How to run ACCESS-AM3 N96 post-2012?

Atmosphere AM3
1

Hi AM3 team

I would like to run ACCESS-AM3 N96 for recent years from 2013. I have successfully run it for the year 1982 and 2012, but the run for the year 2013 failed with the following errror message:

????????????????????????????????????????????????????????????????????????????????
???!!!???!!!???!!!???!!!???!!!       ERROR        ???!!!???!!!???!!!???!!!???!!!
?  Error code: 202
?  Error from routine: UP_ANCIL
?  Error message: REPLANCA: PP HEADERS ON ANCILLARY FILE DO NOT MATCH
?  Error from processor: 649
?  Error number: 11
????????????????????????????????????????????????????????????????????????????????

I am wondering whether it is because the ancillary files for sea ice and sst end in 2012, as can be found in the official AM3 alpha release and used here: /home/563/qg8515/roses/am3-test

So I have three questions: 1) Are there N96 SST and sea ice ancillary files for AM3 amip style simulations from 2013 onwards? I am fine for it to be derived from ERA5 or any other reliable sources. 2) If so, do I just need to change the file paths at install_ancil > file > qrclim.seaice/sst? 3) Should similar changes be done for aerosols and ozone? I am using prognostic aerosols.

Regards, QG

2

@qinggangg , the problem definitely is because some ancillaries only have data up to 2012. We haven’t yet created any extra ancillaries in addition to the ones that come with the released configuration. We haven’t yet discussed providing different ancillaries for AM3, but I have opened an issue so the AM3 development team can discuss when this can be provided.

Note that the aerosols and ozone files finish in 2014 so they would need to be modified as well.

We do not yet have a published ancillary suite to produce the ancillary files, but we are working on one. It takes time as the processing of the data for CABLE is different from what is needed for JULES.

Finally, what period would you like to simulate, so we can be sure to produce files that fit?

3

Thank you, @clairecarouge. That’s very nice to know, and I fully understand that time is required to get things done.

Currently, I can run the simulation from 1982-2012. It would be great if I could extend it to 1979 backwards (good to have) and to 2025 forwards (priority). So it would be very nice to have ancillaries up to 2025, for better comparisons with recent satellite and conventional observations. I am not familiar with aerosols and ozone, but for sst and sea ice, I am fine to use data derived from ERA5.

Just to mention that it would be great to have these files before mid-March, so I could exploit the computing resources of this quarter in these simulations. But no pressure if you have other duties. If there is anything I could do myself, please let me know.

4

I am fairly certain you will not have the files before mid-March. Sorry.

But thanks for the extra details, that’s useful.

5

Would you think it’s a good idea to run the model until 2012 this quarter, and then run it to 2025 when the new ancils are available?

6

At this stage, I have no idea if the new ancillaries will be a continuation of the current ones or if they will be based on completely different datasets. So I can’t tell if it will be a good idea to do a continuation run with the new ancillaries or if it will need to be a new run.

7

@qinggangg - a pragmatic thought, you might as well run up to 2012 this quarter given the available compute, then consider future plans (including potential re-runs) once there is more knowledge.

8

Thank you @bethanwhite and @clairecarouge. That makes sense. Please let me know when it is possible to create new ancillaries.

1 Like