Annual Meeting - SCEC.org

Poster Abstracts
agrees well with the ground truth damage assessment map indicated with polygonal zones of 3 different damage levels,
compiled by the government of New Zealand.
GEOMETRY OF FAULT SLIP ZONES AT DEPTH FROM QUANTITATIVE ANALYSIS OF SEISMIC
CATALOGS: METHOD AND RESULTS FOR THE SAN JACINTO FAULT ZONE (B-064)
Y. Ozakin and Y. Ben-Zion
We develop a quantitative method for estimating the geometry of active fault zones at seismogenic depth using typical entries
of earthquake catalogs. In contrast to prior methods that make various assumptions on geometrical properties of the structure
(e.g. sets of planes or fractal network), our method does not presuppose any shape. Each recorded earthquake is assigned a
Gaussian likelihood function with central coordinates and standard deviations in different directions corresponding to the
hypocenter location and errors. The likelihood function is given a weight corresponding to the event size. The sum of the
likelihoods of all earthquakes reported for a given region, with normalization to account for different event numbers in
different locations, provides a likelihood field for the occurrence of slip patches in the examined volume. High contiguous
regions of likelihood values represent large fault zone sections. Synthetic catalogs are used to develop and test normalization
methods that account for typical spatio-temporal variations of seismicity, and to identify characteristic likelihood shapes
corresponding to typical types of structural heterogeneities (e.g. stepovers, parallel faults). In the present application the
method is applied to estimate geometrical properties of the seismicity in the San Jacinto fault zone environment. Preliminary
results show high general complexity and a dipping trend with increasing depth of seismic slip patches along a NW-SE
direction. The continuing work will focus on performing in-depth analysis of segmentation and continuity along different
sections of the fault, and especially the trifurcation area and Hemet stepover region.
AN EARTHQUAKE RUPTURE FORECAST INVERSION APPLIED TO FAULT SYSTEMS IN CALIFORNIA (B-
112)
M.T. Page, E.H. Field, and K.R. Milner
Previous fault-based rupture forecasts for California have been plagued with magnitude “bulges” and fail to simultaneously
fit all available data. We present results from an inversion-based methodology being developed for the 3rd Uniform California
Earthquake Rupture Forecast (UCERF3) that simultaneously satisfies available slip-rate, paleoseismic event-rate, and
magnitude-distribution constraints. Using a parallel simulated-annealing algorithm, we solve for the rates of all ruptures that
extend through the seismogenic thickness on major mapped faults in California. This inverse approach eliminates the need for
expert-opinion voting on rupture rates themselves, allowing for more transparent and reproducible results than previous
earthquake rupture forecasts. The inversion methodology also allows for the incorporation of multi-fault ruptures, which
eliminates the magnitude-distribution misfits that were present in earlier models. Furthermore, we show that the inversion
solutions match slip-rate and paleoseismic event-rate data better than previous models.
THE INFLUENCE OF AMBIENT FAULT TEMPERATURE ON FLASH-HEATING BEHAVIOR (A-094)
F. Passelegue, D.L. Goldsby, T.E. Tullis, and O. Fabbri
Recent friction experiments demonstrate that rocks become dynamically weak due to extreme heating and thermal
degradation of the strength of microscopic asperity contacts on a slip surface – i.e., due to ʽflashʼ heating. Here we test the
effect of an increase of ambient temperature Tf of a slip surface, up to those occurring at the base of the seismogenic zone, on
the flash-heating behavior of quartzite, India gabbro and Westerly granite. Experiments were performed in a 1-atm rotaryshear
apparatus at ambient humidity, wherein a fixed annulus was slid against a rotating plate of the same material at a
normal stress of 5 MPa. Samples were heated by hot plates adjacent to the annulus and plate. Average sliding surface
temperatures were monitored with thermocouples in contact the annulus and plate, and with a handheld IR detector. The
sliding velocity was initially stepped from 10 µm/s to ~0.36 m/s, then subsequently decreased to ~0.25 m/s over ~45 mm of
slip, then to zero within several mm of slip at the end of the test. The friction coefficient below a critical weakening velocity
Vw of 0.10 - 0.15 m/s obtained values of 0.6 - 0.9 depending on the rock type. Above Vw, a dramatic 1/V decrease in friction
with velocity was observed, with values of the friction coefficient as low as 0.4 for quartzite and 0.5 for granite and gabbro at
the highest slip rate. The experiments demonstrate that increasing values of Tf cause 1) an increase in friction coefficient at
quasi-static slip rates, 2) an increase in the value of Vw, and 3) an increase in the value of the weakened friction coefficient
observed for a given velocity V > Vw. The unexpected increase in Vw with increasing values of Tf is consistent with flashheating
theory (Rice, 2006) when appropriate higher-T values of thermal conductivity, heat capacity and contact stress in the
theoretical expression for Vw are adopted.
214 | Southern California Earthquake Center