[MITgcm-support] Set a density distribution

[fancer fancer]

Ok, as I understood from source code, rhoLoc is the density anomaly. Right?

Sincerely,
Serge Semin

On Sat, Apr 23, 2011 at 5:51 PM, fancer fancer <fancer.lancer at gmail.com>wrote:

> Hi, guys.
>
> One more question about density and buoyancy.
> I've find in the source code the next formulas:
> *find_rho.f:*
>
> ...
> refTemp=tRef(kRef)
> refSalt=sRef(kRef)
> dRho = rhoNil-rhoConst
> ...
> rhoLoc(i,j) = rhoNil*(sBeta*(sFld(i,j)-refSalt) -
> tAlpha*(tFld(i,j)-refTemp) ) + dRho
> ...
>
> *calc_buoyance.f*
>
> ...
> buoy(i,j) = -gravity*recip_rhoConst*rholoc(i,j)
> ...
>
> *ini_parms.F*
>
> ...
> recip_rhoConst = 1. _d 0 / rhoConst
> ...
>
> I don't set tRef, therefore refTemp = 0.0.
> I set sBeta = 0.0, rhoNil = rhoConst = 1000.0, therefore dRho = 0.0.
> Now I get:
> rhoLoc = - rhoNil*tAlpha*tFld(i,j)
> buoy(i,j) = -gravity*recip_rhoConst*rholoc(i,j)
>
> What's wrong in my calculations? Or may be something wrong in source code.
> Is rhoLoc a density or something other?
> I've found the next mailing-list:
> http://mitgcm.org/pipermail/mitgcm-support/2006-December/004492.html
> And I am sure that density should be:
>
> rhoLoc(i,j) = rhoNil*(*1 + *sBeta*(sFld(i,j)-refSalt) -
> tAlpha*(tFld(i,j)-refTemp) ) + dRho
>
> Could you please answer me where I wrong?
>
> Sincerely,
> Serge Semin
>
> On Tue, Apr 19, 2011 at 11:22 PM, fancer fancer <fancer.lancer at gmail.com>wrote:
>
>> Matt, Ryan
>>
>> Thanks you very much.
>> This is a very useful approach for me! Now I testing it.
>>
>> Good luck you in your researches.
>> Sincerely,
>> Serge Semin
>>
>>
>> On Tue, Apr 19, 2011 at 1:45 AM, Ryan Abernathey <rpa at mit.edu> wrote:
>>
>>> Serge,
>>>
>>> What Matt described is what many of us do in idealized setups where we
>>> don't care about T & S individually but only about buoyancy in general.
>>> Specifically, I use the following options in my data file.
>>>
>>> eosType='LINEAR',
>>> tAlpha=2.0E-04,
>>> sBeta=0.,
>>> saltStepping = .false.,
>>>
>>> Then the buoyancy is defined as
>>> gravity * tAlpha * THETA
>>>
>>> You can set an initial buoyancy just by setting an initial temperature.
>>> In your data file, this would look like
>>>  &PARM05
>>>  hydrogThetaFile='myHydrogThetaFile.bin',
>>>  &
>>>
>>> -Ryan
>>>
>>> On Apr 18, 2011, at 5:35 PM, Matthew Mazloff wrote:
>>>
>>> Hi Serge,
>>>
>>> You cannot just set density -- as you stated, T and S are the state
>>> variables used by the MITgcm.  If you are running an idealized experiment,
>>> however, you can just turn off salinity stepping (saltStepping = .false., in
>>> data PARM01)  and just use T making your equation of state linear such that
>>> T and rho will be analogous.
>>>
>>> -Matt
>>>
>>>
>>> On Apr 18, 2011, at 2:06 PM, fancer fancer wrote:
>>>
>>> Matt,
>>>
>>> I mean dynamics density, but for initial conditions.
>>> I mean can I set a matrix with 3D density profile of my area, instead of
>>> setting T and S at initial time?
>>> If I understand your answer properly, I can not.
>>> I looked the equations of model and understood what it is impossible.
>>> Every time calculates T and S  through the special equations. Then
>>> density calculates through EOS and desity goes to equation of motion.
>>> MITgcm do not have a equation for  evolution of density. But if it has,
>>> it can be possible. Right?
>>>
>>> Sincerely,
>>> Serge Semin
>>>
>>>
>>> On Tue, Apr 19, 2011 at 12:34 AM, Matthew Mazloff <mmazloff at ucsd.edu>wrote:
>>>
>>>> Hi Serge,
>>>>
>>>> Do you mean reference density or dynamics density.
>>>>
>>>> Dynamic density is not  a model input, but instead calculated from T and
>>>> S using one of the equation of state (EOS) options.  Some of the EOS have
>>>> matlab equivalents in
>>>> http://mitgcm.org/viewvc/MITgcm/MITgcm/utils/matlab/
>>>>
>>>> Reference density is taken as a constant, and should not be spatially
>>>> varying....
>>>>
>>>> -Matt
>>>>
>>>>
>>>>
>>>> On Apr 18, 2011, at 12:18 PM, fancer fancer wrote:
>>>>
>>>> Hi, MITgcm support.
>>>>
>>>> How can I set a density distribution for my computational area?
>>>> I know about rhoRefFile input parameter, but this is a 1D vector, where
>>>> each element for one vertical layer.
>>>> I have a 3D matrix with density profile for my area.
>>>> Is there are some special input parameters or method for integrating it
>>>> into computational area?
>>>>
>>>> In advance thanks for help.
>>>>
>>>> Sincerely,
>>>> Serge Semin
>>>> <ATT00001.txt>
>>>>
>>>>
>>>>
>>>> _______________________________________________
>>>> MITgcm-support mailing list
>>>> MITgcm-support at mitgcm.org
>>>> http://mitgcm.org/mailman/listinfo/mitgcm-support
>>>>
>>>>
>>> <ATT00001.txt>
>>>
>>>
>>> _______________________________________________
>>> MITgcm-support mailing list
>>> MITgcm-support at mitgcm.org
>>> http://mitgcm.org/mailman/listinfo/mitgcm-support
>>>
>>>
>>>
>>> _______________________________________________
>>> MITgcm-support mailing list
>>> MITgcm-support at mitgcm.org
>>> http://mitgcm.org/mailman/listinfo/mitgcm-support
>>>
>>>
>>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mitgcm.org/pipermail/mitgcm-support/attachments/20110425/e1b49d53/attachment-0001.htm>