Difference between revisions of "BBP Validation 2020"

From SCECpedia
Jump to navigationJump to search
 
Line 13: Line 13:
 
** Validation metrics that already exists (duration, inter-frequency correlation, Anderson GOF, etc.)?
 
** Validation metrics that already exists (duration, inter-frequency correlation, Anderson GOF, etc.)?
 
** New validation metrics?
 
** New validation metrics?
* Group validation of RidgeCrest Earthquake Sequence
+
* Events to consider/discussion notes
 +
** Group validation of RidgeCrest Earthquake Sequence (3 main events M5.4, 6.4, 7.1)
 +
** Amatrice:
 +
*** another Italian event
 +
** El Mayor
 +
*** multi-segment but low benefit as recordings are all far field
 +
** Kumamoto (M7, 2016)
 +
*** great event to do
 +
*** could use existing Japanese vmod
 +
*** part of blind simulation exercise at 2021 ESG6 (http://www.esg6.jp/index.html)
 +
*** issue is access to processed data (not done by PEER)
 +
** Taiwan (Meinong, M6.4 2016)
 +
*** issue is access to processed data (not done by PEER)
  
 
== Group validation of RidgeCrest Earthquake Sequence ==  
 
== Group validation of RidgeCrest Earthquake Sequence ==  
 
* New velocity model: South Sierra Nevada (ssn) TODO: post ppt from Rob (2/21/2020)
 
* New velocity model: South Sierra Nevada (ssn) TODO: post ppt from Rob (2/21/2020)
 +
** Focus on near-field (avoid basins); rough proposed box TODO: post pic from Christine (3/18/2020)
 
** Need new GFs for UCSB and new crustal amps for EXSIM (completed 3/25/2020)
 
** Need new GFs for UCSB and new crustal amps for EXSIM (completed 3/25/2020)
 
* Use PEER dataset for ground motions.  
 
* Use PEER dataset for ground motions.  

Latest revision as of 22:42, 25 March 2020

Intro

We discussed the following focus areas to consider for 2020

  • Evaluate source variability and parameters space sensitivity
    • Will require more than 50 realizations maybe 200?
    • Focus in-close where path effects are not controlling the results (within 1 fault length)
      • M-A
      • Area changes (DDW only?)
      • Rupture velocity
      • Dip
      • Moment-rate (using duration)
  • Evaluate simulations for other metrics
    • Validation metrics that already exists (duration, inter-frequency correlation, Anderson GOF, etc.)?
    • New validation metrics?
  • Events to consider/discussion notes
    • Group validation of RidgeCrest Earthquake Sequence (3 main events M5.4, 6.4, 7.1)
    • Amatrice:
      • another Italian event
    • El Mayor
      • multi-segment but low benefit as recordings are all far field
    • Kumamoto (M7, 2016)
      • great event to do
      • could use existing Japanese vmod
      • part of blind simulation exercise at 2021 ESG6 (http://www.esg6.jp/index.html)
      • issue is access to processed data (not done by PEER)
    • Taiwan (Meinong, M6.4 2016)
      • issue is access to processed data (not done by PEER)

Group validation of RidgeCrest Earthquake Sequence

  • New velocity model: South Sierra Nevada (ssn) TODO: post ppt from Rob (2/21/2020)
    • Focus on near-field (avoid basins); rough proposed box TODO: post pic from Christine (3/18/2020)
    • Need new GFs for UCSB and new crustal amps for EXSIM (completed 3/25/2020)
  • Use PEER dataset for ground motions.
    • Some are missing; PEER notified 2/19/2020.
  • Start with M5.4, then 6.4, then 7.1
    • M5.4 SRC from Graves (3/25/2020)