CO2 concentration in LUMIP's esm-ssp585-ssp126Lu

[oyvindseland]

Tammas Loughran contacted Yanchun He, later with cc to me about about GHG values in the esm-ssp585-ssp126Lu simulation.

"The temperature shows a trend consistent with an increasing greenhouse warming that I would expect to see in this simulation. However, the co2 concentration is unexpected. It increases only slightly to just over 405ppm and stays constant for most of the simulation. The spatial distribution of co2 concentration starts with a hemispherical asymmetry and then tends towards a uniform distribution. Hopefully this is simply a mistake in the post-processing (wrong variable or file maybe?), because I can't explain this."

It is is a known issue that in concentration driven set-up (almost all simulations) this value does not make sense and should probably not have been cmorized. However

"The LUMIP protocol says that it should be an emissions-driven simulation. If it was run with co2 emissions, then this issue shouldn't have affected this simulation (if I understood the problem correctly)."

I tried to find out more about the simulation but was not able to find any case information or raw output data, and I do not know who actually did the simulation. Can anyone check / find out how the GHG concentrations are included in this specific simulation?

[JorgSchwinger]

I can confirm this observation, the co2 output in esm-ssp585-ssp126Lu seems to be wrong. It is correct in the cmorized esm-ssp585 simulation, so this does not seem to be a comorization problem. I believe this simulation was run by Yanchao Fan, but I have no idea, whether the case directory is still available somewhere. Based on this output, I see the possibility that the simulation set-up was incorrect, so it would be good to check the set-up in the case-directory. Maybe Jerry has more information?

[tjiputra]

This was run by Lei Cai (lcai@ynu.edu.cn). I contacted him earlier and he told me he has no longer access to Fram and can't check the case. Someone need to reinitiate his access to investigate this further.

[oyvindseland]

Yes, I agree that it is likely to be an error in the set-up, but if we know the adress of the simulation on Nird we can also check the raw output. Can this be checked from any cmorization log ? @YanchunHe

I checked Lei Cai 's directory on the INES project. There are no coupled cases, only land only in that project so it has to be found somewhere else..

[YanchunHe]

@oyvindseland

On NIRD: /projects/NS9560K/noresm/cases/b.e21.NSSP585frc2.f19_tn14.CMIP6-SSP5-8.5-SSP1-2.6Lu.001/LUMIP

This should be the raw output. But the case directory is unfortunately not archived here!

[tammasloughran]

Hi folks,

many thanks for tracking down the raw data for this simulation. Is there any way to make the CO2 concentration data from that available to me?

Kind regards, Tam

[JorgSchwinger]

Hi all

from the raw output I figure that this simulation was not set up as emission driven as it should. Therefore, the CO2 output is meaningless (both in cmorized as well as raw data), and this cannot be fixed without setting up the simulation correctly and re-running it. I am not sure whether we have the capacity to do this.

We should retract this simulation from ESGF.

Sorry Tam, we're not able to provide the correct CO2, at least not in the near future. Thanks for making us aware of this issue!

[tammasloughran]

Okay I understand

[YanchunHe]

This experiment will be retracted (refer to the above-mentioned issue in the noresm2cmor github repository).

[oyvindseland]

Experiment retracted so closing the issue. Open again if the experiment have been corrected or re-run