== Sea ice outputs for CMIP6 with LIM==

'Created by M. Vancoppenolle'

 * '''siitdconc'''

This field refers to « Sea ice fraction in ice categories ». Why are units in « m-1 »? This would mean that the actual fraction is divided by ∆h. I would suggest to have the direct ice fractions (MV). We'll have problems in the 5th category.

 * '''siitemptop and siage'''

The category-averaging is not explicit for these two fields. We would compute this over the ice-covered part of the grid cell. I.e., <Tsu> = ( ∑_cat Tsu_cat . frac_cat) / ( ∑_cat frac_cat)

 * '''siflfwbot'''

Does that include snow melt ?

 * '''Surface energy budget'''

We will have major trouble to output the fields as requested. In NEMO, the exchanged surface fluxes are grouped into « net solar » and « net non-solar ». We do not have access to the individual components (Fsw_up, Fsw_down, Flw_up, Flw_down, Flh, Fsh) on the ocean model grid. We could diagnose Fsw_up and Fsw_down using Fsw_net and albedo.
We could at least provide the net solar and net non-solar atmospheric heat fluxes over sea ice and ocean fractions.

However, the list of fields as presented now does not include these two contributions, which is a pity. 
Otherwise, there would be no information on surface energy budget in sea ice models using LIM, which would be a pity.

 * '''Integrated diagnostics'''

Would be easy to provide but would involve global sums at each model time step, which may significantly increase the cost of the model. More importantly, the ice extent depends / volume on the grid of the model (Canadian Archipelago or not, …). Hence, different models might not be directly comparable to observations. For ORCA2 grid, we had offsets of 106 km2 at least.


 * '''Minimalist documentation'''

A small (minimalist) documentation of codes might be useful with at least ice category boundaries and physical parameters (water, ice, snow densities, specific heat of pure ice, latent heat of fusion of pure ice, thermal conductivity of pure ice, latent heat of sublimation, …). For instance, we don’t plan to provide ice mass since it duplicates volume, but can easily be retrieved.

[https://forge.ipsl.jussieu.fr/nemo/wiki/2016WP/2016SharedActions/Workshop_NEMO_in_CMIP6/ice_outputs Back]