Icepack standalone

[pduarte8]

Hi,

I have just noticed the following commented lines in icepack_therm_vertical, in subroutine frzmlt_bottom_lateral:
!! uncomment to use all frzmlt for standalone runs
!fbot = min (c0, frzmlt)
I wonder if this is really to uncomment in standalone simulations, especially those using a slab-mixed ocean model. These two commented lined were already present in the columnar CICE package, in the same subroutine in file ice_therm_vertical.
Uncommenting this line has a strong effect on model results (as expected...).
Anyone may clarify this? Iam trying to do it myself but any help is welcome.

All the best

Pedro Duarte

 

[dbailey]

Hi Pedro,

This is interesting. I don't typically run with standalone CICE, so haven't noticed this before. When we run coupled, the "frzmlt" term can have either sign. Negative implies heat lost from the ocean and positive implies heat gained by the ocean. I will look into this further.

 

[pduarte8]

Hi

Thanks for your interest. Actually I have analyzed this a bit more and it seems that it is a good thing that it is commented and probably, it is convenient to remove it. The problem I see here is the following (I am not yet 100% sure....): So, this part of the code is used only when frzmlt < 0 and melting is possible. However, when frzmlt is calculated by the slab ocean mixed layer model in icepack_ocean (frzmlt = (Tf-sst)*cprho*hmix/dt) its value is easily quite negative - all it takes is a small difference between Tf and sst for frzmlt become of the order of thousands of Watts. So, later when we equate fbot with frzmlt, if we uncomment the line mentioned in my previous post, then we get a very negative fbot warmign the ice. The does not seem realistic. In fact it seems also to be cooling down the water in the mixed layer leading to a reduction in frzmlt. However, despite this reduction, it remains in the order of tens of Watts and, once gain, leads to high fbot and too much heat from the ocean to the ice. So, I guess this is a line of code used in some experience and abandoned but I may well be wrong....

 

[eclare@lanl_gov]

Hi Pedro,
You are right about this. The code normally only uses part of the available heat from the ocean, as determined from the sea ice calculation, and leaves the rest of it in the ocean. I don't remember which model was having trouble, but I think the ice was too thick and so we were playing around with the heat flux to try to melt more of it. I agree with you that we should just delete these commented lines from the code.

 

[pduarte8]

Hi
Thanks a lot for your answer. Good to know I was right or else a lot of work would be compromised :->
All the best.

Pedro