[MITgcm-devel] Re: Large & Yeager 2004

Tue Dec 19 02:49:26 EST 2006

Hi there,

just to diversify this discussion:

Yesterday I had a chat with our Arctic specialists (Gerdes, Koeberle,  
Kauker, Karcher), and they use bulk formulae by Parkinson and  
Washington (1970ies). Part of this is in pkg/seaice/ 
seaice_budget_ocean/ice.F but only part of it. E.g. (they claim that)  
they don't use downward long wave radiation as an input field because  
the lw-fields are so bad in the Arctic, but instead estimate it from  
the bulk formulae (and I had a look at their code, they actually do  
that with some reference to SST). Apparently this is part of the  
AOMIP (Arctic Ocean Model Intercomparision Project). I find this hard  
to understand/believe, as I don't see how the ocean temperature has  
an influence on the LW-radiation emitted by the atmosphere. But maybe  
I am just narrow-minded? Andrea, you are my only hope of  
clarification ...

Also they use dew point temperature and relative humidity instead of  
specific humidity. It's all so strange ...

As for Large and Yeager: If you want to use the CORE data set, the  
use of the L&Y04 bulk formulae is strongly recommended (by Griffies  
et al in their short documentation on the CORE data). It appears that  
the parameters in bulk formulae are sort of arbitrary but have to  
match thoses used to generate the atmospheric fields.

Martin

On 19 Dec 2006, at 06:00, Andrea Molod wrote:

> hi patrick and martin,
>
> i usually don't put in my 2 cents here, but the ocean
> emissivity issue is usually handled on the atmosphere side
> by setting it to a wavelength-dependant set of numbers that
> are:
>
> 0.9788, 0.9833, 0.9819, 0.9820, 0.9835, 0.9865, 0.9886, 0.9719,  
> 0.9719, 0.9719, 0.9719, 0.9719
>
> for wave bands:
>    band 1:   4.5 -  5.3 um
>    band 2:   5.3 -  5.9 um
>    band 3:   5.9 -  7.1 um
>    band 4:   7.1 -  8.0 um
>    band 5:   8.0 -  9.1 um
>    band 6:   9.1 - 10.2 um
>    band 7:  10.2 - 12.5 um
>    band 8:  12.5 - 14.9 um
>    band 9:  14.9 - 18.5 um
>    band 10: 18.5 - 25.0 um
>    band 11: 25.0 - 35.7 um
>    band 12: 35.7 -  oo  um
>
> this is from a reference called Fu and Liou (i need to look it up....)
>
> so your 0.97 number sounds closer to this.....
>
>
> and we use 0.08 for open ocean albedo - not sure of the source of  
> this number.
>
> andrea
>
> On Mon, 18 Dec 2006, Patrick Heimbach wrote:
>
>>
>> Definitely yes.
>> The "bulk formulae" are almost identical to the Large & Pond ones.
>> It seems to be the "radiation formulae" that matter.
>> exf has default 0.97, Large & Yeager recommend 1.
>> It's a simple runtime parameter in data.exf:
>> ocean_emissivity = 1.,
>>
>>
>> On Dec 18, 2006, at 3:26 PM, Martin Losch wrote:
>>
>>> Patrick,
>>> I have just put a run with large+yeager bulk formulae into the  
>>> queue. Would it make more sense to stop that job and do, what you  
>>> suggest (ocean_emissivity = 1)?
>>> Martin
>>> On 18 Dec 2006, at 20:52, Patrick Heimbach wrote:
>>>> Dimitris,
>>>> an update on this.
>>>> Different people seem to understand different things
>>>> when they talk about "bulk formulae".
>>>> I undetstood the derivation of buoyancy and momentum fluxes
>>>> from atmos. state involving the parameterization of the neutral
>>>> transfer coefficients, their stability corrections and the like,
>>>> thus deriving, taux & tauy, sensible & latent heat fluxes,  
>>>> evaporation.
>>>> The treatment of radiation is somewhat separate, since completely
>>>> decoupled from these "bulk formulae".
>>>> However, in fact it seems that this minor detail might be
>>>> the only relevant part that distinguishes our current
>>>> exf implementation from the Large & Yeager 2004 recommendations,
>>>> i.e. they suggest ocean_emissivity to be set to 1,
>>>> whereas exf has a default of 0.97
>>>> (and bulkf_force a default of 0.985).
>>>> So I'll conduct another run with changed emissivity
>>>> (the 3% difference should in fact have a discernable impact).
>>>> -Patrick
>>>> On Dec 17, 2006, at 12:26 PM, Dimitris Menemenlis wrote:
>>>>> Patrick, thank you for info re bulk formulae.
>>>>> This is indeed very useful information.
>>>>> Dimitris
>>>> ---
>>>> Dr 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
>>> _______________________________________________
>>> MITgcm-devel mailing list
>>> MITgcm-devel at mitgcm.org
>>> http://mitgcm.org/mailman/listinfo/mitgcm-devel
>>
>> ---
>> Dr 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
> _______________________________________________
> MITgcm-devel mailing list
> MITgcm-devel at mitgcm.org
> http://mitgcm.org/mailman/listinfo/mitgcm-devel