Documents and code for canari data management
This repo will eventually include
- all the documents which describe the CANARI Large Ensemble data environment,
- all the code we use to manipulate data in the CANARI workflow
Meanwhile, this is the draft file attributes specification.
We will be doing the following HIST runs from 1950, with SSP370 continuation to 2100
- there are two parent/child relationships here (CMIP6 HIST -> CANARI HIST2 and CANARI HIST2 -> SSP370)
- branch method: "micro initialisation from CMIP6 macro parent" for the first one set, "continuation" for the second.
- Note that the HIST compoenent should be stored as HIST1, which is a variant of the CMIP HIST run, which differs only in the duration.
- (note the slight change in the experiment name, which is only about duration)
- the branch time in the parent and child will be the same
- branch method: "micro perturbed atmosphere initialisation" for the first one, "continuation" for the second.
- (note that the hash in the variant labels should be replaced by the realization index)
- note that P2 is used because we have the corrected ozone runs so this isn't quite vanilla HadGEM3-GC31.
Variant Label | Parent | Initialisation Index | Realiisation Index | Notes |
---|---|---|---|---|
r#i1f1p2 | CMIP6.CMIP.MOHC.HadGEM3-GC31-MM.historical.r1i1p1f3 | 1 | 1-5 | Five variants iniitialised by perturbing initial atmosphere |
r#i2f1p2 | CMIP6.CMIP.MOHC.HadGEM3-GC31-MM.historical.r2i1p1f3 | 2 | 6-10 | Five variants iniitialised by perturbing initial atmosphere |
r#i3f1p2 | CMIP6.CMIP.MOHC.HadGEM3-GC31-MM.historical.r3i1p1f3 | 3 | 11-15 | Five variants iniitialised by perturbing initial atmosphere |
r#i4f1p2 | CMIP6.CMIP.MOHC.HadGEM3-GC31-MM.historical.r4i1p1f3 | 4 | 16-20 | Five variants iniitialised by perturbing initial atmosphere |
We need an additional four simulations to provide initial conditions from different ocean states, these run from 1850 to 1950.
- these runs are initialised from PICONTROL at 50 year intervals
- there is one parent/child relationships here (PICONTROL -> CANARI HIST1)
- branch method "50 year intervals in PICONTROL to get diverse ocean states"
- Note that these ones, unlike all the other ones, will have a different branch time in the child than in the parent, the children start date will always be 1850 the parent start date will be whatever is selected.
- Note that these should be stored as HIST1, which is a variant of the CMIP HIST run, which differs only in the duration.
Variant Label | Parent | InitialisationRealisation Index | RealisationI Index | Notes |
---|---|---|---|---|
r21i5f1p1 | CMIP6.CMIP.MOHC.HadGEM3-GC31-MM.piControl.r1i1p1f1 | 5 | 21 | Just the one run but check ocean state at start |
r26i6f1p1 | CMIP6.CMIP.MOHC.HadGEM3-GC31-MM.piControl.r1i1p1f1 | 6 | 26 | Just the one run but check ocean state at start |
r31i7f1p1 | CMIP6.CMIP.MOHC.HadGEM3-GC31-MM.piControl.r1i1p1f1 | 7 | 31 | Just the one run but check ocean state at start |
r36i8f1p1 | CMIP6.CMIP.MOHC.HadGEM3-GC31-MM.piControl.r1i1p1f1 | 8 | 36 | Just the one run but check ocean state at start |
These are then continued by another set of runs, the second tranch core-set
- there are two parent/child relationships here (CANARI HIST1 -> CANARI HIST2 and CANARI HIST2 -> SSP370)
- branch method = continuation
- can start when a) the macro parent is finishd, and b) checked to have a different state
Variant Label | Parent | InitialisationRealisation Index | RealisationI Index | Notes |
---|---|---|---|---|
r21i5f1p2 | CANARI-LE.NCAS.HadGEM3-GC31-MM.hist1.r#i5f1p1 | 5 | 21 | Continues without perturbation |
r26i6f1p2 | CANARI-LE.NCAS.HadGEM3-GC31-MM.hist1.r#i6f1p1 | 6 | 26 | Continues without perturbation |
r31i7f1p2 | CANARI-LE.NCAS.HadGEM3-GC31-MM.hist1.r#i7f1p1 | 7 | 31 | Continues without perturbation |
r36i8f1p2 | CANARI-LE.NCAS.HadGEM3-GC31-MM.hist1.r#i8f1p1 | 8 | 36 | Continues without perturbation |
We will be doing an additional set of HIST runs from 1950, with SSSP370 continuation based on the Macro Tranch.
- there are two parent/child relationships here (CANARI HIST1 -> CANARI HIST2 and CANARI HIST2 -> SSP370)
- branch method for the first set: "micro perturbed atmosphere initialisation"; "continuation" for the second set
- can start when a) the macro parent is finishd, and b) checked to have a different state
Variant Label | Parent | Initialisation Index | Realisation Index | Notes |
---|---|---|---|---|
r#i5f1p2 | CANARI-LE.NCAS.HadGEM3-GC31-MM.hist1.r#i5f1p2 | 5 | 22-25 | Four variants iniitialised by perturbing initial atmosphere |
r#i61p2 | CANARI-LE.NCAS.HadGEM3-GC31-MM.hist1.r#i6f1p2 | 6 | 27-30 | Four variants iniitialised by perturbing initial atmosphere |
r#i7f1p2 | CANARI-LE.NCAS.HadGEM3-GC31-MM.hist1.r#i7f1p2 | 7 | 32-35 | Four variants iniitialised by perturbing initial atmosphere |
r#i8f1p2 | CANARI-LE.NCAS.HadGEM3-GC31-MM.hist11.r#i8f1p2 | 8 | 37-40 | Four variants iniitialised by perturbing initial atmosphere |