[MITgcm-devel] seaice ice2watr

[Martin Losch]

Hi there,

am I allowed to retire the runtime flag ICE2WATR? 

Reason: because of this flag it's possible to have inconsistent values of ICE2WATR, SEAICE_rhoIce and rhoConst. Currently this is the case for all verification experiments that use pkg/seaice, except for global_ocean.cs32x15/input.icedyn (because that one does not use the PW thermodynamics), so retiring this parameter requires changing all of these verification experiments.

Why does it matter? I would like to change the freshwater units within the seaice_growth routine into mass units (and then add my sublimation code), but because of the ICE2WATR=0.92 in the "data.seaice"-files it is impossible to test whether these modifications affect the solution or not (any change to the code will, because the densities are inconsistent with their ratio). So I have the choice of either retiring this parameter, thereby changing the above experiments, or keeping ICE2WATR as it is, but changing the verification experiments anyway. What do you think?

There are also a couple of other stupid parameter conventions that I could change along with this operation: SEAICE_sensHeat/latentWater/latentIce that are now what their name implies, but rather some coefficients that are the products of latent heat with some other numbers, I could try to bring them in line with the corresponding exf-parameters for consistency?

Martin