Which is the proper variable of Sea Ice freshwater flux into the Ocean

COSIMA
1

Hi Folks,

I am interested in the sea ice freshwater flux into the ocean of ACCESS-CM2 historical data. I grabbed some output variables from CMIP6. But I am not sure which is the proper one, they seem similar.

‘siflfwbot’:

image

‘fsitherm’:

image

‘sidmassth’:

image

Even ‘fsitherm’ and ‘siflfwbot’ have the same standard name, the data are different.
I am not sure which one contains the salty part, or total freshwater, or freshwater compensated by salt. I prefer to use the freshwater flux compensated by salt. :slight_smile:

Many thanks

2

Hi Zijin, I am most familiar with the output from the CICE model so the sidmassth variable called dvidtt in access-om2 output files its the rate of change of the ice volume/mass when multiplied by density in this case due to the ice thermodynamic changes and is not what you are looking for.

Siflwbot is the the fresh water directly from the ice model output.

fsitherm is the melt term in MOM5 code “melt = -Ice_ocean_boundary%salt_flux(i,j)*ice_salt_concentration_r”

so hope that explains the differnce between these two terms. Its never in my checks been a perfect match due to accumulative round off error over monthly averages.

We left both measures in so user could decide which one they wanted to use based on their usual approach.

2 Likes
3

Thank you, Siobhan

I want double check that siflwbot is a freshwater flux and fsitherm is a water flux with salt? Or fsitherm is also a freshwater flux. I didn’t see any salt flux variable in the output file, so I guess it may have calculated the salt in the output.

4

access_fwflux_diags
access_fwflux_diags984×582 74.2 KB

Revisitiing this topic, since we’re still a bit confused. @Zijin has made a nice plot (hopefully attached) comparing sea ice production/melt sidmassth (related to dvidtt in CICE-world) with the 2 sea-ice related freshwater flux variables (fsitherm and silfwbot). The values are presneted as Sverdrups and seperated into freeze and melt.

  1. Both sidmassth and fsitherm approximaletly balance (they melt as much as they freeze) which impies that for ACCESS-CM2 fsitherm does not include snow - most of the other CMIP6 groups did (and it seems to be in silfwbot, since melt puts more FW in the freezing takes out)

  2. this is the real kicker for us - both fsitherm and siflwbot seem to move more freshwater in/out of the ocean than the actual mass of sea ice freeze and melt. Even is the sea ice were completely fresh and all brine was rejected, this would still be a bias, no?

@sofarrell @aekiss any ideas or comments?

1 Like