Earthquake Early Warning Current progress and future prospects

1
Earthquake Early WarningCurrent progress and
future prospects
In the context of the ongoing CISN early warning
project A collaboration between
U.S. Geological Survey
UC Berkeley
Caltech
SCEC/USC
Swiss Seismological Service
2
Earthquake early warning
Operational systems
around the world
Systems under development
Japan
Romania
Taiwan
United States
Turkey
Switzerland
Mexico
Italy
Greece
India
3
CISN statewide testing
August 2006 July 2009
Goal evaluation of early warning methodologies
in real-time
Aug 06
Year 1 Initial real-time outputs from EEW
algorithms Year 2 Adding uncertainty estimates
website display Year 3 Evaluation of past and
future performance
Currently at 14 mths
? specifications for implemented early warning
system
Jul 09

• equipment requirements (stations)
• telemetry
• processing

4
CISN statewide testing
August 2006 July 2009
Goal evaluation of early warning methodologies
in real-time
Algorithm implementation for a statewide system
Algorithms
1. Onsite warning single station Caltech/U.
Taiwan
2. Virtual Seismologist network
approach Caltech
3. ElarmS network approach UC Berkeley
4. others?
5
System overview and current status
Implementation
event processing statewide
web archive and display statewide
waveform processing network hubs
seismic networks
output parameters
Onsite warning

• predicted ground shaking onsite

VS Onsite ElarmS
Caltech
website EEW event database graphical display
Virtual Seismologist
Caltech/USGS

• location
• magnitude
• predicted ground shaking everywhere

Caltech/ETH
VS Onsite ElarmS
ElarmS
Berkeley/USGS

• location
• magnitude
• predicted ground shaking everywhere

USC/SCEC
Berkeley
6
Santa Rosa
August 2, 2006 Magnitude 4.7
Wurman, Allen Lombard, JGR, in press
7
Santa Rosa
August 2, 2006 Magnitude 4.7
Time trigger 3
ShakeMap
AlertMap
8
Sparse stations
ShakeMap
away from the metropolitan areas
Slows the availability of event information Is
this a small to moderate local event of the
beginnings of a large regional event?
San Simeon
February 20, 2007 Magnitude 3.7
9
San Simeon
AlertMap
February 20, 2007 Magnitude 3.7
ShakeMap
10
Finite faults
Realtime backprojection of station displacement
to fault slip
e.g. Chi-Chi earthquake
Realtime mapping of rupture geometry
from Yamada and Heaton
11
What do we need?
some preliminary thoughts
1. Low noise stations the density of current
realtime telemetry stations in the metropolitan
areas is fine need more stations along the
plate boundary where large events may nucleate
2. Best possible data transmission zero delay
is optimal any telemetry delay is lost warning
time current packetization is greatest source
of delay newer systems designed to reduce this,
e.g. Q330
3. Displacement and velocity instruments
particularly important for largest earthquakes
at what cost?
12
Cost
Is early warning too expensive?
Cost of retrofitting buildings
Barker Hall 14 mill
Barrows Hall 20 mill
Hearst Mining 80 mill (base isolation)
Wurster Hall 30 mill
UC Berkeley SAFER program 20 mill per year for
20 years Spent 900 mill so far
13
Cost
Is early warning too expensive?
Cost of early warning
1. The geophysical network
Installation of 600 new stations 6 - 30
mill Network operation 2 - 6 mill per year
2. Transmitting warning information
Existing technologies - weather radios -
satellite and internet communications - wireless
networks
3. Educational program

• Early warning response is specific to individual
  users
• Set in broader context of earthquake preparedness

14
Summary
Earthquake early warning is here developed
and implemented around the world
Existing systems have diverse users private
industry, public authorities, individual
protection
Realtime testing is underway in California
encouraging results for a statewide system
Full implementation will likely require a
doubling of existing geophysical networks yet
the cost is comparable to retrofitting a
single office building