General - Sung

• NeSI allocation about 50% used (expires end of May) (https://quakecoresoft.canterbury.ac.nz/dashboard/)
• NeSI Mahuika Milan - Apptainer (ex-Singularity: Docker for HPC) trial  
• Office layout 
• What IT gear do we need for 2025? : Better monitors (HP). Take RAM and GPU from zoom setup 
• How can we utilize Github project?: https://github.com/orgs/ucgmsim/projects/4 

Cybershake (ie. future simulation) - Joel

• Examine workflow orchestration tools (Pegasus, 'cylc')

Validation (ie. historical simulation) - Sung

• Small magnitude uncertainty (Sarah) : no immediate plan to run sims : Sung to run on KISTI on behalf

Velocity Model - Sung

• Velocity Model (NZVM code re-write)
• Review development vs awp-reverted branch :
  • NZVM MPI AWP implementation  
  • https://uceqeng.slack.com/archives/C3SGQHXT8/p1691367875056399

Source Model - Jake

• Type 5 Source modelling:
  1. Rough stages for implementation:
     1. Scoping and designing file types (DONE)
     2. Developing algorithms to infer jump points and causality (DONE)
     3. Choosing initial fault (TODO; pending Brendon)
     4. Debugging GSF file generation (IN-PROGRESS; have written my own GSF generator)
        1. C/Python comparison: C has 322 unreadable poorly performing lines and memory leaks, Python has 50 lines and no leaks.
     5. Running test simulation (TODO; could probably do now)
     6. Source parameter variance (TODO)
  2. "nice to haves"
     1. Better error output for realisation files (TODO; Using the Python schema library)
     2. NSHM database (IN-PROGRESS; I think it will be a really useful tool once completed)
     3. Some sort of better interface to specify custom type-5 scenarios. Currently done with SrfGen (TODO; Dovetails really nicely with (b) and I have a lot of ideas)
     4. Improving the visualisation tools for realisations, GSF files and SRF files (TODO)

NZGMDB - Joel

• P-wave picker selection (low priority)
  • Perform brief review of existing p- & s-wave pickers and perform comparison/testing
  • Integrate into NZGMDB workflow
• Re-implement/Tidy up GMC workflow
  • Use p-wave picker selected above, as this is the main reason for the GMC pain, also makes it be consistent with rest of NZGMDB workflow (low priority)
• Investigate Ds multiplier and other methods for grabbing the correct window for the waveforms (Low priority)
  • Produce plots / streamlit app to investigate to visually see how changes to ds multiplier affects results
  • Discuss with brendon and team for a good/different approach
• Phase arrival merge of Geonet data (Andrew)
• Fmax calculation (Andrew)
• IM calculation (Andrew)
• Process Waveforms (Joel)
• Merging flatfiles (Joel)

...