Brute force calculation of a limited set of hindcast ensemble means

[Thomas-Moore-Creative]

Deliver limited hindcast data

run at one, two, and three month lead times from both a summer initialisation date of February 2018 as well as a winter initialisation date of June 2018. This will provide hindcast results for March – April - May and July-August-September 2018.

[Thomas-Moore-Creative]

From internal S2 presentation:

Three ensembles for each day n to n-8 days -> 27 ensembles total

Possible steps forward:

• make a list of files manually
• will we need a preprocessor to deal with extra time dimension?

[Thomas-Moore-Creative]

Attempting to put ACCESS-S2 Hindcast (S2HC) data objects through the S2RA export workflow

hindcast_workflow https://github.com/Thomas-Moore-Creative/NCI-ACCESS-S2-ARD/tree/main/notebooks/hindcast_workflow

1. S2HC_brute_force.ipynb --> create NC files of ensemble means with both lead_time and forecast_start_date
2. S2HC_mask_ETBFcrop.ipynb
3. S2HC_regrid.ipynb
4. S2HC_ETBFcalc.ipynb
5. S2HC_ETBF_doc_test_export.ipynb

[Thomas-Moore-Creative]

looks like somehow both SST and SSS have somehow been lost between step one and two?

Duplicate naming in the 2D and 3D files for temp and salt is likely the problem there.

[Thomas-Moore-Creative]

finished up through step 3 generating accessS2.HC.ocean.masked.AUSWCPregion.TUV_025deg.zarr

Problem!!!: T's are missing!?!

need to add generation of T's to step 4.

[Thomas-Moore-Creative]

completed step 4 with generation of T's

new ds_TUV.to_zarr(path+'accessS2.HC.ocean.masked.AUSWCPregion.ETBFvars.zarr',consolidated=True) written moving on to final step

[Thomas-Moore-Creative]

Final step (5) S2HC_ETBF_doc_test_export.ipynb completed for Feb 2018