Capturing global seismic potential
from GEM’s fault, quake, and strain
datasets
Ross S. Stein GEM Scientific Board chair, ...
 Instrumental quakes
 Active faults
 Historical quakes
 Strain rate
 Ground motion
prediction equations
 Exposure
Po...
ISC‐GEM Catalog: 20,000 earthquakes, 1900‐2009
Cut‐off 
Magnitudes:
M≥7.50 since 1900
M≥6.25 since 1918
M≥5.50 since 1965
...
NewOld
ISC-GEM Catalog: New magnitudes, locations, and depths for all
quakes
Guatemala
Puerto Rico
Colombia
Guatemala
Puer...
ISC-GEM vs. Centennial catalogs: Northern Chile
ISC-GEMCentennial
Chile
Pacific
Ocean
Chile
Pacific
Ocean
Storchak et al
(...
How were these results achieved? Uniform relocations
Pre-1918 Centennial locations events were simply adopted from reliabl...
Centennial ISC-GEM
Uniform magnitudes
ISC-GEM uses a unified MW magnitude scale, originating from just four sources
Ms, mb...
ISC-GEM Catalog: The signature of plate tectonics
km
continental
collision
subduction
Rifting
subduction
Backarc
spreading
(Colored by
fault type)
Newly
discovered
active faults
in Myanmar
Sieh et al
(in preparation)
ISC-GEM
Seismic
Catalog
(ora...
A new tool for geologists to upload their faults to GEM
K. Berryman, A. Chistophersen, N. Litchfield, and GEM
IT staff
Seismic source interpretation and upload tool
K. Berryman, A. Chistophersen, N. Litchfield, and GEM
IT staff
GEM Faulted Earth: Uploaded fault traces and regions so far covered
New
Zealand
U.S.
JapanHimalayas
Australia
Alaska
Ocean...
Japan
Taiwan
Java
Tonga
Aleutians
New
Guinea
Philippines
Fault depth (km)
Sumatra
GEM Faulted Earth
40,000 km of
subductio...
GEM Large Historical
Earthquake Catalog
AD 1000-1900
Paola Albini
and
Roger
Musson,
(in prep.)
GEM Large Historical
Earthquake Catalog
AD 1000-1900
Paola Albini
and
Roger
Musson,
(in prep.)
GEM Large Historical
Earthquake Catalog
AD 1000-1900
Paola Albini
and
Roger
Musson,
(in prep.)
Indonesia: GEM Historical Catalog has 12 times more quakes than
NOAA
NOAA Catalog: 6 earthquakes GEM Catalog: 75 earthquak...
1900‐2009
Instrumental and historical catalogs complement each othe
Istanbul
GEM Historical Catalog
M≥7 during AD 1000‐1900
1900‐2009 Istanbul
Instrumental and historical catalogs complement each othe
Instrumental and historical catalogs complement each other
Instrumental and historical catalogs complement each other
GEM Historical Catalog
M≥7 during AD 1000‐1900
White et al., 2004  Ms  = 7.60
Zúñiga et al., 1997  Mw = 8.10
Nishenko & Singh, 1987 (13 MDPs)
ISC‐GEM Mw = 7.69 ± 0.68
En...
Avni et al., 2002
(macroseismic)
Ben-Menahem et al., 1976 ML =
6.2
ISC-GEM Mw = 6.29 ±
0.21
Location uncertainties: 1927 J...
Who needs a Global Strain Rate Model?
Quake rate Strain rate
Wasatch
fault
Cas‐
cadia
If all accumulating strain were rele...
GEM Global Strain Rate Model reveals earthquake potential and
active faults
Warmer colors
indicate high
strain and
thus hi...
GEM Global Strain Rate Model reveals earthquake potential and
active faults
Warmer colors
indicate high
strain and
thus hi...
GEM Strain Rate Model and ISC-GEM Catalog across Eurasia
Strain rate (nanostrain/yr)
Strain rate and large 20th century ea...
Iran: GEM Strain Rate Model and ISC-GEM Catalog
Strain rate (nanostrain/yr)
I R A N
Tehran
Dubai
Tabriz
Baku
C A S P I A N...
Tehran
Northern Italy
Kreemer, in prep.
Pavia Pavia
Emilia
Romagna
ISC‐GEM quakes
Western Turkey
Istanbul
Istanbul
Izmir
Bursa
Kreemer, in prep.
ISC‐GEM quakes
Northern Philippines
Manila
ISC‐GEM quakes
Baguio
Kreemer, in prep.
Chile and ArgentinaKreemer, in prep.
Santiago
ISC‐GEM quakes
Santiago
2011 Landscan
Population
density
1900-2009 ISC-GEM
Earthquake Catalog
Santiago
Global Strain Rate Model helps regional modelers hunt for local faults
Global Strain Rate Model helps regional modelers hunt for local faults
SHARE Model Faults
Global Strain Rate Model helps regional modelers hunt for local faults
Because inadequate fault data is perhaps the greatest weakness of hazard models
Because inadequate fault data is perhaps the greatest weakness of hazard models
GEM’s response to seismic hazard debate
Yan Y. Kagan, David D. Jackson, and Robert J. Geller
Mark W. Stirling
Seth Stein, ...
John Adams
Canada Natl
Haz Model
John Adams
Canada Natl
Haz Model
Mark Stirling
NZ Natl Haz
(Powell co‐PI)
Oliver Boyd
Cen...
Ned Field
Calif Haz
Model
Mark Petersen
US Nat’l Haz Model
David
Jackson
PSHA critic
Graeme
Weatherill, GEM
Danijel Schorl...
• Build a testable global earthquake activity 
rate model using smoothed seismicity 
and GEM strain rate
• Abandon assignm...
More
quakes
Less
quakes
Forecast quake rate (log scale)
Bird, Jackson, Kagan, Kreemer
and Weatherill (in prep.)
GEM Earthq...
Forecast quake rate (log scale)
More
quakes
Less
quakes
GEM Earthquake Activity Rate (GEAR) retrospective forecast for pos...
GEM Strain Rate Model and ISC‐GEM CatalogVolkan Sevilgen (USGS)
Some of the GEM team
Strain Rate
Kreemer, in prep.
Standard deviation of the Strain Rate
Kreemer, in prep.
Signal-to-noise ratio of the strain rate
Kreemer, in prep.
Dilatation (red) and contraction (blue) rates
Kreemer, in prep.
Greece and Crete
Kreemer, in prep.
Athens
Athens
Gulf of Corinth
ISC‐GEM quakes
Regional Examples
Kreemer, in prep.
ISC‐GEM quakes
Regional Examples
Kreemer, in prep.
ISC‐GEM quakes
Regional
Examples
Kreemer, in prep.
Regional Examples
Kreemer, in prep.
Earthquake potential at millennial, century, and decade
scales
Number of
studies
Paola Albini (INGV Milan) and Roger Musso...
Magnitude
Paola Albini (INGV Milan) and Roger Musson (British Geological Survey), Principal Investigators
GEM Large Histor...
11 eqs in NOAA Catalogue vs 20 eqs in GLHE Catalogue
India
Tibet
Pakistan
Himalayas: GEM Catalog has twice the quakes of t...
Regional Examples
Kreemer, in prep.
Regional Examples
Kreemer, in prep.
GEM Earthquake Activity Rate Model
Bird, Jackson, Kagan, Kreemer & Weatherill (in prep.)
Stephen Young, Endurance Re: “Dat...
Winning blend: 
62.5% catalog / 37.5% strain
Mixing parameter, aPure strain
rate
Pure seismic
catalog
Thus far, the best p...
Internal outline
‣ Themes
‣ Difficulty of inferring future quakes from past quakes
‣ Potential to prioritize global seismi...
Capturing global seismic potential from GEM’s fault, quake, and strain datasets Ross S. Stein GEM Scientific Board chair, ...
Capturing global seismic potential from GEM’s fault, quake, and strain datasets Ross S. Stein GEM Scientific Board chair, ...
Capturing global seismic potential from GEM’s fault, quake, and strain datasets Ross S. Stein GEM Scientific Board chair, ...
Upcoming SlideShare
Loading in …5
×

Capturing global seismic potential from GEM’s fault, quake, and strain datasets Ross S. Stein GEM Scientific Board chair, and U.S. Geological Survey geophysicist, GEM REVEAL 2013

7,701 views

Published on

Capturing global seismic potential from GEM’s fault, quake, and strain datasets
Ross S. Stein GEM Scientific Board chair, and U.S. Geological Survey geophysicist

Published in: Technology
0 Comments
3 Likes
Statistics
Notes
  • Be the first to comment

No Downloads
Views
Total views
7,701
On SlideShare
0
From Embeds
0
Number of Embeds
5,439
Actions
Shares
0
Downloads
32
Comments
0
Likes
3
Embeds 0
No embeds

No notes for slide

Capturing global seismic potential from GEM’s fault, quake, and strain datasets Ross S. Stein GEM Scientific Board chair, and U.S. Geological Survey geophysicist, GEM REVEAL 2013

  1. 1. Capturing global seismic potential from GEM’s fault, quake, and strain datasets Ross S. Stein GEM Scientific Board chair, and U.S. Geological Survey geophysicist
  2. 2.  Instrumental quakes  Active faults  Historical quakes  Strain rate  Ground motion prediction equations  Exposure Population Buildings  Vulnerability Damage data Fragility functions  Decision tools  Loss amplifiers  Risk transfer tools  Retrofit cost- benefit tools  Risk rankings and indices Hazard (faulting & shaking) Exposure and Vulnerability (deaths, damage, dollars) Social Impact (mitigation actions) GEM’s GLOBAL DATASETS, a €10M investment
  3. 3. ISC‐GEM Catalog: 20,000 earthquakes, 1900‐2009 Cut‐off  Magnitudes: M≥7.50 since 1900 M≥6.25 since 1918 M≥5.50 since 1965 15,000  seismic  bulletins  from 300  institutions
  4. 4. NewOld ISC-GEM Catalog: New magnitudes, locations, and depths for all quakes Guatemala Puerto Rico Colombia Guatemala Puerto Rico Colombia Depth (km) Storchak, Di Giacomo, Bondár, Engdahl, Villaseñor, Lee, Harris and Bormann (submitted)
  5. 5. ISC-GEM vs. Centennial catalogs: Northern Chile ISC-GEMCentennial Chile Pacific Ocean Chile Pacific Ocean Storchak et al (submitted) Bolivia Bolivia
  6. 6. How were these results achieved? Uniform relocations Pre-1918 Centennial locations events were simply adopted from reliable sourcesIn ISC-GEM, all but 1900-1903 shocks were relocated based on arrival time data 7Storchak et al (submitted)
  7. 7. Centennial ISC-GEM Uniform magnitudes ISC-GEM uses a unified MW magnitude scale, originating from just four sources Ms, mb, mB, Mw, UK, others Storchak et al (submitted)
  8. 8. ISC-GEM Catalog: The signature of plate tectonics km continental collision subduction Rifting subduction Backarc spreading
  9. 9. (Colored by fault type) Newly discovered active faults in Myanmar Sieh et al (in preparation) ISC-GEM Seismic Catalog (orange quakes are shallow, blue are deep) T H A I L A N D B A N G L A - D E S H M Y A N M A R A N D A M A N S H I M A L A Y A S GEM Faulted Earth Kelvin Berryman (GNS Science) Principal Investigator
  10. 10. A new tool for geologists to upload their faults to GEM K. Berryman, A. Chistophersen, N. Litchfield, and GEM IT staff
  11. 11. Seismic source interpretation and upload tool K. Berryman, A. Chistophersen, N. Litchfield, and GEM IT staff
  12. 12. GEM Faulted Earth: Uploaded fault traces and regions so far covered New Zealand U.S. JapanHimalayas Australia Alaska Oceanic transforms Hawaii K. Berryman, A. Chistophersen, K. Haller, Y. Awata, N. Litchfield, P. Tapponnier, and K. Sieh Europe South America Indonesia Middle East Central Asia Middle EastStill to come:
  13. 13. Japan Taiwan Java Tonga Aleutians New Guinea Philippines Fault depth (km) Sumatra GEM Faulted Earth 40,000 km of subduction zones SLAB 1.0 of Hayes & Wald (2010)
  14. 14. GEM Large Historical Earthquake Catalog AD 1000-1900 Paola Albini and Roger Musson, (in prep.)
  15. 15. GEM Large Historical Earthquake Catalog AD 1000-1900 Paola Albini and Roger Musson, (in prep.)
  16. 16. GEM Large Historical Earthquake Catalog AD 1000-1900 Paola Albini and Roger Musson, (in prep.)
  17. 17. Indonesia: GEM Historical Catalog has 12 times more quakes than NOAA NOAA Catalog: 6 earthquakes GEM Catalog: 75 earthquakes NOAA = National Geophysical Data Center/ World Data Service Significant Earthquake Database Java Myanmar Paola Albini and Roger Musson (in prep.)
  18. 18. 1900‐2009 Instrumental and historical catalogs complement each othe Istanbul
  19. 19. GEM Historical Catalog M≥7 during AD 1000‐1900 1900‐2009 Istanbul Instrumental and historical catalogs complement each othe
  20. 20. Instrumental and historical catalogs complement each other
  21. 21. Instrumental and historical catalogs complement each other GEM Historical Catalog M≥7 during AD 1000‐1900
  22. 22. White et al., 2004  Ms  = 7.60 Zúñiga et al., 1997  Mw = 8.10 Nishenko & Singh, 1987 (13 MDPs) ISC‐GEM Mw = 7.69 ± 0.68 Engdahl & Villa., 2002 Mw = 7.40  Abe & Noguchi, 1983 Ms  = 7.70 Calibration issues: 1903 M~7.7 Mexico-Chiapas M. Stucchi (in prep.) 500 km
  23. 23. Avni et al., 2002 (macroseismic) Ben-Menahem et al., 1976 ML = 6.2 ISC-GEM Mw = 6.29 ± 0.21 Location uncertainties: 1927 Jericho earthquake Max Stucchi (in prep.) Amman Jerusalem Tel Aviv
  24. 24. Who needs a Global Strain Rate Model? Quake rate Strain rate Wasatch fault Cas‐ cadia If all accumulating strain were released seismically, the quake rate should be proportional to strain rate 2000-2011 GPS velocities used by Kreemer et al for the GEM Strain Rate Model Gutenberg-Richter a-value from declustered ANSS catalog (Arnaud Mignan, ETH Zurich) Western United States
  25. 25. GEM Global Strain Rate Model reveals earthquake potential and active faults Warmer colors indicate high strain and thus high quake rates Kreemer et al (in prep.) 4,000 velocities in 2004 model, 20,000 in GEM’s Canada South America Mexico Caribbean Alaska U.S.
  26. 26. GEM Global Strain Rate Model reveals earthquake potential and active faults Warmer colors indicate high strain and thus high quake rates Kreemer et al (in prep.) China TaiwanIndia Java Japan Sumatra Iran Philippines Tibet Indian Ocean New Guinea East    African rift 4,000 velocities in 2004 model, 20,000 in GEM’s
  27. 27. GEM Strain Rate Model and ISC-GEM Catalog across Eurasia Strain rate (nanostrain/yr) Strain rate and large 20th century earthquakes are correlated M≥6 earthquakes GEORGIA IRAN TUNISIA TIBET CRETE KYRGYZSTAN INDIA ITALY PAKISTAN JORDA N ALGERIA SERBIA TURKEY But strain exceeds seismicity in Himalayas, Tehran, Baku, North Anatolian fault, Greece ISAREL
  28. 28. Iran: GEM Strain Rate Model and ISC-GEM Catalog Strain rate (nanostrain/yr) I R A N Tehran Dubai Tabriz Baku C A S P I A N A Z E R B A I J A N M≥6 earthquakes 1.5 million 2.1 million 12.0 million 1.5 million Population Kreemer (in prep.) Shiraz
  29. 29. Tehran
  30. 30. Northern Italy Kreemer, in prep. Pavia Pavia Emilia Romagna ISC‐GEM quakes
  31. 31. Western Turkey Istanbul Istanbul Izmir Bursa Kreemer, in prep. ISC‐GEM quakes
  32. 32. Northern Philippines Manila ISC‐GEM quakes Baguio Kreemer, in prep.
  33. 33. Chile and ArgentinaKreemer, in prep. Santiago ISC‐GEM quakes Santiago
  34. 34. 2011 Landscan Population density 1900-2009 ISC-GEM Earthquake Catalog Santiago
  35. 35. Global Strain Rate Model helps regional modelers hunt for local faults
  36. 36. Global Strain Rate Model helps regional modelers hunt for local faults
  37. 37. SHARE Model Faults Global Strain Rate Model helps regional modelers hunt for local faults
  38. 38. Because inadequate fault data is perhaps the greatest weakness of hazard models
  39. 39. Because inadequate fault data is perhaps the greatest weakness of hazard models
  40. 40. GEM’s response to seismic hazard debate Yan Y. Kagan, David D. Jackson, and Robert J. Geller Mark W. Stirling Seth Stein, Robert Geller, and Mian Liu September/October 2012 2012
  41. 41. John Adams Canada Natl Haz Model John Adams Canada Natl Haz Model Mark Stirling NZ Natl Haz (Powell co‐PI) Oliver Boyd Central US Haz Model Marco Pagani GEM Marco Pagani GEM Seth Stein,  PSHA critic Alex  Allmann Munich Re Laura Peruzza European  Models Laura Peruzza European  Models Bring warring parties together on a mountain top for 3 days to develop new  strategies, and to agree on tests of seismic hazard assessment Mark  Leonard,  Australia Natl Haz Model Ray Durrheim So.  Africa  Model Anke Friederich European geology Anke Friederich European geology USGS Powell Center Workshops: Harnessing the community
  42. 42. Ned Field Calif Haz Model Mark Petersen US Nat’l Haz Model David Jackson PSHA critic Graeme Weatherill, GEM Danijel Schorlemmer GEM Testing and  Evaluation Facility Martin Käser Munich Re Martin Käser Munich Re Mark Stirling NZ Natl Haz (Powell co‐PI) Mark Stirling NZ Natl Haz (Powell co‐PI) Morgan Page Calif Haz Model Gavin Hayes Subduction zones USGS Powell Center Workshops: Harnessing the community
  43. 43. • Build a testable global earthquake activity  rate model using smoothed seismicity  and GEM strain rate • Abandon assignments of ‘maximum  earthquake magnitude’ for a scientifically  defensible alternative • Test post‐1996 ground motions against the 1996 US Natl Seismic Hazard model Powell Center Groups: Essential projects
  44. 44. More quakes Less quakes Forecast quake rate (log scale) Bird, Jackson, Kagan, Kreemer and Weatherill (in prep.) GEM Earthquake Activity Rate (GEAR) retrospective forecast for post-2005 M≥5.75 quakes Best forecast is from 37.5% GEM Strain Rate Model and 62.5% pre-2005 seismic catalog
  45. 45. Forecast quake rate (log scale) More quakes Less quakes GEM Earthquake Activity Rate (GEAR) retrospective forecast for post-2005 M≥5.75 quakes Best forecast is from 37.5% GEM Strain Rate Model and 62.5% pre-2005 seismic catalogBird, Jackson, Kagan, Kreemer and Weatherill (in prep.)
  46. 46. GEM Strain Rate Model and ISC‐GEM CatalogVolkan Sevilgen (USGS)
  47. 47. Some of the GEM team
  48. 48. Strain Rate Kreemer, in prep.
  49. 49. Standard deviation of the Strain Rate Kreemer, in prep.
  50. 50. Signal-to-noise ratio of the strain rate Kreemer, in prep.
  51. 51. Dilatation (red) and contraction (blue) rates Kreemer, in prep.
  52. 52. Greece and Crete Kreemer, in prep. Athens Athens Gulf of Corinth ISC‐GEM quakes
  53. 53. Regional Examples Kreemer, in prep. ISC‐GEM quakes
  54. 54. Regional Examples Kreemer, in prep. ISC‐GEM quakes
  55. 55. Regional Examples Kreemer, in prep.
  56. 56. Regional Examples Kreemer, in prep.
  57. 57. Earthquake potential at millennial, century, and decade scales Number of studies Paola Albini (INGV Milan) and Roger Musson (British Geological Survey), Principal Investigators GEM Large Historical Earthquake Catalog: 832 M≥7 quakes during AD 1000-1900
  58. 58. Magnitude Paola Albini (INGV Milan) and Roger Musson (British Geological Survey), Principal Investigators GEM Large Historical Earthquake Catalog: 832 M≥7 quakes during AD 1000-1900 Earthquake potential from millennial, century, and decade record
  59. 59. 11 eqs in NOAA Catalogue vs 20 eqs in GLHE Catalogue India Tibet Pakistan Himalayas: GEM Catalog has twice the quakes of the NOAA Catalog
  60. 60. Regional Examples Kreemer, in prep.
  61. 61. Regional Examples Kreemer, in prep.
  62. 62. GEM Earthquake Activity Rate Model Bird, Jackson, Kagan, Kreemer & Weatherill (in prep.) Stephen Young, Endurance Re: “Data Quality was identified as a serious issue in 2011. The Tohoku event had not been fully considered in the event set, leaving decided gaps in knowledge.” From the Font Line, Bermuda Reinsurance (2012) Stephen Young, Endurance Re: “Data Quality was identified as a serious issue in 2011. The Tohoku event had not been fully considered in the event set, leaving decided gaps in knowledge.” From the Font Line, Bermuda Reinsurance (2012)
  63. 63. Winning blend:  62.5% catalog / 37.5% strain Mixing parameter, aPure strain rate Pure seismic catalog Thus far, the best predictors of future quakes are past quakes and strain rate Information score (binary bits per quake)  GEM Earthquake Activity Rate (GEAR) experiments find ‘seismic gap theory’ (next quakes in areas of high strain and fewest past quakes) fails Bird, Jackson, Kagan, Kreemer & Weatherill (in prep.)
  64. 64. Internal outline ‣ Themes ‣ Difficulty of inferring future quakes from past quakes ‣ Potential to prioritize global seismic threat if data and methods are uniform ‣ Potential of testable PSHA if it is global and as uniform as possible ‣ Data ‣ GEM Faulted Earth ‣ GEM Historical earthquake catalog ‣ ISC‐GEM Instrumental quake catalog ‣ GEM Strain Rate Model ‣ Powell process: Collaborative, open, problem‐solving ‣ GEAR ‣ Landing ‣ In the earth sciences, data trumps all, and so…  ‣ The GC datasets are a gift to science, commerce, and humanity

×