When Good Quakes Go Bad Or How Do Small Quakes End Up Being Big And Why Cant Something Be Done About

1
When Good Quakes Go BadOrHow Do Small Quakes
End Up Being BigAndWhy Cant Something Be Done
About It?
David Oppenheimer Presented to CISN Steering and
Advisory Committees at Caltech, 27 September 2007
2
Where do Noise Events Come from?

• Case 1
• Bad radio repeater at Oroville causes
  interference on 7 stations
• Automated picks associate into Cape Mendocino
  quake
• Coda durations last for 300-400s (M5)

3
Where do Noise Events Come from?

• Case 2
• Small quake occurs at Geysers
• Automated picks are ok, but Oroville noise adds
  2 bad picks.
• No codas on BG stations. Oroville again causes
  codas of 300-400s (M5)

4
Chronology of Event 40201796
5
Solutions

• Make picker smarter
• RT differentiation of noise from signal
• Squash bad magnitudes faster
• Stifle ShakeCast generation

6
Making systems smarter
7
Making systems smarter
8
Making systems smarter
Or, better, eliminate the problem
9
What have we learned?

• Fine tune criteria for generating ShakeMaps
• ShakeCast needs to get smarter
• Need to dynamically remove bad stations from
  system
• Clients ascribe (too?) high level of reliability
  to CISN products.
• CISN staff should work with clients to explain
  limitations