How to access CESM2 piControl r2i1p1f1 and r3i1p1f1 outputs of pr and evspsbl in Amon

[feikan]

Hi everyone.

I am working with the CMIP6 LUMIP deforest-globe experiment using CESM2 and comparing the results against piControl. The deforest-globe experiment provides three variant members (r1i1p1f1, r2i1p1f1, r3i1p1f1) for pr and evspsbl in Amon, but when accessing CESM2 piControl data via ESGF, I can only find r1i1p1f1. Where can I obtain the piControl outputs of pr and evspsbl for r2i1p1f1 and r3i1p1f1? Thanks~

 

[asphilli]

Hi Fei Kan,
There is only one CESM2 ~1-degree resolution piControl member that NCAR submitted to CMIP6: r1i1p1f1. If I am not understanding your query correctly, please let me know!
Best,
Adam

 

[feikan]

Hi Adam,


Thanks a lot for the clarification — that helps.

I realize I may not have described my question very clearly before. My original assumption was that each CESM2 deforest-globe variant would correspond to a different piControl variant, since the deforestation signal is usually evaluated relative to piControl.

After your reply, I checked the metadata of the three deforest-globe files (r1i1p1f1, r2i1p1f1, r3i1p1f1) for evspsbl. All three list
parent_variant_label = r1i1p1f1 and the same branch_time_in_parent (182500), as the image below, which seems to indicate that they were all branched from the same piControl member at the same timestamp.




Given this, I’m a bit unsure what exactly distinguishes r1, r2, and r3 in the deforest-globe experiment. Are these simply different realizations generated through small perturbations at branching?

Thanks again for your help.

Best,
Fei

 

[asphilli]

Hi Fei,
I have verified what you show within the netCDF global file attributes of the 3 experiments. I consulted our run database, and the metadata presented match what is shown in the global file attributes as before. Errors do exist in our run database though, and I do not really buy that they were all spun off from the same year of the piControl. So I went to the case directories, and they show the following:

b.e21.B1850.f09_g17.CMIP6-deforest-globe.001/env_run.xml: <entry id="RUN_REFCASE" value="b.e21.B1850.f09_g17.CMIP6-piControl.001">
b.e21.B1850.f09_g17.CMIP6-deforest-globe.002/env_run.xml: <entry id="RUN_REFCASE" value="b.e21.B1850.f09_g17.CMIP6-piControl.001">
b.e21.B1850.f09_g17.CMIP6-deforest-globe.003/env_run.xml: <entry id="RUN_REFCASE" value="b.e21.B1850.f09_g17.CMIP6-piControl.001">
b.e21.B1850.f09_g17.CMIP6-deforest-globe.001/env_run.xml: <entry id="RUN_REFDATE" value="0501-01-01">
b.e21.B1850.f09_g17.CMIP6-deforest-globe.002/env_run.xml: <entry id="RUN_REFDATE" value="0841-01-01">
b.e21.B1850.f09_g17.CMIP6-deforest-globe.003/env_run.xml: <entry id="RUN_REFDATE" value="0871-01-01">
(The run names shown to the left of the lines above match the cesm_casename global file attributes of the CMIP6 data files for each run.)

Based on this information, I can confidently say that the metadata information in the data files is incorrect, that each of the 3 runs were spun off from the piControl, and that the spin off years are 504 (for r1i1p1f1), 841 (for r2i1p1f1) and 871 (for r3i1p1f1).

Best,
Adam

 

[feikan]

Hi Adam,

Thanks a lot for checking this so carefully. Your reply really cleared up my confusion. I really appreciate your time and patience.

Best,
Fei