Land and carbon group 4: Ability to routinely compare runs/configurations of coupled model against a set of observations

[tammasloughran]

This is part of the land surface working group's list of action items to meet the requirements for model evaluation for CMIP7.

Description

Likely utilise iLAMB running against multiple simulations of ACCESS – will need to ensure that ACCESS output is sufficient (STASH) and in correct format for iLAMB (CF compliant)

Timeframe

?

[rbeucher]

Hi @tammasloughran,

Please give us more information.
Are you talking about non-cmorised outputs of ACCESS-ESM ACCESS-CM? Or raw outputs of the couple models?

The next version of the ILAMB-Workflow will support non-cmorise data of ACCESS-ESM /ACCESS-CM, data that went through the ACCESS-Archiver. ACCESS-Archiver converts the UM outputs to NetCDF, among other things....

[rbeucher]

@rhaegar325 might be able to help here

[rhaegar325]

Hi, @tammasloughran , really glad to hear from you. However, I am not sure if I understand your question, what does sufficient(STASH) means? could you please give me an example so we could understand your issue more clear?

[tammasloughran]

This is part of the land surface working group's list of action items to meet the requirements for model evaluation for CMIP7.

Are you talking about non-cmorised outputs of ACCESS-ESM ACCESS-CM? Or raw outputs of the couple models?

Either. In essence, the data should be compatible with ILAMB eventually. If we can cut out the CMORization entirely then that would be great.

what does sufficient(STASH)

Once we have defined the observations we want to compare to in ILAMB, we need to make sure that ACCESS outputs the required variables required for the comparison. This may necessarily involve declaring new STASH variables in the UM.

@har917 anything you can add here?

[har917]

On the STASH question - the UM provides great flexibility when creating output, to the extent that only a small fraction of the available output is typically generated from a single simulation. As a general principal we will need to ensure that the model output that we're looking to compare to observation products via iLAMB is made available - implying a potential need to modify the STASHC (the file that controls what gets output) and/or connect CABLE variables through to the appropriate STASH fields and/or adding new STASH variables within the UM.

We should not assume that the STASHC used for CMIP will be all encompassing for the purposes of routine model evaluation.

Examples where this could be relevant include

• extracting tile level outputs (e.g. to compare above-ground biomass of forests/grass against remote sensing products)
• comparing diffuse/direct and visible/near-infrared land albedos to observations
• comparing more detailed model variables against re-analyses.

Examples of where we have had to add new STASH variables to the UM in the past include

• for the tiled soils (soil moisture, soil temperature, soil carbon)
• some of the land nitrogen variables, and
• all of the phosphorus variables.

[access-hive-bot]

This issue has been mentioned on ACCESS Hive Community Forum. There might be relevant details there:

https://forum.access-hive.org.au/t/land-surface-model-evaluation-metrics/1989/1