[Mitgcm-support] Re: KPP units

[mitgcm-support at dev.mitgcm.org]

Hi Dimitri,

 You're back, but now Alistair is on holiday for two weeks so
you'll have to make do with me!!

 Just to clarify further. As you noted, the convention described
in FFIELDS.h is that the FFIELD data structures use the same units 
**and sign** as internal tendency variables. However, input 
files etc... can be, and often are, in different units. So for example 
heatflux input datasets are often in W/m^2 with +ve into the
atmos. 

 I imagine the internal dynamics "FFIELDS" convention may well not make sense 
for all components of the ECCO environment. In an analogous way, our coupled run
fluxes are exchanged in terms of W/m^2 (latent, sensible heat), 
m/s (precip. , evap) etc... In these cases our convention is that
units appropriate to the context are used and that the
units need to be clearly documented, along with the variable definiton 
etc..., in an appropriate place in the actual code.

Chris