[MITgcm-devel] seaice

[Martin Losch]

Hi Dimitris,
as a matter of fact I use exf_bulkformulae, and thus CORE-winds in  
two cases: one with seaice, one without. In a third case I use  
trenberth wind stress and ncep heat fluxes. The circulation (ACC) is  
reasonable with trenberth and with CORE-winds (with bulkformulae)  
without seaice, but bad with CORE-winds and seaice.
You are probably right about the strength of the ncep winds, but I  
guess this is a problem one order of magnitude smaller than mine.

Martin

On Feb 11, 2006, at 10:30 PM, Dimitris Menemenlis wrote:

> Martin, this is just a very wild guess but the big difference  
> between the integrations that have realistic ACC and those that do  
> not in your message seems to be the usage of NCEP wind stress vs  
> NCEP 10-m winds, respectively.  Am I correct?
>
> From what I have seen (and I still owe a couple of plots to Andrea  
> on this topic) is that NCEP wind stress is much more realistic than  
> NCEP 10-m winds in coastal regions.  Over land, NCEP appears to  
> have very large drag coefficient, so that 10-m winds are very  
> small, even though wind stress remains about the same magnitude.   
> These weak winds leak over to coastal regions near land, due to  
> coarseness of NCEP reanalysis.
>
> Dimitris
> _______________________________________________
> MITgcm-devel mailing list
> MITgcm-devel at mitgcm.org
> http://mitgcm.org/mailman/listinfo/mitgcm-devel