[MITgcm-support] Problems with heat flux

[Martin Losch]

Hello Renske,

I like your topography, it looks pretty in the ascii-output (o:

 From  your output file, everything seems to be OK (btw, with  
debugLevel>1 in data, namelist PARM01 you would have got a lot of the  
output that you now painfully programmed).
Are you sure that your surface flux is non-zero?

It looks like your setup has idealised geometry. If you are willing  
to send me your "code" directory, all you data*-files (including  
eedata), in other words the "input" directory, and a matlab script  
that generates the input files for topography and surface flux, I can  
offer to have a closer look.

Martin

On 23 May 2008, at 13:17, Renske Gelderloos wrote:

> Dear all,
>
> I'm trying to run a hydrostatic simulation with a heat flux at the  
> surface. As a basis I used exp2, and added a binary file for  
> surfQFile in input/data (and also changed the topography and number  
> of tiles). To keep things simple, the heat flux is uniform and not  
> time dependent.
>
> The problem is now that the binary file is read correctly by the  
> model, but it doesn't do anything with it. So the temperatures  
> after a couple of months is exactly the same as it was in the  
> initial conditions. Using some basic print statements I found out  
> that "gT" and "theta" are changed in "thermodynamics", "dynamics",  
> and "tracer_correction_step". I don't really understand how they  
> are changed and what goes wrong.
> Furthermore (I'm not sure whether or not this is the cause of the  
> malfunctioning of the heat flux in my case, or that I just don't  
> understand this part of the code), I think something strange  
> happens in the k-numbering in the "dynamics" part. In the loop  
> where the tendency is calculated (in mom_fluxform) the k is changed  
> in the last loop over j from k=1 to k=0, which doesn't exist. I  
> added the output.txt file for some reference and the extra print  
> output.
>
> Is there perhaps some extra switch I should turn on? Or is there  
> something else I should change?
>
> Thanks for any help!
>
> Renske Gelderloos