[MITgcm-devel] ICE workshop - recap and homework!

[Martin Losch]

Hi Hong,

yes, you are right, this factor was missing in exf_radiation.F leading to too much incoming long wave radiation. In a uncoupled set-up (with a prescribed atmosphere), this will probably balanced in part by the turbulent fluxes (by restoring to atemp). In seaice_solve4temp.F it was hard-coded.

M.

On Apr 11, 2012, at 6:45 PM, Hong Zhang wrote:

> Martin Losch wrote:
>> Hi Hong,
>> 
>> this is something I introduced. It really should be the default, because without this flag, you will not be able to get a consistent long wave radiation budget. See this thread:
>> <http://mitgcm.org/pipermail/mitgcm-devel/2011-December/005079.html>
>> especially:
>> <http://mitgcm.org/pipermail/mitgcm-devel/2011-December/005080.html>
>> I am not surprised that it leads to different ice thicknesses, because the flag changes the radiation budget. You might be able to recover somethings similar as before with setting the ice_emissivity=0.97 in data.exf. Now you use the default, which is 0.95
>> 
>>  
> Hi Martin,
> thanks for the reply and the clarification.
> In addition to ice_emissivity (0.97 or 0.95) in seaice_solve4temp.F,
> another piece of code could cause the difference with this flag
> is exf_radiation.F
>>            lwflux(i,j,bi,bj) =
>>     &          ocean_emissivity*stefanBoltzmann*
>>     &          ((theta(i,j,k,bi,bj)+cen2kel)**4)
>>     &          - lwdown(i,j,bi,bj)
>> #ifdef EXF_LWDOWN_WITH_EMISSIVITY
>>     &           *ocean_emissivity  !!!0.97
>> C     the lw exitance (= out-going long wave radiation) is
>> C     emissivity*stefanBoltzmann*T^4 + rho*lwdown, where the
>> C     reflectivity rho = 1-emissivity for conservation reasons:
>> C     the sum of emissivity, reflectivity, and transmissivity must be
>> C     one, and transmissivity is zero in our case (long wave radiation
>> C     does not penetrate the ocean surface)
>> #endif /* EXF_LWDOWN_WITH_EMISSIVITY */
> I'll try to see if we could recover the old result
> by just changing ice_emissivity without touching the code above.
> 
> cheers
> hong
> 
> _______________________________________________
> MITgcm-devel mailing list
> MITgcm-devel at mitgcm.org
> http://mitgcm.org/mailman/listinfo/mitgcm-devel