Rivers in nested grids

General scientific issues regarding ROMS

Moderators: arango, robertson

Post Reply
Message
Author
jpringle
Posts: 107
Joined: Sun Jul 27, 2003 6:49 pm
Location: UNH, USA

Rivers in nested grids

#1 Unread post by jpringle »

Dear all --

I am having some trouble with nesting and river inflows. I have justed started to implement rivers entering into my domain. Some rivers enter where the fine grid overlaps the coarse grid. When I run the model with nesting, it blows up quickly. When I run with rivers in just coarse grid (without nesting), or with nesting and no rivers, the model runs fine.

Right now, when I have the river enter in the fine grid, I do not have a river enter in the same location is the coarse grid. Is this correct? What is best practice in the case where a river enters where both grids overlap:

A) River in just the fine grid
B) River in both grids.

I know with two-way nesting the fine grid overwrites the coarse grid where they overlap. But the volume flux at the edge regions comes from the coarse grid, and because the river volume flux propogates through the model in the fast barotropic time-stepping, perhaps river input (or lack thereoff) in the coarse grid can affect non-overlap regions after only a single baroclinic timestep.

Anyway, I am interested in what people have done that works. Is there a best practice yet?

Thanks,
Jamie

User avatar
wilkin
Posts: 875
Joined: Mon Apr 28, 2003 5:44 pm
Location: Rutgers University
Contact:

Re: Rivers in nested grids

#2 Unread post by wilkin »

Jamie,

I have run nested grids with rivers but made a point of choosing my nest perimeter to avoid having a river source in the contact point regions. That just seemed to be asking for trouble because it might require special handling of the fluxes on those faces.

You are correct that in two-way nesting any rivers that enter the coarse grid within the region encompassed by the fine will be irrelevant. The solution in those cells gets overwritten with the fine-to-coarse average. This is true of the entire coarse grid solution.

We still satisfying the CFL condition for the fast barotropic mode in each grid so I don't see how this would allow barotropic signals to leap over the contact points region.

In my runs I've left the rivers in the coarse grid where the nest is - simply to avoid having to create a new rivers file when switching between nesting, and not.

With nesting on, if you have rivers in the coarse but not the fine I think you will see the river source disappears.

John.
John Wilkin: DMCS Rutgers University
71 Dudley Rd, New Brunswick, NJ 08901-8521, USA. ph: 609-630-0559 jwilkin@rutgers.edu

jpringle
Posts: 107
Joined: Sun Jul 27, 2003 6:49 pm
Location: UNH, USA

Re: Rivers in nested grids

#3 Unread post by jpringle »

While waiting for some other runs to complete, I did some simple test cases with nested grids, and can confirm that what John said is true.

Jamie

Post Reply