Meeting Minutes
Date: December 7, 2023
Attendees:
- Micael Oliveira @micael
- Ezhil Kannadasan @ezhilsabareesh8
- Siobhan O’Farrell @sofarrell
- Anton Steketee @anton
- Harshula Jayasuriya @harshula
- Dougie Squire @dougiesquire
Changes in Meeting Format:
- First monthly meeting dedicated to updates from TWG.
- Second monthly meeting focused on discussing issues, planning, and prioritizing work.
Ezhil:
- Developed the IAF Configuration for MOM6-CICE6-WW3.
- Compared MOM6-CICE6-WW3 IAF data with NDBC buoy data; good match for lower HS, deviation for larger HS.
- Resolved a bug in existing MOM6-CICE6 IAF configuration;
year_alignin IAF should matchyear_startin streams XML.
Discussion on year_align for RYF Configuration:
- Correct
year_aligncrucial to avoid incorrect forcing. - Debate on ideal
year_alignfor RYF: use 1900 (start of RYF dataset) or 0000 (simulation start year).
Dougie:
- Completed first successful om3 run with bling, focusing on CO2 and O2 concentrations.
- Addressed handling of fields not provided by the atmosphere.
- Discussed reading capabilities for files not provided by model components.
Anton:
- Identified a bug in openmp hindering parallel reading.
- Working on adding chunking info to CICE output.
- Progressing with CICE standalone on Gadi; exploring different CICE versions and potential CICE6-WW3 integration using NUOPC.
Michael:
- Discussed git workflow for patch file creation and PRs; emphasized unified repo and configuration branching.
- Collaborating on name list file comparison between configurations.
- Covered profiling challenges, work with cmake, and Python scripts for configuration management.
- Raised questions on the applicability of optimization tools across different models.
Harshula:
- Spack
dev-buildallows compiling/installing directly from git cloned directory. - CI can check syntax and test build of the changes in a PR.
- Spack v0.21 was released in November. Has problems with building
parallelio. - Helping Sean improve the CABLE build system, at the moment.
- Observed 2% performance drop in access-om2 for 1-degree configurations compared to COSIMA.https://forum.access-hive.org.au/t/testing-spack-model-component-builds-for-access-om2/1567/2
Two-Week Planning:
- Ezhil: Continue development and analysis in IAF configurations, comparison with satellite data and seek BOM assistance to check the flags and input parameters of ww3.
- Anton: Finalize CICE standalone; collaborate on analysis recipes.
- Dougie: Assess output from coupled bling runs; seek NCAR MOM assistance.
- Michael: Prioritize optimization and profiling for an OM3 configuration.
Next Meeting:
- Scheduled for December 20, 2023, to discuss further planning.