surface mass sources (but not rivers)

Discussion of how to use ROMS on different regional and basin scale applications.

Moderators: arango, robertson

Post Reply
Message
Author
jklinck
Posts: 32
Joined: Mon Jun 30, 2003 2:29 pm
Location: CCPO/ODU, USA

surface mass sources (but not rivers)

#1 Post by jklinck » Thu Oct 08, 2015 7:04 pm

I would like to use roms to demonstrate the abyssal circulation in a homogeneous density model by applying surface mass sources and sinks. Is there a way to add mass to change d zeta/dt?

ana_psource seems to be designed as a lateral mass source (rivers) which also adds momentum along with mass. I am not sure how to add a localized mass source and have a distributed mass sink. I could add a mass source/sink at every grid point but this seems a bit clumsy. Especially since the vertical distribution of the mass source is also required.

Has anyone tried to add mass through the surface (adding rain but not just as a salt source)? I would be interested in hearing suggestions.

Thanks,

John

User avatar
kate
Posts: 3757
Joined: Wed Jul 02, 2003 5:29 pm
Location: IMS/UAF, USA

Re: surface mass sources (but not rivers)

#2 Post by kate » Thu Oct 08, 2015 8:01 pm

Ray-Qing Lin was working on that roughly 15-20 years ago and had a lot to say about it. She says it was easy to get wrong. I never saw the code.

User avatar
wilkin
Posts: 513
Joined: Mon Apr 28, 2003 5:44 pm
Location: Rutgers University
Contact:

Re: surface mass sources (but not rivers)

#3 Post by wilkin » Fri Oct 09, 2015 12:05 pm

Look at the LwSrc option. I think that does what you want. grep through the code to see what it does, but just be aware that it interacts with it's counterpart option LuvSrc and a logical negation of LuvSrc can be relevant.

You'll see in omega.F that the massflux of the sources impacts w, which is what you want.
Be sure sure you make LuvSrc = false or you get partially doubled sources.

The LwSrc method applies the source directly to the cell centered flux divergence, whereas LuvSrc adds the fluxes to cell faces.
John Wilkin: DMCS Rutgers University
71 Dudley Rd, New Brunswick, NJ 08901-8521, USA. ph: 609-630-0559 jwilkin@rutgers.edu

jklinck
Posts: 32
Joined: Mon Jun 30, 2003 2:29 pm
Location: CCPO/ODU, USA

Re: surface mass sources (but not rivers)

#4 Post by jklinck » Thu Oct 15, 2015 5:02 pm

I was able to set up the source/sink flow based on John's suggestion (Thanks!). I can get a linear abyssal circulation driven by surface sources and sinks using LwSrc=true and LuvSrc=false, and using ana_psource.h to control sources/sinks.

The model limits the number of sources to 200. I have a 100x150 grid. I tried to have 10 source points and 14990 sink points but the model failed (I then found out about the 200 source limit). Rather than increase the number of sources (in mod_sources.h), I set up a distributed set of sinks every 10 grid points (150 sinks) and 10 source points. The distributed sinks cause some variability in SSH and ubar,vbar, but the model behaves reasonably well. I could get a smoother solution to have sinks every 5 points, but I did not take the time to do this.

Anyway, it is possible to coerce ROMS to solve abyssal circulation (shallow water model) driven by sources and sinks.

Post Reply