[MITgcm-support] KPP & OB file

[Martin Losch]

Hi Riema,

In order to use KPP you have to put "kpp" into your local  
packages.conf and you have to set #define SHORTWAVE_HEATING in your  
local CPP_OPTIONS.h because kpp requires a two band incoming  
radiation (short wave= visible light and long wave = the remaining  
part). The model will stop if you don't have short wave heating defined.
You don't need to set any other parameters, unless you are unhappy  
with the choice of default values for, say, Ricr, etc. (see  
kpp_readparms.F). I have never changed them so far, but it's  
possible. KPP then compute vertical mixing coefficients etc based on  
your present vertical stratification, vertical shear, surface fluxes,  
etc. (see, eg., Large et al 1994 for details)

OBCS:
exp4 gives an example of how to specify CONSTANT (in time) open  
boundary values from files:
  useOBCSprescribe = .TRUE., (enables reading from files in general)
  OBNuFile = 'OBmeridU.bin', (example of a file)
[...]
these files were generated with exp4/input/gendata.m. The script tell  
you that a filed for the northern (southern) boundary should have the  
dimension (nx,nr) and for eastern (western), (ny,nr).

If you want to specify time dependent boundary values you can do that  
too. For example, you have hourly data you have create a field  
(nx,nr,nt), where nt is the number of "hours" you have. and then you  
use the parameters
   periodicExternalForcing=.TRUE.,
   externForcingPeriod=3600.,
in the file "data", in namelist PARM03. If your data is the same  
after say 12 hour (ie., that you want to use the fields repeatedly)  
then add this parameter in data
  externForcingCycle=43200.,
These parameters that control reading external forcing data (also  
surface forcing if you use any) in general. the drawback is that all  
forcing data have to be at the exact same time levels. The advantage  
is , that it's simple.
The alternative is to use the exf-package, where you can use  
different frequencies etc for different input fields, but it's also  
far more complicated.

Does that help?

M

On 2 Mar 2007, at 02:45, Riema Rachmayani wrote:

> thanks martin for KPP...KPP is  special in that you have to set  
> #define SHORTWAVE_HEATING in  CPP_OPTIONS.h in addtion...so you  
> mean that KPP is used ONLY when we define shortwave_heating??so i  
> dont need to change Ricr  for my internal wave case...because i  
> thought it influence of dynamic stability...
>
> one more question :
>
> now i'd like to ask about open boundaries file, as in \verification 
> \exp4\input\data.obcs :
>
> # Open-boundaries
>  &OBCS_PARM01
>  OB_Jnorth=80*-1,
>  OB_Jsouth=80*1,
>  OB_Ieast=42*-1,
>  OB_Iwest=42*1,
> #useOrlanskiNorth=.TRUE.,
> #useOrlanskiSouth=.TRUE.,
> #useOrlanskiEast=.TRUE.,
> #useOrlanskiWest=.FALSE.,
>  useOBCSprescribe = .TRUE.,
>  OBNuFile = 'OBmeridU.bin',
>  OBSuFile = 'OBmeridU.bin',
>  OBWuFile = 'OBzonalU.bin',
>  OBEuFile = 'OBzonalU.bin',
>  OBWsFile = 'OBzonalS.bin',
>  OBWptrFile(1) = 'OBzonalS.bin',
>  &
> # Orlanski parameters
>  &OBCS_PARM02
> #Cmax=0.45,
> #cVelTimeScale=1000.,
>  &
>
> in the blue line means that data  open boundaries is given from  
> data observation/prediction?? may i know how to read and to write  
> (format) for that open bondaries file?? is the data read every time  
> step or every one hour??how to interpolate them?? in what routines  
> that files read??
>
> in my case, i'm gonna use tidal prediction data of eta (elevation)  
> from ORITIDE from Ocean Research Instute, University of Tokyo, the  
> data is in every one hour, and usually we use spline cubic  
> subroutine to interpolated one hour data become every time step  
> data, so that every time step data/value used for open boundaries  
> input/value...how to modify my case in MITgcm?? because as far i  
> can see there is no open boundaries file for eta...
>
> thank you,
> regards
> rima
>
>
> Send instant messages to your online friends http:// 
> uk.messenger.yahoo.com
> _______________________________________________
> MITgcm-support mailing list
> MITgcm-support at mitgcm.org
> http://mitgcm.org/mailman/listinfo/mitgcm-support