WOMBAT: issue with units of ice-to-ocean algae and nitrate fluxes

[dougiesquire]

In MOM5, the units of the prognostic tracer stf field (surface tracer flux) should be "rho * m/s * tracer conc". E.g., WOMBAT prognostic tracers carry units of xmol/m3 (x=m for all tracers except iron which has x=u), so their stf units should be kg*xmol/m5/s. This is the case for DIC, aDIC, O2 and iron (I think, though the provenance of the dust.nc input file is unknown).

However, phytoplankton and nitrate surface tracer fluxes are set directly to the ice-to-ocean algae and nitrate fluxes received from the coupler, which I think have units of mmol/m2/s. I think these fields should be multiplied by rho0. I.e. the surface fluxes are approximately ~1000 too small at the moment.

[pearseb]

That's a good catch Dougie. I really haven't looked closely at these fluxes at all, but I know that there is a lot of room for improvement here.

pearse

[aekiss]

ping @hakaseh

[hakaseh]

I had a quick look and I agree that I had assumed incorrect units for those fluxes, as mmol/m2/s.
I had noticed in the past that these ice-to-ocean fluxes had negligible effects on surface ocean phy or no3 and now it makes sense why. Thanks @dougiesquire for careful check.
Should I fix this?

[aekiss]

Is this error only in the ice->ocean flux or also in the ocean->ice flux?

[aekiss]

Should density be rho0 or ice density? rho0 seems right to me, but I thought I should ask.

[dougiesquire]

Should density be rho0 or ice density? rho0 seems right to me, but I thought I should ask.

Yeah I think rho0. That's what's done for the other surface flux terms in WOMBAT.

[dougiesquire]

Is this error only in the ice->ocean flux or also in the ocean->ice flux?

Sea surface concentration of nitrate and phytoplankton in mmol/m^3 are passed back to the ice (see here and here). This looks correct.

[dougiesquire]

Should I fix this?

@hakaseh, if you have time to submit a PR that would be welcome. I'd be happy to review.

[aekiss]

Is this error only in the ice->ocean flux or also in the ocean->ice flux?

Sea surface concentration of nitrate and phytoplankton in mmol/m^3 are passed back to the ice (see here and here). This looks correct.

Hm, so exchange with sea ice acts as a sink of ocean nutrient and phyto, since 99.9% of it won't come back to the ocean (if we neglect productivity in the ice).

[dougiesquire]

Hm, so exchange with sea ice acts as a sink of ocean nutrient and phyto, since 99.9% of it won't come back to the ocean (if we neglect productivity in the ice).

I may be misunderstanding, but I don't think the ice removes any nitrate/phytoplankton from the ocean. An amount 1000 times too small is fluxed into the ocean from the ice, but nothing is fluxed out. I'm not sure what happens on the CICE side.

[pearseb]

I think it’s likely that if the coupling to sea-ice is done properly, it is as Andrew says, that there is a sink of nutrient and phytoplankton to the sea ice. The phytoplankton in sea ice has to come from somewhere and I would be surprised if Hakaseh didn’t seed the sea ice phytoplankton with phytoplankton from the ocean. From: Dougie Squire ***@***.***> Date: Thursday, 16 May 2024 at 9:47 AM To: COSIMA/access-om2 ***@***.***> Cc: Pearse J Buchanan ***@***.***>, Comment ***@***.***> Subject: Re: [COSIMA/access-om2] WOMBAT: issue with units of ice-to-ocean algae and nitrate fluxes (Issue #282) Hm, so exchange with sea ice acts as a sink of ocean nutrient and phyto, since 99.9% of it won't come back to the ocean (if we neglect productivity in the ice). I may be misunderstanding, but I don't think the ice removes any nitrate/phytoplankton from the ocean. A amount 1000 times too small is fluxed into the ocean from the ice, but nothing is fluxed out. I'm not sure what happens on the CICE side. — Reply to this email directly, view it on GitHub<#282 (comment)>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/AA7GKGGXASF43I63YNSEHLLZCPXX5AVCNFSM6AAAAABGTZWBKGVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDCMJTGY3DAOBSGI>. You are receiving this because you commented.Message ID: ***@***.***>

[aekiss]

Ocean nutrient and phyto is coupled to ice (as well as the reverse) https://github.com/COSIMA/1deg_jra55_ryf/blob/0d412448f6ff5a4ce74987e3e8e0d20d3043334b/namcouple#L322-L331

[ofa001]

@aekiss that suggests that the correction needs to be applied both ways, if it was in error that way too.

[anton-seaice]

Hm, so exchange with sea ice acts as a sink of ocean nutrient and phyto, since 99.9% of it won't come back to the ocean (if we neglect productivity in the ice).

Does this mean it accumulates in an unrealistic way in the sea-ice? Possibly not because the sea-ice will melt and so the nutrient/phyto disappears over summer?

[aekiss]

What I was wondering is if 1 mol (of N or phy) from the ocean arrives as 1 mol in ice, then when it returns to the ocean does that 1 mol in the ice becomes 0.001 mol in the ocean? So the round trip makes 99.9% of N or phy disappear?

[aekiss]

If so, it's actually worse to have the error only in the ice->ocean coupling (not ocean -> ice), because then it acts as a sink of N & phy from the ocean.

[dougiesquire]

Okay I'm a bit confused by some of the discussion here and I think I must be missing something. Looking at the code, what I see happening from the ocean's perspective is:

• the ocean receives surface fluxes of N & phy from the ice
• WOMBAT runs and the surface fluxes are included when calculating the new surface concentrations of N & phy
• the surface concentrations of N & phy are passed back to the ice

If the surface fluxes are 1000x too small, isn't this effectively like having no coupling of N & phy from the ocean's perspective? What am I missing?

[pearseb]

Dougie is there a loss of N and Phy when the ocean passes these terms to the ice?

[dougiesquire]

Dougie is there a loss of N and Phy when the ocean passes these terms to the ice?

If there is, I can't see it. All that's passed to the ice is the surface concentration at the end of the ocean timestep.

[aekiss]

Ah yes, apologies for the red herring, I think you're right @dougiesquire. According to the comments in namcouple, the coupler sends surface concentrations (not fluxes) from mom to cice, and the fluxes are only from cice to mom. Sorry I didn't look more carefully.

[dougiesquire]

From the ice's perspective on the other hand, it fluxes N & phy to the ocean but doesn't see what it sent reflected in the concentrations that are returned.

[pearseb]

Hmmm... So the coupling wasn't a complete coupling. If there wasn't an error in units by 1e-3 then the sea ice would be adding substantial quantities of nitrogen and phytoplankton to the ocean without there being a opposing sink of these fields. We should revisit this in the future and make the coupling complete.

[aekiss]

doesn't see what it sent reflected in the concentrations that are returned

(or only sees it reflected at 0.1% the level it should be)

[aekiss]

Should there also be an atmospheric dust input of nutrient to the ice? Or is that insignificant?

[dougiesquire]

Hmmm... So the coupling wasn't a complete coupling. If there wasn't an error in units by 1e-3 then the sea ice would be adding substantial quantities of nitrogen and phytoplankton to the ocean without there being a opposing sink of these fields. We should revisit this in the future and make the coupling complete.

I'm not sure about this. The ice decides how much N & phy should be fluxed between the ice and ocean based on the surface concentrations. The ocean updates the surface concentrations based on those fluxes (and other things) and the process repeats. The fluxes can presumably go negative so the ice can take N & phy from the ocean.

[dougiesquire]

Should there also be an atmospheric dust input of nutrient to the ice? Or is that insignificant?

I'm not sure, but relatedly the flux of atmospheric dust to the ocean does not get modified by sea ice.

[pearseb]

Wait so there are fluxes? This is getting a bit confusing to be honest.

RE ice and dust: Yes, the ice does accumulate important quantities of dust, with particular relevance for iron fluxes to surface waters during the melt period. This is something that we should look into.

[anton-seaice]

@pwongpan - we're not sure how the ice-to-ocean (and back) fluxes for phytoplankton and nitrate surface tracer fluxes are calculated. They are calculated within CICE, but we are not sure what process that represents (e.g. is it only associated with freeze/melt, or is it something else?)

[hakaseh]

I will have a closer look at the code next week to recall how I did the two way coupling.

[ofa001]

@PearceB and @pwongpan Cice does handle dust, as does Wombat, but I think you told me it was just a climatology, you can pass a dust variable from the UM if we want to think of coupling to there as well. But then I think we need tothink of whether this is "the CICE5 set up like @hakaseh used, what we want to potentially use in CICE6-MOM6 where mushy is an option (though not when couple dot the UM) or when in the older CICE version in ESM1.6 which still uses CICE4.1 which wont be able handle anything more sophisticated ice BGC wise that in CICE5 (we have to use the zero layer version of the thermodynamics there as well.

[hakaseh]

Sorry for the late reply (I'm attending JpGU this week).

I am not sure if I understand the issues here except for the incorrect units for ice-to-ocean flux, but I am guessing there are some questions about whether tracer concentration in sea ice is updated according to the ice-ocean flux?
In CICE5, tracer concentration in sea ice is updated in subroutine skl_biogeochemistry.
The ice-ocean flux is determined based on melting/growth rate of sea ice which is done here.

[dougiesquire]

I am not sure if I understand the issues here except for the incorrect units for ice-to-ocean flux

Thanks @hakaseh. I don't think there are any other issues - just some communication breakdown.

In CICE5, tracer concentration in sea ice is updated in subroutine skl_biogeochemistry.
The ice-ocean flux is determined based on melting/growth rate of sea ice which is done here.

Thanks for checking this and clarifying

[access-hive-bot]

This issue has been mentioned on ACCESS Hive Community Forum. There might be relevant details there:

https://forum.access-hive.org.au/t/cosima-twg-meeting-minutes-2024/1734/10

[access-hive-bot]

This issue has been mentioned on ACCESS Hive Community Forum. There might be relevant details there:

https://forum.access-hive.org.au/t/access-om2-release-information/1602/4