Optimizing Diagnostics for MOM Model Outputs: Enhancing Analysis and Data Management

Technical
1

Dave and I are planning to make modifications to the diag table for 1/4 MOM5 outputs. As a result, I will be organizing an online meeting to present these changes. We welcome anyone interested to join us in the meeting.

Date: 13 June

Time: 11am - 12.30

Purpose and the target model:

The job we are playing now is based on present-day control run.
We have discovered that the 1/4 MOM output is based on variable names rather than the original formats of ocean_month or ocean_daily. This poses two issues: firstly, the current CMORise tool is unable to handle this format. Under such a format we would generate 224,000 files if we run a 1000-year pi-Control simulation; secondly, conducting analysis on such a vast amount of data becomes extremely challenging. As a solution, we have decided to switch back to using the ocean_month and ocean_daily formats.

Additionally, we have noticed that the 1-degree MOM in CM2 sometimes produces unreasonable kinetic energy (KE) values. To address this, we have developed a tool called ocean_ke that allows us to assess the performance of KE. However, once we make changes to the diag_table, the associated scripts will also need to be modified accordingly.

During the meeting, I will discuss the process of setting up a diagnostic table for MOM. Additionally, I will address some potential issues related to CM2.

-~-~-~-~-~-~-~-~-~-~-~-~-~-~-~-~-~-~-~-~-~-~-

Video link

– Do not delete or change any of the following text. –

Welcome to CSIRO Webex Conferencing featuring audio, video and presentation capabilities. You can join this conference from:

Desktop or Mobile Devices

Once connected to your meeting remember to start your audio and video.
Video Conferencing (VC) Rooms
Use the remote control or touch panel and dial the number indicated below:
CSIRO VC Room 26548760899# (hash required)
External VC Room 26548760899@csiro.webex.com
Phones - AUDIO ONLY
Primary Australia +61 2 6246 4433
Quick Dial +61262464433,26548760899%23%23
Other Global Numbers https://conferencing.csiro.au/webex-meetings.php?ref=global_call
Meeting Number/Access Code 2654 876 0899
Password (if prompted) 2gMaA9pq3pb

1 Like
2

I can’t see any details of the date and time for the meeting.

The title says “How to setup a diagnostic table for MOM?”. Does that mean you need assistance with the technical details of doing this? Or are you asking for feedback on what diagnostics to include?

Is this is for an ACCESS-CM2-025 run? Can you give some more context on what the run will be? Is it pre-industrial control?

1 Like
3

Thanks for updating with more details @ars599.

I’ll ping a few people to make sure they see this: @aekiss @rbeucher @MartinDix @dougiesquire

1 Like
4

Terribly sorry I missed this @ars599.

I hope it was a fruitful discussion.

Could you add some notes on the outcome to this topic to keep everyone in the loop? (myself included)

5

Many thanks, Aidan, and no worries.

Managing 224,000 files can be quite challenging if we run a 1000-year pi-Control simulation. However, we believe it would be a great idea to provide two versions of diag_tables for the community:

  1. The first version would cater to COSIMA and OM2 users, featuring an output format based on one variable per output file.
  2. The second version would adhere to the standard ocean_month output format, facilitating compatibility with the current cmorise tool, APP4, and other analysis scripts. In this version, we might also consider excluding Fabio’s outputs and other unnecessary variables to minimize the overall file size.

We welcome any additional feedback and suggestions from the community.