experiment descriptions for snapsi catalogue records

[charliepascoe]

Defining experiment descriptions for snapsi catalogue records
They should have enough information to be meaningful yet generic enough to be applicable to all the snapsi data providers.

For comparison here are the github issues for confirming the general descriptions of the CCMI-2022 experiments
refD1 cedadev/ccmi-2022#59
refD2 cedadev/ccmi-2022#60
senD2-sai cedadev/ccmi-2022#61
senD2-ssp126 cedadev/ccmi-2022#62
senD2-ssp370 cedadev/ccmi-2022#63

[charliepascoe]

The experiment_id and accompanying experiment field in the SNAPSI_CV.json have a list of experiment names and short descriptions.
https://github.com/cedadev/snap/blob/main/Tables/SNAPSI_CV.json

[charliepascoe]

There is also info about the experiments in section 2 Overview and Motivations on page 4 of the snapsi GMD paper
https://gmd.copernicus.org/preprints/gmd-2021-394/gmd-2021-394.pdf

[charliepascoe]

And a more context here https://www.sparc-climate.org/2021/04/30/new-set-of-controlled-numerical-experiments-stratospheric-nudging-and-predictable-surface-impacts-snapsi/

[aph42]

Here are a set of first drafts for a template of the catalogue record. Following the CCMI-2022 example, I've included a 'model intro' section that would include a brief description of the model components, an 'experiment description' section that describes the experiment, and a 'SNAPSI' intro that describes the project as a whole. Each record would include these three paragraphs. The sub-experiment ids correspond to different initialization dates; I've just indicated how they should be interpreted.

Model intro

This dataset contains model data for SNAPSI experiment 'free' generated by the CNRM-CM 6.1 model produced by the modelling team at Météo France. The model consists of the ARPEGE-Climat 6.3 atmospheric model, the NEMO 3.6 ocean model, the GELATO 6 sea ice model and the ISBA-CTRIP land surface model.

SNAPSI intro

The SNAPSI project is a model intercomparison project to study the role of the stratosphere in subseasonal forecasts following stratospheric sudden warmings and the representation of stratosphere-troposphere coupling in subseasonal forecast models.

Experiment intro

free

The free experiment is a set of retrospective, 45-day, 50-member ensemble forecasts. The forecasts are initialized on the date indicated by the sub-experiment id; for instance, the sub-experiment 's20180125' is initialized on 25 January 2018. The ocean, sea-ice, land-surface and ozone are all initialized and run prognostically.

control

The control experiment is a set of retrospective, 45-day, 50-member ensemble forecasts. Following the initial date, the stratospheric zonal mean temperatures and zonal winds are nudged towards the time-evolving climatological state. The forecasts are initialized on the date indicated by the sub-experiment id; for instance, the sub-experiment 's20180125' is initialized on 25 January 2018. The ocean, sea-ice, land-surface and ozone are all initialized and run prognostically.

nudged

The nudged experiment is a set of retrospective, 45-day, 50-member ensemble forecasts. Following the initial date, the stratospheric zonal mean temperatures and zonal winds are nudged towards the observed time-evolving state. The forecasts are initialized on the date indicated by the sub-experiment id; for instance, the sub-experiment 's20180125' is initialized on 25 January 2018. The ocean, sea-ice, land-surface and ozone are all initialized and run prognostically.

control-full

The control-full experiment is a set of retrospective, 45-day, 50-member ensemble forecasts. Following the initial date, stratospheric temperatures and horizontal winds are nudged towards the time-evolving climatological state. The forecasts are initialized on the date indicated by the sub-experiment id; for instance, the sub-experiment 's20180125' is initialized on 25 January 2018. The ocean, sea-ice, land-surface and ozone are all initialized and run prognostically.

nudged-full

The nudged-full experiment is a set of retrospective, 45-day, 50-member ensemble forecasts. Following the initial date, stratospheric temperatures and horizontal winds are nudged towards the observed time-evolving state. The forecasts are initialized on the date indicated by the sub-experiment id; for instance, the sub-experiment 's20180125' is initialized on 25 January 2018. The ocean, sea-ice, land-surface and ozone are all initialized and run prognostically.

[aph42]

From Damien: He suggests the model intro be modified to

This dataset contains model data for SNAPSI experiment 'free' produced by the seasonal prediction research team at Météo-France. It is generated with the coupled climate model CNRM-CM 6.1. The model consists of the ARPEGE-Climat 6.3 atmospheric model, the NEMO 3.6 ocean model, the GELATO 6 sea ice model and the ISBA-CTRIP land surface model.

to be more precise about the team that produced the runs.

[mollymacrae]

Draft CEDA catalogue records:

'free' experiment example:
https://catalogue.ceda.ac.uk/uuid/bbe6dfaa0c9a4dfb8a0e7f131cc4d0b4

Météo France dataset collection (which links to all the individual experiment records in 'Related Records' section):
https://catalogue.ceda.ac.uk/uuid/201dac63fd1046e39299801dbf47dd3f

google doc for progress of data received at CEDA and record links:
https://docs.google.com/spreadsheets/d/1BytwB7yYPJ9E7p-cgBa7E356DjdQWrvVmBqkf4Pqk9o/edit#gid=0

In the 'Details/Docs' section are the two SNAPSI specific links Charlotte provided above, should we also add the reference publication for CNRM-CM 6.1 listed in the SNAPSI paper? For now I have added it just to the dataset collection record but can remove it or add it to all records.

Please add any comments or changes and we can keep updating these as modifications are suggested here!