sediment to a sink

Sediment modeling collaborators: issues, applications, information exchange

Moderators: arango, robertson, rsignell

Post Reply
Message
Author
antoinemarie
Posts: 52
Joined: Mon Mar 07, 2011 5:37 pm
Location: HYDRATEC

sediment to a sink

#1 Post by antoinemarie » Tue Jul 16, 2013 3:05 pm

Hi ROMS users,

I try to create a sink with a netcdf forcing file. How can I include sediment in a sink? Because my sediment concentration varies with time. So I can not set a fixed value (0 for instance), otherwise the concentration blow-up.

Thank you in advance for your answers :)

Antoine

User avatar
kate
Posts: 3700
Joined: Wed Jul 02, 2003 5:29 pm
Location: IMS/UAF, USA

Re: sediment to a sink

#2 Post by kate » Tue Jul 16, 2013 4:59 pm

For the sources and sinks, there is an LtracerSrc with a value for each active and passive tracer. In the sediment.in, there are MUD_Ltsrc and SAND_Ltsrc for the sediment tracers.

Sinks behave better if you don't set any of the tracers on them, while sources behave better if you do set tracer values. If you only have sinks, set the switches all to false. If I had both, I'd probably hack the code for pointwise control.

antoinemarie
Posts: 52
Joined: Mon Mar 07, 2011 5:37 pm
Location: HYDRATEC

Re: sediment to a sink

#3 Post by antoinemarie » Wed Jul 17, 2013 7:01 am

Thank you Kate for your answer

I have both source and sink, how can I hack the code? Moreover, I have the same problem with the temperature of my sink.

User avatar
kate
Posts: 3700
Joined: Wed Jul 02, 2003 5:29 pm
Location: IMS/UAF, USA

Re: sediment to a sink

#4 Post by kate » Wed Jul 17, 2013 4:50 pm

antoinemarie wrote:Moreover, I have the same problem with the temperature of my sink.
Of course you do. All sources without tracer values are using an unconditionally unstable downwind advection scheme. Ditto sinks with specified tracer values.
I have both source and sink, how can I hack the code?
First of all, search for the point source code in ROMS and figure out what's doing what.

1. I would add a logical array, dimensioned by the number of sources. Add the code to read it from ana_psource.h and/or from the sources file.
2. Use it per source instead of using LtracerSrc.

Note that you'll need to provide tracer values for all source/sinks for all tracers, but you'll only be using some of them.

antoinemarie
Posts: 52
Joined: Mon Mar 07, 2011 5:37 pm
Location: HYDRATEC

Re: sediment to a sink

#5 Post by antoinemarie » Wed Jul 24, 2013 7:40 am

Thank you Kate for your advice, but it seems a little bit complicated for me, so I try another way.

Eventually, I have defined an open boundary with radiative conditions in my closed model, then I extend a row of my grid to this boundary to create a sink. But the volume conservation is not respected. I will try another simulation with a clamped conditions for zeta on the open boundary in order to force my model to keep his initial level.

Post Reply