2008/03/26 CSEP meeting

Attendees: Phil, Yan, Fabian, John, Jeremy, Masha

1. Introducing Western Pacific testing region into the CSEP:

   • Amit identified 11 extra events with Mw >= 5.8 in CSEP input catalog vs. UCLA input catalog:

     1. Yan updated his input catalog with missing events

   • Fabian has finished CMT catalog import into QuakePy, will work on acceptance tests

   • To automate Yan's models:
     1. Download and convert CMT catalog in 'ndk' format to one-line-per-event format is done
     2. Write Python wrapper to invoke Yan's model
     3. Write conversion program to populate ForecastML template with Yan's forecast
     4. Will need to split long-term and short-term forecasts into 2 separate files for evaluation purposes:
        • Currently written to the same file
     5. Evaluation test for the model is RELM L-test
2. Software development:
   • Masha reported the status of CSEP V8.4 release candidate:
     1. csep-cert ran successful build for SVN tags/Version_8_3_24 code base
        • Will rename the latest candidate tag to Version_8_4_0 for the release
     2. Nightly Dispatchers run against the same code base
   • Fabian will commit the changes for cumulative plots rejection bars, Masha will re-tag repository and follow usual steps for getting release candidate running on csep-cert machine

   • csep-op will need some upgrading (SciPy, GMT) in preparation for Version 8.4.0

     1. Can skip one day of processing if required - easy to catch up

   • Discussion of Scheduler requirements:

     1. Masha will present purpose/issues/possible solution in the meeting on Friday, March 28, at 11:00am
3. Phil brought up hardware "refresh":
   • To replace CSEP machines:
     1. Integration
     2. Certification
     3. Operational
     4. Re-processing

   • Operational web-server is running on scec

   • Masha asked if we should have operational machine per testing region