regression testing with master against Robin's validation benchmark

1. [Timeseries] Enable LF to output Acceleration ascii files

2. Script to enable LF runs to Calculate IM

BB calculation simply reformats LF or HF if either one is missing. ()

(Plan : 2~3 hours)

Restart dumping large amount of data: Upon successful completion of LF, remove checkpoint files

Part of e2e test - minor issues emerged

• LF completion test now checks for zero values in the output

(HF not running reliably : NeSI issue - suspected broken nodes - reproducible by NeSI crew : to alert Brendon)

(HF seed change doesn't affect the result - using default seed)

LF-only/HF-only runs verbally planned. Need written plan - also to figure out what 'BB' features are required.

Restart cleaned up on successful completion of LF

Subduction cybershake (v19p4) from NZFM (just subduction-interface)

• Magnitude Scaling for subduction interface/slab faults
• BB/IMCalc

• HF finding bugs.

• Verify new VM (by plots)
• Stoch file errors with 2km (Robin suggested to try 4 km dx/dy instead )
• SRF validation for 5.4.2 (Shallow Crustal & subduction)
  
  • 50 srf - Single Fault (any) : check stats for consistency - 4 cases
• Wiki Page for CS 19p5
  • New/Maybe/No science for CS
  • Shallow crustals first, subduction later
• Run GMSimViz (Hik)
  
  

(BB/IMCalc done for FiordSZ09 which revealed Fd-station list issue (rotation not considered))

(HF doesn't run for unknown reason: Subduction HF bug finding has been shelved in lieu of potential 'fixes' with new SRF generation or larger dx/y in stoch. Issue found with HikWelMax,HikRauk)

(4km dx/dy works for generating Stoch File, but not 2km. Waiting for new subduction SRFGen (RG) before testing further.)

Cybershake v19p5

GMSimViz for HikWelMax takes >1.5 weeks to run on 12 cores on Hypocentre. Should re-evaluate running it for longer duration sims as a diagnostic.

• Plan Submission of SRF / VM on a realisation by realisation  level
• Remove outdated workflow from NHM2SRF
• Change HYP01 → REL01 in generation / post process code
• Remove Cybershake configuration file (.json)
  (Leave)

(Low priority)

Version parameter removed from cybershake config, awaiting pull request.

Fix disaggregation

Developing unit test/ regression test : laundry list of testing (Claudio to get involved)

Discuss User-class (general purpose, early access, internal)

Upload Waveform for 1 station (prototype)

Test AWS lambda + S3 for waveform retrieval

Ground motion integration

(Disaggregation issue fixed, BB done sanity check)

2 & 3) SeisFinder Test & Release Plan

e2e_test:

1. source generation in validation 
2. ensure that different GMSim versions can be handled

Claudio

Melody

Probability calculations for Jessee 17/ Zhu16 general (to be verified) (1 day analysis)

Investigate shp files

(Verified and checked.)

(Waiting on Liam / Amelia for confirmation to run new version of infrastructure files / provide additional verification.)

1. Version number for GMSim workflow can be any string
2. Reduce extents for VMs out of (165,180) longitude range (June, Ethan)

1. Done (mystery developer)
2. Resolved with temporary fix. Ensure that SRF bounds are within VM domain as a part of vm validation.