numerical origin. When a weak 5 cm/s ambient current is present in the no-wind case, both the CENTRAL and standard MPDATA schemes produce a false fresh- and dense-water source just upstream of the river inflow due to a standing two-grid length oscillation in the salinity field. The recursive MPDATA scheme completely eliminates the false dense water source, and produces results closest to the grid-converged solution. The results are shown to be very sensitive to vertical grid resolution, and the presence of wind forcing dramatically changes the nature of the plume simulations. The implication of these idealized tests for realistic simulations is discussed, as well as ramifications on previous studies of idealized plume models.

Additional publication details

Publication type:
Publication Subtype:
Journal Article
Modeling surface trapped river plumes: A sensitivity study
Series title:
Proceedings of the International Conference on Estuarine and Coastal Modeling
Year Published:
Publisher location:
Reston, VA, United States
Contributing office(s):
Woods Hole Coastal and Marine Science Center
14 p.
Larger Work Type:
Larger Work Subtype:
Journal Article
Larger Work Title:
Proceedings of the International Conference on Estuarine and Coastal Modeling
First page:
Last page:
Conference Title:
Proceedings of the 6th International Conference on Estuarine Coastal Modeling
Conference Location:
New Orleans, LA, USA
Conference Date:
3 November 1999 through 5 November 1999
United States
Other Geospatial:
Gulf of Maine