[MITgcm-support] is it feasible to use tRef, sRef as initial conditions

Michael Spall mspall at whoi.edu
Wed Feb 26 12:39:59 EST 2014


I would also worry about initializing with a density gradient along a boundary
without a balanced flow through the boundary.  You will get large vertical
velocities next to the boundary, which can cause the model to blow up.

Mike

Sent from my iPad

> On Feb 26, 2014, at 7:29 AM, Dimitris Menemenlis <dmenemenlis at gmail.com> wrote:
> 
> Typically when we initialize model from a climatology, we have to use
> much shorter time step initially, e.g., first day or so, to reduce CFL shocks.
> As you ruhn the model check the "advcfl" diagnostics in the STDOUT
> files and make sure they remain below about 0.5.  Then you can gradually
> increase time step.  The CFL violation often occurs not because the
> initial fields are not sufficiently smooth in the horizontal, but rather because
> there are vertical instabilities in these fields, i.e., dense water over light water.
> 
>> On Feb 26, 2014, at 9:15 AM, 李志远 <oceanlizy at 163.com> wrote:
>> 
>> From: Li zhiyuan <oceanlizy at 163.com>
>> To: MITgcm <mitgcm-support at mitgcm.org>
>> Subject:"is it feasible to use tRef,sRef  as initial conditions "
>> my dear MITgcm users:
>>       There had been a question troubled me for a long time . I used mitgcm to simulate a wind_diriven circulation in an semi-enclosed domain ,such as " N24-N41,E117-E132" . When I use a pre-processed initial temperature and salt input file  (such as Levitus 01 datas) ,the model run some steps then occured "Nan" in the STDOUT file ,but when I use
>>  " tRef= 26.5, 25.5, 24.4, 23.1, 21.9,
>>        20.5, 18.7, 16.4, 13.6, 11.3,
>>        8.9,  6.5,  4.6,  4.6,  4.6,
>>        4.6,  4.6,  4.6,  4.6,  4.6,
>>        4.6,  4.6,  4.6,  4.6,  4.6,
>>        4.6,  4.6,  4.6,  4.6,  4.6,
>>        4.6,  4.6,  4.6,  4.6,  4.6,
>>       4.6,
>>  sRef= 33.28, 33.59, 33.87, 34.13, 34.36,
>>        34.53, 34.61, 34.64, 34.56, 34.48,
>>        34.39, 34.35, 34.38, 34.38, 34.38,
>>        34.38, 34.38, 34.38, 34.38, 34.38,
>>        34.38, 34.38, 34.38, 34.38, 34.38,
>>        34.38, 34.38, 34.38, 34.38, 34.38,
>>        34.38, 34.38, 34.38, 34.38, 34.38,
>>       34.38,
>>  the model run without errors . I have learnd that  this maybe a CFL problem , because  the initial conditions are not  smooth enough  . But I have smooth the data for many times , I think it is smooth. But things went wrong !!! I do not know  how to make the initial conditions "enough smooth" . 
>> So ,I doubt that  is it feasible for me to just use  the above " tRef ,sRef " as a initial conditions .  I want to force the model with wind force,  surface heatflux and salt flux  for few years until the model run steadily ,when I can get a steady state  " temperature , salt filed ", and 
>> the" u,v velocity filed" . I guess if feasible , it will take  a much longer time to get such a steady state .  Am  I  right ?
>> 
>>  Any  quick reply will be great appreciated !
>> Yours  Li zhiyuan
>> Best regards  ! 
> 
> _______________________________________________
> 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/20140226/1cfeadff/attachment-0001.htm>


More information about the MITgcm-support mailing list