Insights Into Tectonic Regionalisation in Seismic Hazard Analysis and Computational Methods for Delineation

1. Insights Into Tectonic Regionalisation in Seismic Hazard Analysis
and Computational Methods for Delineation
Yen-Shin Chen1, Graeme Weatherill2, Marco Pagani3,
GEM Working Group on Tectonic Regionalisation
1PhD Student, Understand & Managing Extremes School, Pavia, Italy
2GEM Hazard Scientist, Pavia, Italy
3GEM Hazard Coordinator, Pavia, Italy
26th, August, 2014 2ECEES Istanbul, Turkey

2. Precedents for Tectonic Regionalisation
Jordan (1981)
USGS (1997)
Douglas (2007) Kagan et al. (2010)

3. Tectonic Regionalisation in Seismic Hazard Analysis
GMPEs
• Active shallow crust
• Subduction Interface/Intra-slab
• Stable Continent Region
• Volcanic
• Non-subduction deep focus
• Oceanic Travel Path
• Q, attenuation
• κappa
Magnitude Scaling Relations
• Active shallow/subduction/SCR
• Crustal age/structural maturity
• Orogenic?
• High/low stress drop
Tectonic Proxies
• MMAX (e.g. EPRI, 1994) or Corner Magnitude
• Recurrence (e.g. Truncated | Tapered G-R)
• Seismogenic depth distribution?

4. Tectonic Regionalisation: Basic Objective Questions, Comments
Current Regionalisation methods are mainly performed by
subjective judgements
– basis for delineation of zones remains unclear
Our goal is to design an quantitative and objective definition of
a regionalisation scheme/methodology
– a global data driven and transparent, scalable, replicable approach

5. Possible Relevant Datasets for Seismic Hazard Regionalisation
Seismic Moment Rate
QLG Global Composite
(from Brian Mitchell, personal comm.)
Shear Wave Velocity Variation at Depth of 175 km
(Mooney, 2012)
Global Continent Crust Age Map(modified from CCGM)

6. Softening the Decisions: Implementation in a Fuzzy Framework
not Active Active
10-9 Ex: strain Rate
Definitely
Definitely
not
• Crisp Classification

7. Softening the Decisions: Implementation in a Fuzzy Framework
not Active Active
10-9 Ex: strain Rate
Definitely
Definitely
not
•Fuzzy Classification
Softening the boundary
10-9 Ex: strain Rate
degree of belief
Definitely
Definitely
not
Based on knowledge …
ex:
If Strain rate is high, then
high chance to be Active
• Crisp Classification

8. Example of Fuzzy Framework
Smoothed Seismic
Moment Rate
Exploratory Data Analysis
Log10(Mo)(N-myr-1-km-2) Log10(Mo)(N-myr-1-km-2)
Normalized recording number
Stable Active
Log Mo(N-m yr-1 km-2)
Stable Continent Regions by Johnson, 1994

9. Example of Fuzzy Framework
Smoothed Seismic Moment Rate
Fuzzy Framework:
If moment rate is high, and
S velocity var. is low
Deg. of “Active” is high
Log Mo(N-m yr-1 km-2)
Shear Wave Velocity Variation
Shear wave velocity var.(%)
(Mooney, 2012)
Stable
Active
Shear wave velocity variation
Normalized recording number
Exploratory Data Analysis

10. Result - Degree of Being Active
SCR regions defined by Johnson, 1994
Degree of being “Active”

11. Defuzzication– Active/ Stable Delineation
Stable
Active
SCR regions defined by Johnson, 1994

12. Tectonic Regionalisation Model Drive From our Study
Stable Continent Region, Non-Craton
Stable Continent Region, Craton
Subduction
Active Continent Shallow Region
Stable Oceanic Region Active Oceanic Region

13. Please attribute to the GEM Foundation with a link to -
www.globalearthquakemodel.org
Except where otherwise noted, this work is licensed under:
creativecommons.org/licenses/by-nc-nd/4.0/