[MITgcm-devel] vectorizing seaice and exf

[Martin Losch]

Yes, that should work. But the question was rather, do we want that?  
I remember, that Jean-Michel had a problem with such a solution, but  
I am be completely wrong.

Martin

On 29 Sep 2007, at 00:45, Patrick Heimbach wrote:

>
> Hi Martin,
>
> if I remember correctly,
> the safeguard in case of #ifdef ALLOW_ATM_WIND
> was to define
>            sh(i,j,bi,bj) = MAX(wspeed(i,j,bi,bj),uMin)
> so as to avoid vanishing ustar:
>            ustar(i,j)  = rdn(i,j)*sh(i,j,bi,bj)
>
> The analogue would be to define a wStress pendant
> that is always nonzero, and use this one instead of
>            tau(i,j)    = wStress(i,j,bi,bj)
>            ustar(i,j)  = sqrt(tau(i,j)/atmrho)
>
> Is this OK or does this give rise to spurious motions?
> (case wstress == 0 should be highly unlikely in open ocean)
>
> -p.
>
> On Sep 28, 2007, at 3:51 AM, Martin Losch wrote:
>
>> Hi,
>>
>> I just checked in my new exf_bulkformulae.F
>> Patrick, please check, if it does not break the adjoint.
>>
>> One issue remains (nothing to do with vectorization):
>> Jean-Michel, I remember that we talked about this already, but I  
>> forgot the solution. For the case that I want to read wind stress  
>> along with ALLOW_BULKFORMULAE (#undef ALLOW_ATM_WIND), the field  
>> wStress is computed from input-fields sqrt(ustress^2+vstress^2).  
>> If stress=0 (for some reason), ustar=0 in exf_bulkformulae  
>> resulting in a division by zero. Do we want to safeguard against  
>> that? And if so, how? Add a minimum wstress or ustar?
>>
>> Martin
>>
>> On 27 Sep 2007, at 22:47, Patrick Heimbach wrote:
>>
>>> Martin,
>>>
>>> great news.
>>> I reckon you are confident that the
>>>     IF (atemp ... = 0) THEN
>>> doesn't cause additional problems...
>>>
>>> Re. routine, I had a quick look, and it seems ok.
>>> One thing is that the store directives for
>>> rdn, ustar, qstar, tstar, tau
>>> are now broken.
>>> Could you add "(i,j)" to these?
>>>
>>> I guess you'll run testreport before checkin(?)
>>>
>>> Cheers
>>> -Patrick
>>
>> _______________________________________________
>> MITgcm-devel mailing list
>> MITgcm-devel at mitgcm.org
>> http://mitgcm.org/mailman/listinfo/mitgcm-devel
>
> ---
> Patrick Heimbach | heimbach at mit.edu | http://www.mit.edu/~heimbach
> MIT | EAPS 54-1518 | 77 Massachusetts Ave | Cambridge MA 02139 USA
> FON +1-617-253-5259 | FAX +1-617-253-4464 | SKYPE patrick.heimbach
>
>
> _______________________________________________
> MITgcm-devel mailing list
> MITgcm-devel at mitgcm.org
> http://mitgcm.org/mailman/listinfo/mitgcm-devel