[MITgcm-support] Re: question about seaice coupling to

[Martin Losch]

Dimitris,

On Monday, May 3, 2004, at 06:20 PM, Dimitris Menemenlis wrote:

>> old topic, but what about using some estimate of the Ekman layer depth
>> as the depth for the proxy-geostrophic velocities? The Ekman depth,
>> where according to theory the Ekman spiral has reversed the velocity
>> direction, is sqrt(A_v/(.5*f)). For A_v=1.e-3 (values from the
>> global_ocean.90x40x15/global_ocean.cs32x15), this ranges ranges from 
>> 4m
>> in high latitudes to 50m near the equator. Below this depth the wind
>> driven velocity is 36% of the surface value, below twice  this depth 
>> it
>> is 14% etc (you know what exponential decay looks like (o: ). So if 
>> you
>> use something like 4 or 5 times the Ekman depth, you are well below 
>> the
>> Ekman layer (but not necessarily below the mixed layer!).
>
> Martin, do you want to experiment with this and maybe add it
> as a runtime option for initialization of pkg/seaice?
This would be very interesting, and indeed, I will think about this. 
I'll also ask around among our sea-ice modellers here, and how they do 
they coupling (frankly, I don't think the do any coupling so far).
>
> The other option of course is to compute surface geostrophic
> velocity explicitly from eta.
As soon as loading of sea-ice is included, this may not work? I don't 
know...
>
> The whole issue of ice-ocean drag and ice-air drag is treated
> in a very primitive way right now.  For example, presence of ice
> is ignored for momentum flux to the ocean.
>
> There are many improvements that one could think about
> But it would be good if this is done within the context of
> detailed model/data comparisons and/or data assimilation.
Agreed!

Martin