UCERF3 Deformation Model Approaches
Tim Dawson (California Geological Survey)*
Diablo Canyon Seismic Source Characterization (SSC) Model Update – SSC Workshop 2
November 6 – 7, 2012 San Luis Obispo, CA
*with contributions from Tom Parsons (USGS) and Ray Weldon (U of O)
Proper SSHAC Equipment
From
Hanks et al. (2009)
SSHAC Questions:
• How uncertainty is incorporated in individual models
and between models?
• How is geologic, geodetic, and seismicity data
incorporated?
• How is aseismic deformation treated?
• How should these models be considered for a sitespecific study such as DCPP?
UCERF3:
Uniform California Earthquake Rupture Forecast
Working Groups on California Earthquake Probabilities (WGCEPs)
?
UCERF2
2008
UCERF3
2013
System-Level Science: the integration of diverse sources of knowledge
about the constituent parts of a complex system with the goal of
obtaining an understanding of the system's properties as a whole
UCERF Ingredients:
Physics – scaling laws
UCERF3 Faults
Components of the Uniform Earthquake Rupture Forecast 3
Components of the Uniform Earthquake Rupture Forecast 3
UCERF2
UCERF3
ABM
NeoKinema
On fault
Geologic
Faults
Finish this
In addition to a geologicallybased model, geodetic data
has been used directly to
produce more quantitative,
kinematically consistent
geodesy-based models
Off Fault
C Zones
Zeng
Distribution of Geologic Slip Rate sites
Distribution of GPS observations
Geologic Model: Statewide Slip Rate Database
•
315 reported geologic rates in
compilation
•
173 UCERF3 fault sections
with a reported geologic rate
(out of ~350)
•
Only 62 fault sections with
multiple reported geologic slip
rates
•
50 fault sections with slip rates
over different time intervals
Not surprisingly, dataset is relatively
sparse spatially and temporally!
Geologic Deformation Model
•
Best estimate rates provided for
each UCERF3 fault section
•
Every fault section also assigned
slip rate bounds (from reported
uncertainties, USGS Q-faults
database slip rate category, or
correlation to recency of activity
•
Uncertainties for geologic slip
rates are difficult to quantify.
Quality of slip rates used in
model vary wildly, uncertainties
sometimes reported, reported
rate may not be representative
of fault and geologic expertise
applied for best estimate rate.
•
Documented in Table 1 of
Appendix B
UCERF3 Block Models
For geodesy-based block
deformation models, a Statewide
block model was defined by the
geologists and geodecists.
Subset of geologic slip rates
provided to geodetic modelers to
use as additional constraints in the
geodetic models
Slip rate bounds from geology also
used in some models (e.g. Zeng)
Block Models only provide slip
rates for faults coincident with
block boundaries…slip rates for
non-block boundary faults set at
50% of intra-block strain rate.
UCERF3 Deformation Models evaluation
UCERF3 Deformation Models evaluation
UCERF3: Aseismicity
UCERF3 updated creep database (Appendix D of UCERF3 report, in
review)
•
Incorporates alignment arrays, creep meter measurements, plus creep
modeled from InSAR.
UCERF3: Aseismicity
UCERF2 approach: Creep
rate/slip rate = aseismicity
factor
However, creep is thought to
be shallow phenomenon and
decreases with depth
If this is the case, then
UCERF2 over-reduced the
moment rate on faults.
UCERF3: Aseismicity
New UCERF3 moment
reduction model takes into
account that creep rate
decreases with depth and is
also a function of creep rate (v)
and total slip rate (V)
Model based on formulation of
Savage and Lisowski (1993)
In our model, for a creep rate
(v) about ½ of the slip rate,
creep depth is about half of
locking depth
As creep rate (v) approaches
the slip rate (V), the creep rate
at depth becomes more like the
creep rate at the surface
UCERF3: Aseismicity
Moment reduction is applied in two
ways:
Applied as an area reduction for
faults with moment reductions
factors of < 0.9
Applied as a slip rate reduction
for faults with moment reductions
of > 0.9 (e.g., the fastest creeping
faults like the creeping section of
the SAF)
Default value of 0.1 for faults
without specific creep data
SSHAC Questions:
How uncertainty is incorporated in individual models and
between models?
Four models considered for UCERF3, range of
models provide a range of slip rates that (hopefully)
span the uncertainties.
SSHAC Questions:
How is geologic, geodetic, and seismicity data incorporated?
Both geologic and geodetic data are incorporated in the
deformations models. Models include a “pure” geologic model,
plus 3 others types of models that use geodetic data, plus
geology to varying degrees.
SSHAC Questions:
How is aseismic deformation treated?
New compilation of creep data for California (Weldon
and others, Appendix C)
New model for moment reduction from creep rates that
takes into account creep decreases with depth and is
rate dependent.
Faults without specific creep data get a default 0.1
moment reduction
SSHAC Questions:
How should these models be considered for a sitespecific study such as DCPP?
Caveats:
• UCERF3 is a Statewide model, seismic sources are
often coarsely represented and the model is an
approximation of earthquake rates at a system-wide
level.
• Earthquakes rates are calculated in a uniform manner,
for consistency across the State.
• Caution that seismic sources, slip rates, and model
results should not be simply extracted for use in sitespecific studies.
SSHAC Questions:
How should these models be considered for a sitespecific study such as DCPP?
Other Points:
• Geologic and Geodetic data represent observations
over different time scales. Differing resulting may
represent rate changes through time across the system
and should be considered because we may not be able
to tell which one is currently more appropriate to use.
• Some model results may be artifacts resulting from
over-simplications. Hazard may simply be shifting from
one fault system to another.
SSHAC Questions:
How should these models be considered for a sitespecific study such as DCPP?
However:
• UCERF3 has produced a number of databases (slip
rates, paleoseismic recurrence, paleoseismic slip,
creep compilations, historical seismicity) that may be
useful for other studies.
• Also, (as you will hear) different approaches used in
developing the deformation models may be useful for a
site-specific study in exploring the epistemic
uncertainties that a single model (e.g. geologicallybased deformation model) can’t produce.
SSHAC Questions:
How should these models be considered for a site-specific
study such as DCPP?
Finally:
Assuming that UCERF3 (or elements of it) is adopted in to the
USGS National Seismic Hazard Map, it will inevitably be used
as a point of comparison to site-specific studies, such as at
DCPP. Because the models are different (uniform Statewide vs
site specific), it will be important to recognize the models are
different and understand how and why the models are different.
Availability of UCERF3 Products
UCERF3 Main Report and Appendices:
• Currently in review
• Are available upon request
• Finalized by the end of 2012 (?)
Comments and Questions ?