Compare_SSH_model_obs.ipynb relies on hardcoded paths

[charles-turner-1]

This recipe:

https://github.com/COSIMA/cosima-recipes/blob/3cf879f02c65d6abcdde9a46eacfc40d0ee4eadf/Recipes/Appetisers-Easy/Compare_SSH_model_obs.ipynb

relies on hardcoded paths to a dataset in ua8 (cell 8):

filenames = glob.glob("/g/data/ua8/CMEMS_SeaLevel/timeseries/*.nc")
cmems = xr.open_mfdataset(filenames, parallel=True)

obs_ssh = cmems.adt
obs_ssh = obs_ssh.sel(time=slice(start_time, end_time))
obs_ssh = obs_ssh.rename('adt_cmems')

ua8 is being decomissioned. Is the data hosted in another project now @paolap? - it looks like you're CI on that project.

If so, we should add the new data to the catalog & update this recipe to remove the hardcoded paths.

[adele-morrison]

@rbeucher do we have SSH data stored somewhere at ACCESS-NRI now that we can replace this with?

[rbeucher]

Yes we do. Let me see what we can do with that one.

[chrisb13]

As earlier, trying to join ua8 I get

ARC Centre of Excellence for Climate System Science data Since the Centre of Excellence for Climate Extremes (CLEX) has closed, this project is getting decommissioned. No new members will be accepted.

I'm currently working on the OM3 equivalent of this (1 and 2, summary here).

I think /g/data/av17/access-nri-temporary/OM3/CMEMS/SEALEVEL_GLO_PHY_L4_MY_008_047/cmems_obs-sl_glo_phy-ssh_my_allsat-l4-duacs-0.125deg_P1D_202411 might be a viable alternative but will have a better idea soon.

This doesn't address @charles-turner-1 point about hard coded paths however.

[paolap]

saw this only now, most of data that used to be in ua8 and were interest was expressed has been moved to others project like jt48, I can't remember where this specific dataset has gone to. We transferred some of the data years ago, the NRI was informed of this and had a chance to copy the data before it was gone

[chrisb13]

Thanks @paolap, I think it's good practice to update it to a newer dataset in any case.

[rbeucher]

We have it in kj13. Not sure it has been transferred to av17 but that could be done if needed.

[rbeucher]

/g/data/kj13/datasets/cosima/CMEMS/SEALEVEL_GLO_PHY_L4_MY_008_047

[chrisb13]

Oh thanks @rbeucher, so is that the same data as this /g/data/av17/access-nri-temporary/OM3/CMEMS/SEALEVEL_GLO_PHY_L4_MY_008_047/?

[chrisb13]

From @rbeucher

I did download the original data from CMEMS. I don't know what COSIMA did.

[chrisb13]

I'm currently working on the OM3 equivalent of this (1 and 2, summary here).

Now done, see this post for summary.

I think it would be better to update this recipe to use ssh^2 (timestep level output) for the variability. I'll do that soon for OM3 in the above repo'.