Source Parameters of the 15 November 1995 Border Town, Nevada, Earthquake Sequence
GENE A. ICHINOSE1,
KENNETH D. SMITH,
and
JOHN G. ANDERSON1,
University Nevada Reno Seismological Laboratory
Mackay School of Mines
1Also at the University Nevada Reno, Dept. Geological Sciences
Mail Stop-174
Reno, NV, 89557-0141
phone (702) 784-4265
fax (702) 784-1833
email ichinose@seismo.unr.edu
url http://enigma.seismo.unr.edu
Bulletin Seismological Society of America, Vol 87. No. 3, pp. 652-667, June 1997
Abstract
The 15 November 1995, Border Town, Nevada earthquake sequence occurred west
of the California-Nevada border approximately 20 km northwest of Reno, Nevada.
The largest earthquake of the sequence (Mw=4.5) was
widely felt throughout the Reno-Sparks-Truckee region.
This event occurred
on a west dipping high angle normal fault at a depth of 14 km and shows
dip-slip motion on a preferred fault-plane orientation of strike N10°E,
dip -70°W, and rake -75°. We have relocated 27 aftershocks and one
foreshock of the event using records from the local network and two portable
digital instruments. The largest aftershocks also align along a N10°E
trend and define the preferred fault plane. All of the aftershocks occur within
a small volume with a 2- by 2-km horizontal extent and between depths of 10 and 14 km.
Simultaneous determination of Mo, fc, and kappa is made by fitting
spectra using various starting models (i.e., initial static stress drop and kappa) based
on an f-2 spectral shape. Mo ranges from
1018 to 1023 dyne cm for 14 events and hypocentral distances were
less than 15 km. We fit S-wave Spectra with different starting models to test the stability
of static stress drop. The results show that static stress drop and kappa converged to
different values depending on the starting model and the magnitude because of an attenuation-source
dimension trade-off. For example, a model starting with a static stress drop of 10 Bars
produced an average static stress drop of 27 Bars and an
apparent scaling breakdown below Mw 3. Starting the search at 50 Bars produced an average
static stress drop at 50 Bars produced an average static stress drop of 69 Bars and no
magnitude dependence. We cannot resolve any breakdown in self-similarity because we find plausible
spectral models for all magnitudes where static stress drop is nearly constant at 60 Bars.
Seismograms of the four largest aftershocks, Mw 3 to 3.4, were deconvolved
using smaller aftershocks as empirical Green's functions. Event radii estimated from
the pulse widths yield dynamic stress drop averaging 34 Bars, within the range of
the frequency-domain results.
Introduction: Historical to Modern Seismicity
An Mw earthquake on 15 November 1995 in the Border Town, Nevada, area was
felt throughout the Reno-Sparks-Truckee region (origin time = 20:34 UTC, latitude=39° 38.8'N,
longitude=120° 0.8'W, and depth = 13.7 km). It is the largest event along the eastern Sierra
Nevada since the September 1994 (Mw5.9) Double Spring Flat earthquake that occurred 40
miles south of Carson City, Nevada (dePolo et al., 1994). Reno and Border Town are near the
western margin of the Basin and Range Province, which is classically regarded as an area of
crustal extension. Rogers et al., (1991) reports a maximum average return period of 27 years for
M > 7 earthquakes in the Basin and Range based on the historical seismicity. In addition, Ryall (1977)
and Wallace (1987) estimate an average return period from 300 to 10,000 years for any individual
major range-bounding fault. The recurrence estimates, and relative uncertainties, for
major earthquakes and rupture of specific fault zones in the Reno-Carson City
area have not yet been bully developed.
The Reno area has experienced a number of these moderate to strong earthquakes. The
constraints on the locations and magnitudes of historical events are summarized by
dePolo et al. (1996) in a report on a scenario earthquake for the northern Nevada region. Altogether,
dePolo et al. (1996) identified 13 earthquakes with magnitudes greater than 6.0 in the Reno
area. The largest of these historical earthquakes, both in the 1860's are an M7 near
Pyramid Lake (Bell et al. 1979) and an M6.7 earthquake in the Olinghouse area (Toppozada et al., 1981).
Surface faulting from the Olinghouse earthquake has been recognized by Sanders et al. (1979). Moderate
to large earthquakes within close proximity to the Reno area could potentially cause severe damage
to structures and, possibly loss of life.
Figure 1 shows the locations of earthquakes from the University of Nevada Reno Seismological Laboratory
(UNRSL) catalog from 1960 through 1996 including the 1995 Border Town sequence. Also shown are
two historical moderate-sized earthquakes near Verdi, Nevada, that were strongly felt in Reno
and caused minor damage to structures. The Verdi earthquakes, M 6.4 February 1914 (Slemmons et al 1965)
and M 6.0 December 1948 (E. J. Bell, written communication., 1982; dePolo et al., 1996), locate in
an area of persistent modern seismicity near the 1995 Border Town sequence and may be an area
of concern for continued moderate-sized earthquakes. The 1966 M 6.0 Truckee, California,
earthquake is the most recent moderate-sized event that was felt strongly in the Reno area (Fig. 1.).
Displacement during the 1966 earthquake was left-lateral strike-slip on a northeast striking
high-angle fault (Kachadorrian et al., 1967; Greensfelder, 1968; Tsai and Aki, 1970; dePolo et al., 1996).
A study of the seismotectonics of the Truckee-Lake Tahoe region by Hawkins et al. (1986) includes
the first compilation of earthquake source parameters from three-component records for sources near
the Reno area. Hawkins et al. focused on the seismic hazard associated with local reservoirs along the
Truckee River drainage. They relied on the background seismicity recorded at fairly large distances
as the data set to study source parameters.
Portable digital recorders deployed within hours of the M 4.5 Border Town earthquake to supplement
the permanent network provide the first near-source digital recordings of earthquake activity within 20 km
of the Reno area. First, we relocate the aftershocks and obtain focal mechanisms to support interpretations
of tectonic processes north of the Reno area. Finally, the limited spatial extent of the aftershock sequence
allows us to more confidently assess source parameters over a range of seismic moments.
Regional Geology
Miocene-aged andesites and rhyolites, Pliocene- to
Pleistocene-aged basalts and tuffaceous rocks, and basin filled with Tertiary
and Quaternary sediments dominate the surface geology in most of the
northeastern Sierra Nevada near the Reno area. This stratigraphy represents active Tertiary extensional
tectonics (Thompson, 1964; Slemmons, 1966; Bonham, 1969; Lovejoy, 1969; Gilbert et al., 1973). Prior
to extensive Tertiary deformation and volcanism, the region was in a volcanic back-arc setting
of Mesozoic compressional tectonics along the subduction boundary (Speed, 1979; Oldow, 1983). The
basement rocks derived from this period are composed of Mesozoic-aged metasediments and metavolcanics
that are intruded by Cretaceous-aged granitic plutons of the Sierra Nevada. Some exposures of the pre-
Cretaceous sedimentary sequences remain as isolated roof pendants over these Mesozoic intrusive
throughout the region (Bonham, 1969; Oldow et al., 1984). The complex volcanic stratigraphy from
several deformation cycles and several periods of Pleistocene glaciation makes the Quaternary
deformation history difficult to interpret. Quaternary and Holocene faulting has been recognized
throughout the Border Town, Reno, Carson City, and Carson Valley areas (Bell, 1984). The
relationship between the active range-bounding normal faults, adjacent to and outboard of the
Sierran block, and northwest- and northeast-trending strike-slip faults of the Walker Lane belt
(Fig. 1) suggests a complex tectonic model to account for the distributed deformation process
(Stewart, 1988).
We find that the normal-faulting Border Town earthquake took place on a west-dipping high-angle normal
fault near the intersection of the Last Chance and Peavine Peak fault zones (Fig. 1). The Last Chance fault
zone extends from Verdi, Nevada, northward to the Diamond Mountains (Hawkins et al., 1986). The mapped length
of the main fault trace is 54 km (Burrnett and Jennings, 1962), and the sense of motion is not known. The
1948 Verdi earthquake has been located near the southern end of this fault zone (Fig. 1). The
Peavine Peak fault zone borders the north end of Peavine Peak and trends generally west-northwest,
possibly intersecting the Last Chance fault (LCF) at a high angle in the Border Town area. In the
Border Town-Lemmon Valley area just north of Reno, several north to northeast trending exposure of
Mesozoic metasediments and metavolcanics are bounded by suspected high angle Quaternary faults (Bell, 1984).
Data and Instrumentation
The Border Town sequence occurred in the northwestern section of the UNRSL
regional Western Great Basin Seismic Network (WGBSN). Data from two near-source,
( 13 km hypocentral distance), portable digital recorders BRD1 and BRD2 (Figs. 2 and 3)
and a permanent long period station at WCN (Fig. 1) at Washoe City, Nevada ( 44 km
hypocentral distance ), were used in the waveform analysis. RefTek portable recorders
were used with Teledyne S-13 1 Hz velocity sensors. Accurate timing was maintained using
GPS clocks, and the recorders
were configured to record at 100 samples/sec. The portable instruments recorded 228
aftershocks, and of these, only 28 events were large enough to trigger the WGBSN. The
main event was recorded on scale at WCN.
Figure 1.
Earthquake Relocations and Focal Mechanisms
We relocated 28 earthquakes of the Border Town sequence using a layered velocity
model (Table 1) and program FASTPONG (Herrmann et al., 1981). Figure 2 shows the relocated
epicenters on an aerial photo that illustrates the local relief, and Figure 3 shows
the focal mechanisms and two cross-sectional views. The velocity model is used
internally at the UNRSL for western Great Basin earthquakes. After an initial inversion
for the location, the resulting average station corrections were applied in the final
relocation (Table 2). Hypocenter depths were well constrained by P- and S-wave arrival times at
stations BRD1, BRD2, and BAB, which are located within 1 to 10 km of the main event (within one
focal depth).
TABLE 1. Velocity model used in hypocenter relocations and
waveform modeling
---------------------------------------------------------------------
Layer Depth 1 Thickness a Qa B QB p
(km) (km) (km/s) (km/s) (g/cc)
---------------------------------------------------------------------
1 0.0 1.0 4.0 50 2.3 150 2.20
2 1.0 1.0 5.2 150 3.0 200 2.40
3 2.0 3.0 5.5 300 3.2 600 2.50
4 5.0 23.0 6.0 300 3.5 600 2.75
5 28.0 12.0 6.5 1000 4.5 800 2.80
6 40.0 oo 7.8 1000 4.7 2000 3.32
---------------------------------------------------------------------
1 Depth to top of layer
TABLE 2. Earthquake Relocations
----------------------------------------------------------------------------------------------
Date UTC Latitude Longitude Depth Md #Picks Az gap Rms Dmin
Yr-Mo-Da Hr:Min:Sec (o) (o) (km) o (sec) (km)
----------------------------------------------------------------------------------------------
95-11-15 20:18:48.46 39N39.02 120W 0.80 12.48 3.0 13 51 0.05 9.49
95-11-15 20:33:58.86 39N38.78 120W 0.81 13.71 4.6 14 50 0.10 9.25
95-11-15 20:36:32.27 39N38.70 120W 0.84 11.43 2.3 12 74 0.06 9.12
95-11-15 20:38:22.69 39N38.90 120W 0.95 11.49 1.9 13 73 0.05 9.19
95-11-15 20:38: 9.30 39N38.64 120W 1.51 13.78 0.9 6 160 0.08 8.26
95-11-15 20:59:43.63 39N38.93 120W 1.07 12.00 2.1 13 73 0.04 9.08
95-11-15 21: 9:28.61 39N38.74 120W 1.06 13.25 2.2 14 73 0.11 8.90
95-11-15 21:22:16.91 39N39.04 120W 1.36 11.14 1.5 7 156 0.07 8.88
95-11-15 21:58:19.11 39N38.56 120W 0.72 13.13 3.5 14 74 0.08 9.14
95-11-15 23: 1: 1.36 39N38.80 120W 0.73 10.68 1.5 11 74 0.04 9.35
95-11-16 0: 7:20.60 39N39.01 120W 1.40 10.60 1.2 8 106 0.09 8.80
95-11-16 1:39: 8.45 39N39.03 120W 1.03 10.47 1.2 10 108 0.06 1.43
95-11-16 7:37:53.62 39N38.76 120W 0.41 10.49 2.5 14 74 0.07 1.90
95-11-16 8:44:38.89 39N38.96 120W 0.22 7.75 1.6 12 112 0.14 2.31
95-11-16 14:59:24.44 39N38.92 120W 0.68 10.78 3.8 15 50 0.06 1.67
95-11-16 14:60:17.30 39N39.07 120W 1.17 9.92 1.1 9 157 0.07 1.37
95-11-16 16:11:10.10 39N39.09 120W 1.28 10.25 1.1 10 107 0.09 1.33
95-11-16 16:16:45.27 39N38.83 120W 1.19 10.73 1.5 11 108 0.07 1.00
95-11-16 17:34:56.20 39N39.15 120W 0.95 10.32 1.1 10 108 0.07 1.67
95-11-16 18:47:17.98 39N38.81 120W 0.65 10.64 2.1 15 74 0.05 1.63
95-11-17 0:32: 6.59 39N38.90 120W 0.88 11.08 1.7 14 73 0.10 1.43
95-11-17 10:36: 3.37 39N39.27 120W 1.00 10.39 1.3 6 145 0.09 1.81
95-11-17 18:16:55.67 39N39.11 120W 0.34 11.48 3.4 15 50 0.09 2.28
95-11-17 19:48: 4.55 39N39.17 120W 1.18 10.10 1.4 11 93 0.06 1.52
95-11-20 9: 2:21.89 39N39.28 120W 1.14 11.94 1.6 11 72 0.04 9.41
95-11-25 14: 4:32.90 39N38.73 120W 0.87 12.22 1.9 12 74 0.04 9.11
95-11-30 19: 2:22.35 39N40.15 120W 3.10 7.14 2.3 3 121 0.05 3.76
95-12-18 22:40:29.89 39N39.10 120W 0.90 11.52 1.6 12 73 0.05 9.46
----------------------------------------------------------------------------------------------
The program FPFIT (Reasenberg et al., 1985). was used to calculate P-wave first motion mechanisms
for the 28 relocated events. The larger events show generally north-striking high angle normal-
faulting mechanisms (Fig. 3). At smaller magnitudes, the solutions are less constrained.
The entire sequence is confined to a small volume at a depth range of about 10 to 14 km with
with a horizontal extent of about 2 by 2 km (Fig. 3). There is an apparent northwest alignment to
the aftershock activity as shown in Figure 3, but all of the M > 3 events align along a
north-northeast strike direction (Figs. 3 and 4) consistent with the strike direction of the
steeply west-northwest-dipping fault plane from the short-period focal mechanisms. Also,
we found coherency in the WCN waveforms at periods < 1 second for 10 of the larger aftershocks,
supporting the consistency of the first motion mechanisms for those events. We could
interpret the scatter in the hypocentral distribution to the northwest to be activity in
the hanging wall block of the mainshock fault plane or attributed the scatter to poorly
constrained hypocenter locations. We interpret the operative fault plane to strike N10°E and
dip 70° to the west-northwest into the Sierran range front based on the near vertical
hypocenter distribution (Fig. 3) and presence of many active high angle faults in this region.
This fault geometry projects to the surface near a suspected Quaternary fault east of White Lake
(Figs. 2 and 3) mapped by Bell (1984). Figure 2 also includes the east-west error estimates of
the surface projection from the focal mechanism inversion. If this is the operative fault, then
it might imply that the LCF, to the west, is a steeply dipping fault most likely accommodating
strike-slip motion. If the LCF is an east-dipping normal fault with a dip of 60°, then the
Border Town event would be in the LCF footwall, but it is unlikely that the operative fault would
project into the LCF hanging wall. , the Border Town area is near the intersection of the
Peavine Peak and Last Chance fault zones (Fig. 1), suggesting an area of possible stress concentration that
may result in activity on a variety of faults and higher levels of background seismicity.
For the main event, synthetic seismograms were computed for a double-couple point source
in a layered half-space using a discrete wavenumber summation method (Zeng and Anderson, 1995). We
compared the WCN waveforms to an input fault orientation of strike N8°E, dip 69°W, rake -75°
and a seismic moment of 6.2x1022 dyne cm (Dreger, written comm., 1995). These source
parameters provide good fits to WCN waveforms at 1 Hz for all three components of ground motion (Fig. 4).
The modeled main event source parameters also agree with the UCB moment-tensor solution and short-period
focal mechanisms.
Scaling Relationships of Source Parameters
Numerous microearthquake source studies have shown an apparent stress drop-seismic moment scaling
breakdown for earthquake magnitudes less than 3 (e.g. Archuleta et al., 1982; Fletcher et al., 1986;
Aki, 1987). The fundamental questions are whether the inferred scaling breakdown is a site
effect or a property of the earthquake source. Dieterich (1986) provided a numerical model for the
nucleation of seismic slip based on rock mechanics where aseismic slip across some critical source
dimension is required before dynamic rupture. This has led to the suggestion of a minimum rupture
dimension, perhaps about 100 m (Archuleta et al., 1982; Aki, 1987), or a nucleation patch too small
to be detected (Abercrombie, 1995). Abercrombie (1995) postulates that a minimum source dimension
could be controlled by a geometrical constraint from the fault-zone geometry, similar to the constraint
of the seismogenic thickness of the crust and the effect it has on the Mo-L scaling
breakdown of large earthquakes (Shimazaki, 1996). Sacks et al. (1995) suggest that source scaling relations
could be controlled by quanta characteristics. Such models predict an apparent breakdown at
low magnitudes.
Alternatively, other studies have concluded that the apparent scaling breakdown at low magnitudes
reflects an attenuation effect. Hanks (1982), Frankel (1982), Anderson et al. (1984), and Frankel et
al. (1989) noted a correlation between the absence of high-frequency radiation in earthquake
spectra and the receiver site geology. These studies suggest that severe attenuation produced the
apparent scaling breakdown. In light of these findings, Anderson (1986) presented a model in which the
site and path attenuation parameters trade off with the corner frequency at low magnitudes. This
fundamentally limits the ability to resolve earthquake source properties with decreasing magnitude.
In a study of near surface effects, Abercrombie (1995) examined microearthquakes recorded both in a
deep borehole environment (2.5 km depth) and at the surface. Although an order of magnitude variation
in static stress drop was observed in that study over a range of magnitudes, earthquakes recorded on the
deep borehole instrument confirm the contribution of near-surface attenuation in source scaling relationships.
Furthermore, source parameters from extremely small seismic events recorded in an underground experiment
demonstrate that scaling relationships hold for magnitudes far less than that observed for tectonic
earthquakes (Gibowicz et al., 1991). Gibowicz and others recorded magnitudes between -3.6 and -1.9
which source radii from 0.3 to 1.0meters, which is far less than a suggested minimum source dimension
of 100 meters.
The resolution of source parameters with decreasing magnitudes is fundamental interest in the derivation
of earthquake scaling relationships and the determination of the stress drops of local earthquakes.
Estimates of seismic moments, source dimensions, attenuation, and static stress drops of 15 Border Town
earthquakes were determined by fitting near-source S-wave spectra. For the purpose of comparison,
seismic moments and dynamic stress drops were calculated for the four largest aftershocks using an
empirical Green's function technique. Since many source parameters cannot be uniquely determined
(Boatwritght, 1985; Anderson, 1986; Hough et al., 1995a; Hough 1995b), the ability to resolve the
stress drop at low magnitudes is examined based on a comparison of the results of the two methods.
Figure 2.
Figure 3.
Figure 4.
Spectral Analysis
Waveforms recorded on the near-source digital instruments were analyzed by modeling the S-wave
displacement spectra. We find best fits to an idealized one-corner spectral shape derived by
Boatwright (1978) following the methodology of Hanks and Wyss (1972), Lindley et al. (1992), and
Abercrombie (1995). The shape of the displacement spectra used in the forward model is
Equation 1.
where omega(f) is the Fourier displacement spectral amplitude at frequency f, omegao
is the low frequency asymptote, kappa is the spectral decay parameter, and fc is the
corner frequency. The high-frequency fall-off is fixed as (eta=2) to minimize the number of
unknown parameters. The Simplex method (Nelder and Mead, 1965) is used to find the best-fitting
model by varying the parameters omegao, kappa, and fc in equation (1). This
process involves a nonlinear optimization algorithm that minimizes the sum of the squared
residuals between the forward calculation and the observed spectral shape thorough an iterative
scheme (Caceci and Cacheris, 1984).
An example illustrating the modeling procedure is shown in Figure 5. We attempt to minimize the
P-wave contamination and isolate the SH-wave energy by using only the transverse component of the
S-wave arrivals. The (SH) time series are band passed at 0.02 to 50 Hz from velocity seismograms
that have been instrument corrected, rotated, and windowed prior to a standard fast Fourier transform
(FFT). We used a 1-sec Hanning-tapered (10%) window for the data and a half-second window just
before the signal window to establish the noise level. The noise spectrum was typically an order
of magnitude below the signal out to 20 Hz.
Figure 5.
There is good justification for fixing the parameter eta by assuming an f-2 spectral shape
at high frequencies (e.g., Aki, 1967; Brune, 1979; Lindley et al., 1992; Abercrombie, 1995). This
parameter characterizes the amount of high frequency energy radiated from the rupture initiation of
a circular fault (Brune, 1970). Other fault geometries can radiate different amounts of energy, but
f-1.5 must be the upper limit to prevent the source from radiating infinite energy. Aki
(1967) and Abercrombie (1995) concluded that a high-frequency fall off of f-2 was a good
average but also noted that significant variation in eta exists.
The shape of the displacement spectrum at high frequencies is characterized by an exponential decay,
where kappa is primarily controlled by the site effects, but also incorporates the path attenuation
as a function of distance (Anderson and Hough, 1991). As mentioned above, Anderson (1986) points out
a trade off between kappa and fc from uncorrected spectra. For instance, as the magnitude
decreases and the corner frequency increases for a Brune (1970) source spectrum with a constant static
stress-drop, the effect of kappa is to cause the spectrum to roll off at frequencies less than the
corner frequency. If this apparent corner, caused by attenuation, is instead assumed to be related
to the source dimension, one would mistakenly conclude that the constant stress drop scaling breaks
down.
A test for the stability of model parameters is made by checking if the model parameters converge
regardless of their starting values. Since omegao is usually well determined, we
first compute the seismic moment using Keilis-Brook (1957),
Equation 2.
where we assume a material density, rho=2.7g cm-3, and S-wave velocity, beta=3.3 km/s.
DELTA is the hypocentral distance, Rrho,theta = 0.6 is the S-wave RMS radiation pattern
over the focal sphere, and there is a free-surface correction of 1/2. A starting value for the
static stress drop is then used to determine a starting value for fc in equation 1. The
value for the fc parameter is therefore determined in the starting model by the static
stress drop and Mo through the relationship for an equidimensional fault (Brune, 1970).
Equation 3.
where c is either the P- or S-wave velocity near the source.
The initial values given to static stress drop and kappa for two models
are as follows: (model 1) static stress drop=10 bars and kappa 30 msec and
(model 2) static stress drop=50 bars and kappa = 40 msec. The inversions were
allowed to run for 50 iterations or until a maximum accepted model error of 1x10-4
cm-sec-2 was reached. By using different starting model parameters, we can explore
the trade off between fc and kappa. If static-stress drop can be well resolved, then
solutions should converge regardless of starting model parameters.
Seismic Moment, Stress Drop, and Kappa Results from Spectral Modeling
Tables 3a and 3b list the source parameters estimated from the spectral modeling. Figure 6
shows two separate plots of Mw versus static stress drop models for the
two initial static stress drops using only S-wave spectra. The Mw versus
kappa values are also plotted below. We did not find any systematic difference in the estimates
of kappa between stations BRD1 and BRD2. We also find that the different initial parameters
result in convergence to very different spectral parameters and a dependence between convergence
and magnitude.
Figure 6.
TABLE 3a. Earthquake Source Parameters 1 from initial /\-o of 10
bars
------------------------------------------------------------------------------------------
Station /\1 Julian UTC Oo fc k M0 Mw /\-o r
(km) Day hr:min (cm.s) (Hz) (ms) dynexcm bars (m)
------------------------------------------------------------------------------------------
WCN 45.6 319 2034 6.75x10-3 1.5 62 6.84x1022 4.5 47.4 859
BRD1 10.6 320 0139 1.51x10-5 9.3 27 3.54x1019 2.3 5.6 141
BRD1 10.6 320 0737 1.13x10-4 11.1 27 2.67x1020 2.9 71.6 118
BRD2 12.6 320 0737 1.15x10-4 8.8 28 3.21x1020 2.9 43.5 148
BRD1 7.7 320 0844 3.02x10-5 9.9 28 5.15x1019 2.4 10.0 131
BRD2 9.8 320 0844 3.71x10-6 18.9 26 8.11x1018 1.9 10.9 69
BRD1 10.6 320 1459 1.02x10-3 5.0 28 2.41x1021 3.5 60.3 260
BRD2 12.6 320 1459 5.10x10-4 5.1 27 1.43x1021 3.4 36.6 258
BRD1 10.0 320 1500 3.15x10-5 15.6 26 7.02x1019 2.5 52.7 84
BRD2 11.9 320 1500 1.68x10-5 13.6 26 4.46x1019 2.4 22.2 96
BRD1 10.3 320 1611 4.44x10-6 13.0 29 1.02x1019 1.9 4.4 101
BRD2 12.2 320 1611 2.32x10-6 16.0 28 6.31x1018 1.8 5.2 81
BRD1 10.8 320 1616 6.71x10-6 21.7 26 1.61x1019 2.1 32.7 60
BRD2 12.9 320 1616 3.09x10-6 18.4 26 8.83x1018 1.9 10.9 71
BRD1 10.5 320 1734 3.21x10-6 23.7 26 7.47x1018 1.9 19.8 55
BRD2 12.2 320 1734 2.57x10-6 20.2 26 6.96x1018 1.8 11.4 64
BRD1 10.8 320 1847 4.36x10-5 14.0 27 1.04x1020 2.6 57.2 93
BRD2 12.7 320 1847 4.01x10-5 14.7 32 1.13x1020 2.6 71.5 89
BRD1 11.2 321 0031 4.59x10-5 7.6 28 1.14x1020 2.6 10.0 171
BRD2 13.0 321 0031 2.78x10-5 8.6 28 8.06x1019 2.5 10.0 152
BRD1 11.2 321 1035 2.23x10-5 9.7 28 5.56x1019 2.4 10.0 135
BRD2 12.7 321 1035 9.59x10-6 8.0 30 2.72x1019 2.2 2.7 163
BRD1 11.7 321 1816 1.18x10-3 3.8 29 3.08x1021 3.6 32.7 346
BRD2 13.1 321 1816 4.80x10-4 4.5 28 1.40x1021 3.4 25.0 291
BRD1 10.2 321 1947 6.76x10-6 19.8 26 1.53x1019 2.1 23.6 66
BRD2 12.0 321 1947 4.81x10-6 16.2 26 1.28x1019 2.0 10.9 80
BRD1 8.1 334 1902 1.78x10-5 15.1 26 3.20x1019 2.3 21.9 86
BRD2 9.7 334 1902 1.24x10-5 15.9 26 2.67x1019 2.2 21.4 82
------------------------------------------------------------------------------------------
1 S wave horizontal components, 2 Hypocentral distance
TABLE 3b. Earthquake Source Parameters 1 from initial /\-o of 50
bars
-------------------------------------------------------------------------------------------
Station /\2 Julian UTC Oo fc k M0 Mw /\-o r
(km) Day hr:min (cm.s) (Hz) (ms) dynexcm bars (m)
-------------------------------------------------------------------------------------------
WCN 45.6 319 2034 6.75x10-3 1.6 65 6.84x1022 4.5 58.7 800
BRD1 10.6 320 0139 1.51x10-5 21.3 38 3.54x1019 2.3 67.9 61
BRD1 10.6 320 0737 9.38x10-5 10.5 38 2.21x1020 2.8 50.7 124
BRD2 12.6 320 0737 1.15x10-4 9.2 40 3.21x1020 2.9 50.0 141
BRD1 7.7 320 0844 3.02x10-5 19.5 36 5.15x1019 2.4 76.0 67
BRD2 9.8 320 0844 3.71x10-6 34.6 38 8.11x1018 1.9 66.3 38
BRD1 10.6 320 1459 1.02x10-3 5.4 38 2.41x1021 3.5 75.3 241
BRD2 12.6 320 1459 1.71x10-4 9.0 38 4.78x1020 3.1 69.7 144
BRD1 10.0 320 1500 3.15x10-5 20.1 38 7.02x1019 2.5 112.7 65
BRD2 11.9 320 1500 2.40x10-5 19.6 38 6.38x1019 2.5 95.9 66
BRD1 10.3 320 1611 4.44x10-6 30.4 38 1.02x1019 1.9 56.8 43
BRD2 12.2 320 1611 2.32x10-6 35.7 38 6.31x1018 1.8 57.1 36
BRD1 10.8 320 1616 6.71x10-6 29.4 38 1.61x1019 2.1 80.9 44
BRD2 12.9 320 1616 4.52x10-6 28.9 38 1.29x1019 2.0 61.9 45
BRD1 10.5 320 1734 3.21x10-6 36.2 38 7.47x1018 1.9 70.1 36
BRD2 12.2 320 1734 5.05x10-6 29.1 38 1.36x1019 2.0 66.6 45
BRD1 10.8 320 1847 4.34x10-5 15.7 38 1.04x1020 2.6 79.7 83
BRD2 12.7 320 1847 3.97x10-5 14.3 38 1.12x1020 2.6 65.5 91
BRD1 11.2 321 0031 4.95x10-5 11.9 41 1.23x1020 2.7 41.2 109
BRD2 13.0 321 0031 2.35x10-5 15.0 38 6.81x1019 2.5 45.9 87
BRD1 11.2 321 1035 1.72x10-5 18.3 38 4.29x1019 2.4 52.3 71
BRD2 12.7 321 1035 4.56x10-6 29.9 38 1.29x1019 2.0 68.2 44
BRD1 11.7 321 1816 1.14x10-3 4.9 39 2.96x1021 3.6 68.5 267
BRD2 13.1 321 1816 4.80x10-4 6.2 38 1.40x1021 3.4 64.7 212
BRD1 10.2 321 1947 6.76x10-6 28.7 38 1.53x1019 2.1 72.0 45
BRD2 12.0 321 1947 4.81x10-6 30.3 38 1.28x1019 2.0 71.1 43
BRD1 8.1 334 1902 1.78x10-5 24.2 38 3.20x1019 2.3 90.5 54
BRD2 9.7 334 1902 1.24x10-5 25.5 38 2.67x1019 2.2 88.3 51
-------------------------------------------------------------------------------------------
1 S wave horizontal components, 2 Hypocentral distance
Seismic moments for these earthquakes range from 6 x 1018 to 7x1022 dyne
cm. Moment magnitudes are systematically a tenth of a magnitude unit smaller at station
BRD1 than BRD2, possibly from the assumption in radiation pattern that depends on the location
of the stations with respect to the fault plane geometry. This difference could also be caused
by a site effect.
The difference between the static stress drop and kappa values for the two different initial conditions
indicates that different combinations of fc and kappa values can fit the S-wave
spectra equally well. This trade-off between the model parameters prevents the unique determination
of static stress drop. For events with magnitudes less than 3, static stress drop is only resolved
to within 3 to 113 bars, and kappa values at BRD1 and BRD2 could only be resolved to within 26 and 41
msec. The results show a weaker trade off for the larger aftershocks for magnitudes between 3 and 4
and therefore produce a smaller uncertainty of 25 to 75 bars and an average of 58 bars. The
Mw 4.5 mainshock was the only event that is not affected by the starting conditions and
retained a stable static stress drop of 53± 5 bars. This is expected since the attenuation at
the corner frequency of this event is not significantly affected by kappa. These results in this
analysis reveal that most static stress drop estimates are at least greater than the minimum starting
values of model 1. The maximum limits of static stress drop estimates for the smaller magnitudes
can actually be driven by numerical conditions (i.e., maximum allowable misfit and maximum iterations)
rather than driven by any data constraints. We do not see this for well-constrained, larger-magnitude
events that always returned to a stable static stress drop value under different numerical conditions.
The estimates for path attenuation made by measuring kappa also depend on the starting model. Both
stations have kappa approximately 27 msec for model 1 and 38 msec for model 2. We did not
find any difference in kappa between BRD1 and BRD2. The estimate of kappa at WCN is much
larger at 63 msec. Although observations of kappa at a single station tend to increase with
epicentral distance, in this case, there is also a possible contribution from differences in site
effect. Also, the specific path from Border Town to WCN crosses a geothermal area south of Reno,
which could contribute to the large WCN kappa value.
Empirical Green's Function Analysis
In the previous section, we modeled Mo and static stress drop with a
quasi-dynamic source model and without considering site effects. The empirical Green's function
(EGF) technique accounts for the site and path effect directly by taking the ratio of a
small- and large magnitude colocated event pair with the same ray paths. This technique
isolates the source properties of the larger event and has been applied in a number of other studies
(e.g., Mueller, 1985; Frankel et al., 1986; Mori et al., Hough et al., 1991, 1995a). This
technique also takes into account the possibility that the rupture might have some dynamic
property and is constrained to a specific spectral shape.
We windowed P- and S-wave arrivals to determine the source parameters of four Border Town
aftershocks using the EGF technique. We chose event pairs with the largest
magnitude differences and with similar waveform shapes. Earthquake pairs include a large and small
event with at least a difference of one magnitude unit. We feel that the proximity
of hypocenters from the relocations, within a cluster, justifies using EGF analysis for events with similar
waveforms (Frankel et al., 1986). The windowed waveforms were Hanning tapered (10%) before deconvolution
(Fig. 7).
The deconvolution is performed using a time-domain technique rather than by spectral division
(Sipkin et al., 1992; Gurrola et al., 1995). Time-domain deconvolution has the advantage over
spectral division techniques in that we can estimate the trade off between the smoothness of the
source time function and the measurement of the event parameters, in particular, the pulse width. In
contrast, spectral division traditionally uses a water-level threshold to fill spectral "holes"
and can be a more subjective process (Gurrola et al., 1995).
A sharp pulse is identified after the deconvolution process and the source dimension is estimated by
measuring the half-duration pulse width tau1/2. The pulse width is measured from the
onset of the pulse to the first zero crossing on the deconvolved source time function. The
source radius is then determined from tau1/2 with a modification for the source
directivity,
Equation 4.
where c is either the P- or S-wave velocity. The rupture velocity is assumed to be 0.8beta. The
angle from the strike of the rupture direction to the azimuth of the station theta is estimated to be
less than 45° for a station located directly above a normal fault. Since the actual rupture
direction is unknown, we fix theta to 45°. For a unilateral rupture up-dip, the source radius
can theoretically vary by a factor of 2 from directivity, causing about a factor of 10 uncertainty in
stress drop.
Seismic Moment and Stress Drop Results from Empirical Green's Function Analysis
Empirical Green's function analysis was performed for only four events because Green's functions
could not be found for smaller events. The results produced both simple and complex source time
functions (Fig. 7). Events 13 and 20 show simple pulses, whereas events 15 and 23 appear to be
complex ruptures composed of two subevents separated by about 0.1 sec. The initial pulses are invariably
short, and all are about the same width, implying a fault radius of 220 meters. We believe
this is about the smallest dimension resolvable in the presence of attenuation.
The seismic moment is computed by integrating the source time function and multiplying by
the moment of the Green's function (Mori et al., 1990), shown as the shaded area in Figure 7.
The area under the pulse is substituted for the low-frequency asymptote in equation 2. Mw
from this method does not differ significantly from estimates determined from the spectrum.
Figure 7.
Dynamic stress drops are estimated using the relationship (7 Mo/ 16 r-3)
(Eshelby, 1957; Brune, 1970), where r is calculated from equation (4). If r is an upper bound
to the radius the the values of dynamic stress drop are lower than static stress drop. Figure 8b
shows dynamic stress drop as a function of Mw. We report the dynamic stress drop and
seismic moment from the initial subevent of the source time functions in Table 4, so the
seismic moment and dynamic stress drop are expected to be slightly lower than estimates in the
spectral analysis for complex events. The P-wave dynamic stress drop for events 15 and 23 are 32 and 47
bars, respectively, and the S-wave dynamic stress drop is 16 bars for event 23. P-wave and S-wave
dynamic stress drop values for events 13 and 20 range from 18 to 72 bars. These dynamic stress drop
values are within the range found in the previous spectral analysis section.
Discussion
Uncertainties and Possible Errors for the Determination of Event Stress Drops
There is not enough resolution of the corner frequency below M 3 to either rule out or confirm
a breakdown in scaling relationship. Starting model 1 yields a magnitude dependence of stress
drop on moment, while starting model 2 has no dependence. The difference is caused by a trade-off
between fc and kappa. At low magnitudes, the minimum in the curve fitting algorithm is
replaced by a low-lying plateau where a range of parameters satisfies the same spectral shape. The
range of allowed values of stress drop spans an order of magnitude for magnitudes near 2. As
magnitude increases, the trade-off becomes less severe and is largely gone above M 3. Thus, we
can only report a wedge shape range of possible values of stress drop for magnitudes less than 3 that
result from the difference between the two starting models. Previous studies, with better resolution
at low magnitudes, do not indicate a scaling breakdown in the small-magnitude range, but
a large scatter in stress drop also persists in these results (e.g., Abercrombie 1995; Hough 1995b).
Estimates of stress drop for large magnitudes in Nevada also need to be resolved for scaling
relationship in this region. Considering Figure 9, the stress drop estimates from this article
and other Basin and Range earthquakes are consistent with stress drops in southern California and
elsewhere.
Using the mean radiation pattern rather than an SH wave radiation pattern correction can
produce an uncertainty as high as a factor of 4 in seismic moment in the low-frequency
range at which omegao is measured (Vidale, 1989). The seismic moments in
Tables 3a and 3b show about a factor of 2 increase at BRD1, but this is not considered significant
since a factor of 2 in seismic moment cannot account for an observed stress drop scatter of a
factor of 10.
The state of stress as a function of seismogenic depth and tectonic environment has been investigated
(e.g., Sibson, 1974), but its relationship to earthquake stress drop remains ambiguous. If the
tectonic stress is driving static stress drop, then regions of thrust faulting might be expected to have a
relatively higher stress drop than normal-faulting regions. A weak dependence has been found by
McGarr (1984), and Lindley et al. (1992) found a conflicting dependence between stress drop and
tectonic environment. An alternative hypothesis would place the dependence of stress drop on tectonic
environment as secondary relative to the frictional strength of the fault, at least for
earthquakes as deep as the events in this study. Numerical simulations using a rate-dependent fault
friction produce a scatter in the Mo-L scaling law over two orders of magnitude in seismic
moment (Madariaga et al., 1996). These simulations also demonstrated that a fault embedded in a homogeneous
medium can spontaneously become complex, revealing features like self-healing pulses found by Heaton (1990)
and premature fault locking, a concept introduced by Brune (1970). Thus, some of the range of
observed stress drop may be irreducible, and the relationships to the faulting process may remain
ambiguous until there is a large increase in high quality data.
Comparison between Time-Domain and Frequency Domain Techniques
The comparison of static stress drop from the spectral analysis and the dynamic stress drop from the EGF
analysis provides an addition constraint on stress drop estimates. The estimates of stress drop of
four events examined using both techniques are nearly the same. This result is significant because
different assumptions are applied in each method. The EGF analysis makes no assumption of the
source spectral shape, while the frequency domain inversion is constrained to one corner frequency
and a f-2 spectral fall-off. Differences between Mo and stress drop are
possibly due to directivity effects that were not well constrained. Hough et al. (1995a) found
little difference in directivity corrected and noncorrected stress drop measurements but suggests
that directivity effect can become quite large in the direction of rupture. Their study also
points out the subjectivity of measuring tau1/2 in terms of picking the narrowest pulse
width or from the zero crossings.
Comparison of Attenuation Results
The range of kappa values are similar to those in the western Nevada (Hough et al., 1989). The
site geology at BRD1 is granitic, and BRD2 lies on volcanic rocks. The kappa at WCN is similar
to western Nevada (Mina, NV) and the Imperial Valley, California (Anderson, 1991). Hough et al. (1989)
interpret the larger kappa values in western Nevada relative to southern California as a result
of an increase in the near surface attenuation.
Comparison with Previous Stress Drop Studies
The range of stress drops measured from tectonic earthquakes in the Basin and Range intermountain
region, the Cajon Pass in southern California, and from this study are compared in Figure 9. Stress
drop estimates from intermountain region normal-faulting earthquakes are comparable with the
results from this study. Three M > 6 normal-faulting earthquakes of the Fairview Peak-Dixie Valley,
Nevada, sequence have estimated stress drops ranging from 60 to 70 bars (Doser, 1986). Average
dynamic stress drop for aftershocks of the 1983 Ms 7.3 Borah Peak, Idaho, earthquake
are 33 ± 16 and 77 ± 52 bars for two individual fault segments (Boatwright, 1985). Mayeda
et al. (1996) estimated dynamic stress drops in the western US with attenuation corrections based
on regional coda envelopes from broadband stations. They show a variance in stress drop from 4 to 88
bars (average is 30 bars) for Basin and Range earthquakes.
Hawkins et al. (1986) used spectral analysis on 17 events near the Dog Valley Fault, Truckee, California,
recorded at WCN for Mw 0.5 to 4.8. Their stress drops at WCN range from 0.005 to 133 bars,
with more than 78% below 1 bar. We estimate a strong attenuation effect (kappa=64msec) in this
study along similar path to WCN; therefore, stress drops for events below M 4 in Hawkins et al. (1986)
are most likely underestimated and therefore not included in Figure 9.
Figure 8.
Figure 9.
Conclusions
The November 1994 M 4.5 Border Town, Nevada, earthquake most likely occurred on a N10°E striking,
steeply west-northwest dipping (70°), high angle fault. Simple projection of the fault plane to
the surface approximates the trace of a suspected Quaternary fault (Bell, 1984), which implies, if this
is the operative fault, that the Last Chance fault, to the west, is also high angle fault
accommodating strike-slip motion or normal motion or both.
Stress drop estimates from aftershock S-wave spectra do not appear to have a strong scaling in the
Mw 1.8 to 4.5 range. A test for the stability of stress drop reveals a range of stress drop
that converges at higher magnitudes ( around 60 bars ). Since we find that this dominant scatter in
stress drop depends on magnitude and the starting model, we infer that this is caused from an attenuation
source dimension trade off. Empirical Green's function analysis of only the four largest aftershocks
supports the stability of these stress drop estimates above M 3. Both time-domain and frequency domain
stress drop estimates are nearly constant at 60 bars down to M 3, but the estimates scatter over an
order of magnitude for M < 3. The Mw 4.5 mainshock has a stable stress drop of 60 bars,
which is within the scatter found in other Basin and Range studies.
Acknowledgments
We acknowledge Glenn Biasi and others involved with instrument deployment and
spending time servicing the sites. We would like to thank Yuehua Zeng for the
waveform modeling codes. Phase data was provided by
University Berkeley California, Northern California Earthquake
Data Center and Bill Walter. This manuscript has improved through the
suggestions of two anonymous reviewers.
This work was made possible through financial support
provided by U.S. Geological Survey NEHRP grant 1434-94-G-2479.
References
Abercrombie, R. E. (1995). Earthquake source scaling relationships
from -1 to 5 ML using seismograms recorded at 2.5-km
depth, J. Geophys. Res., 100, 24015-24036.
Aki, K. (1967). Scaling law of seismic spectrum, J. Geophys.
Res. 72, 1217-1231.
Aki, K. (1987). Magnitude-frequency relation for small earthquakes: A clue
to the origin of fmax of large earthquakes, J. Geophys.
Res. 92, 1349-1355.
Anderson, J. G. and S. E. Hough (1984). A model for the shape of
the Fourier amplitude spectrum of acceleration at high frequencies,
Bull. Seism. Soc. Am. 74, 1969-1994.
Anderson, J. G. (1986). Implication of attenuation for studies of
the earthquake source, in Earthquake Source Mechanics, Editor,
M. Ewing, Geophysical Monograph 37, American Geophysical
Union 6, 311-317.
Anderson, J. G. (1991). A preliminary descriptive model for the distance
dependence of the spectral decay parameter in southern California,
Bull. Seism. Soc. Am., 81, 2186-2193.
Archuleta, R., E. Cranswick, C. Mueller, and P. Spudich (1982). Source
parameters of the 1980 Mammoth Lakes, California, earthquake sequence,
J. Geophys. Res. 87, 4595-4607.
Bell, J. W. (1984). Quaternary fault map of Nevada Reno sheet, Nevada
Bureau of Mines and Geology map 79.
Bell, E. J. and D. B. Slemmons (1979). Recent crustal movement in the Sierra Nevada-Walker Lane
region of California-Nevada Part II. The Pyramid Lake right-slip fault zone segment of the
Walker Lane, Tectonophysics 52, 571-583.
Boatwright, J. (1978). Detailed Spectral analysis of two small
New York state earthquakes, Bull. Seism. Soc. Am. 68, 1117-1131.
Boatwright, J. (1985). Characteristics of the aftershock sequence
of the Borah Peak, Idaho, earthquake determined from digital
recordings of the events, in Proceedings of Workshop XXVIII,
on the Borah Peak, Idaho, Earthquake, U.S. Geol. Surv., Open
File Rep. 85-290A, 311-344.
Bonham, H. F. (1969). Geology and mine deposits of Washoe and Story Counties, Nevada,
Nevada Bureau of Mines and Geology Bulletin 70.
Brune, J. N. (1970). Tectonic stress and the spectra of seismic
shear waves from earthquakes, J. Geophys. Res. 75, 4997-5009.
Brune, J. N. (1971). Correction, J. Geophys. Res. 76, 5002.
Burnett, J. L. and C. W. Jennings (1962). Chico sheet, Geologic
map of California, Calif. Division Mines and Geology, scale
1:250,000
Caceci, M. S. and W. P. Cacheris (1984). Fitting curves to
data, Byte Magazine, May 1984, 340-362.
dePolo, D., K. Smith, R. Anooshehpoor, C. M. dePolo, and K. Priestley (1994). Double
Spring Flat earthquake, western Nevada, 12 September 1994 (abstract), Am. Geophys. Union 1994
Fall Meeting Special Session.
dePolo, C.J., J.G.Rigby, G.L. Johnson, S. L. Jacobson, and J. G. Anderson (1996). Planning
scenario for a major earthquake in western Nevada, Nevada Bureau of Mines and Geology, Special
Publication 20.
Dieterich, J. (1986). A model for the nucleation of earthquake slip,
in Earthquake Source Mechanics, Editor, M. Ewing,
Geophysical Monograph 37, American Geophysical Union 6, 37-47.
Doser, D. I. (1986). Earthquake processes in the Rainbow
Mountain-Fairview Peak-Dixie Valley, Nevada, Region 1954-1959,
J. Geophys. Res. 91, 12572-12586.
Eshelby, J. D. (1957). The determination of the elastic field of
an ellipsoidal inclusion and related problems, Proc. R. Soc.
London A241, 376-396.
Fletcher, J. B., F. L. Vernon, J. N. Brune, T. C. Hanks, and J.
Berger (1986). The effects of attenuation on the scaling of
source parameters for earthquakes at Anza, California, in
Earthquake Source Mechanics, Editor, M. Ewing, Geophysics Monograph 37,
American Geophysical Union 6, 331-338.
Frankel, A. (1982). The effects of attenuation and site response on the spectra of
microearthquakes in the northeastern Caribbean, Bull. Seism. Soc. Am. 72, 1379-1402.
Frankel, A. and L. Wennerberg (1989). Microearthquakes spectra from the Anza, California, seismic
network: site response and source scaling, Bull. Seism. Soc. Am. 79, 581-609.
Frankel, A., J. Fletcher, F. Vernon, L. Haar, J. Burger, T.
Hanks, and J. Brune (1986). Rupture characteristics and
tomographic source imaging of Ml 3 earthquakes near Anza,
southern California, J. Geophys. Res. 91, 12633-12650.
Gibowicz, S. J., R.P. Young, S. Talebi, and D. J. Rawlence (1991). Source parameters of seismic events at the
underground research laboratory in Manitoba, Canada: scaling relations for events with moment magnitude
smaller than -2, Bull. Seism. Soc. Am. 81, 1157-1182.
Gilbert, C. M. and M. W. Reynolds (1973). Character and chronology of basin development, western
margin of the Basin and Range Province, Geol. Soc. Am. Bull. 84, 2489-2510.
Greensfelder, R. (1968). Aftershocks of the Truckee, California, earthquake of September 12, 1966,
Bull. Seism. Soc. Am. 58, 1607-1620.
Gurrola, Harold G., Eli Baker, and J. Bernard Minster (1995).
Simultaneous time domain deconvolution with application to
receiver functions, Geophys. J. Intl. 120, 537-543.
Hanks, T. C. and Max Wyss (1972). The use of body-wave spectra in
the determination of seismic-source parameters, Bull. Seism.
Soc. Am. 62, 561-589.
Hanks, T. C. (1982) fmax, Bull. Seism. Soc. Am. 72, 1867-1879.
Hawkins, F. F., R. LaForge, and R. A. Hansen (1986). Seismotectonic study of the
Truckee/Lake Tahoe Area Northeastern Sierra Nevada, California for Stampede, Prosser Creek, Boca, and
Lake Tahoe Dams, Seismotectonic Rept. No. 85-4, U.S Bureau of Reclamation, Denver, CO., 1-210.
Heaton, T. (1990). Evidence for and implications of self healing pulses of
slip in earthquake rupture, Phys. Earth Planet Interiors 64, 1-20.
Herrmann, R., S. Park, and C. Wang (1981). The Denver earthquakes
of 1967-1968, Bull. Seism. Soc. Am. 71, 731-745.
Hough, S. E. and J. G. Anderson (1989). Attenuation in western
Nevada: Preliminary results from earthquake and explosion
sources, Geophys. Res. Lett. 16, 207-210.
Hough, S. E., L. Seeber, A. Lerner-Lam, J. C. Armbruster, and H.
Guo (1991). Empirical Green`s function analysis of Loma
Prieta aftershocks, Bull. Seism. Soc. Am. 81, 1737-1753.
Hough, S. E. and D. S. Dreger (1995a). Source parameters of the 23
April 1992 M 6.1 Joshua Tree, California, earthquake and its
aftershocks: Empirical Greens`s function analysis of GEOS
and Terrascope Data, Bull. Seism. Soc. Am. 85, 1576-1590.
Hough, S. E. (1995b). Can you believe observational constraints on
earthquake source scaling?, EOS Trans. Amer. Geophys. Union 76, F406.
Kachadorrian, R. R. F. Yerkes, and A. O. Waanemen (1967). Effects of the Truckee, California,
earthquake of September 12, 1966, U.S. Geol. Surv. Circular 537.
Keilis-Brook, V. I. (1957). Investigation of the mechanism of
earthquakes, Tr. Inst. Geofis. Akad. Nauk, SSSR, 40,
(Engl. transl., Sov. Res. Geophys. Ser., 4, 1960).
Lees, J. M. (1995). Xmap8: A Free Program for Three-Dimensional
GIS, Seism. Res. Lett. 66, 33-37.
Lindley, G. T. and R. Archuleta (1992). Earthquake source
parameters and the frequency dependence of attenuation at
Coalinga, Mammoth Lakes, and the Santa Cruz Mountains,
California, J. Geophys. Res. 97, 14137-14154.
Lovejoy, C. M. P. (1969). Mount Rose, northern Carson Range, Nevada; new light on the
late Cenozoic tectonic history of the Sierra Nevada from a classic locality, Geol. Soc. Am. Bull. 80, 1833-1842.
Madariaga, R. and A. Cochard (1996). Dynamic friction and the origin of the complexity of earthquake
sources, Proc. Natl. Acad. Sci. 93, 3819-3824.
Martinelli, Diane M. (1989). Geophysical investigations of the
northern Sierra Nevada-Basin and Range Boundary,
West-Central Nevada and East-Central California, Univ. Nevada
Reno, M.S. Thesis, Reno, NV, 1-127.
Mayeda, K. and W. R. Walter (1996). Moment, energy, stress drop, and source
spectra of western United States earthquakes from regional coda envelopes,
J. Geophys. Res. 101, 11195-11208.
McGarr, A. (1984). Scaling of ground motion parameters, state of stress, and focal
depth, Bull. Seism. Soc. Am. 89, 6969-6979.
Mori, J. and A. Frankel (1990). Source Parameters for small
events associated with the 1986 north Palm Springs, California
earthquakes determined using empirical Greens functions,
Bull. Seism. Soc. Am. 80, 278-295.
Mueller, C. S. (1985). Source pulse enhancement by deconvolution
of an empirical Green`s function, Geophys. Res. Lett. 12, 33-36.
Nelder, J. A. and R. Mead (1965). A simplex method for function
minimization, Computer Journal 7, 308-313.
Oldow, J. S. (1983). Tectonic implications of a late Mesozoic fold and thrust belt in
northwestern Nevada, Geology 11, 542-546.
Oldow J. S., H. G. Ave Lallemant, and W. J. Schmidt (1984). Kinematics
of plate convergence deduced from Mesozoic structures in the western Cordillera, Tectonics 3, 201-227.
Reasenberg, P. A. and D. Oppenheimer (1985). FPFIT, FPPLOT,
FPPAGE: Fortran computer programs for calculating and
displaying earthquake earthquake fault-plane solutions,
U.S. Geological Survey, Open File Rep. 85-739.
Rogers, A. M., S. C. Harmsen, Edward J. Corbett, K. Priestley,
and Diane dePolo (1991). The seismicity of Nevada and some
adjacent parts of the Great Basin, in Neotectonics of North
America, Editor, D. B. Slemmons, Geol. Soc. Am.,
Decade Map Vol. I, 153-183.
Ryall, A. S. (1977). Earthquake hazard in the Nevada region,
Bull. Seism. Soc. Am. 67, 517-532.
Sacks I. S., and P. A. Rydelek (1996). Earthquake "quanta" as an explanation
for observed magnitudes and stress drops, Bull. Seism. Soc. Am. 85, 808-813.
Sanders, C. O. and D. B. Slemmons (1979). Recent movements in the central
Sierra Nevada-Walker Lane region of California Part III, the Olinghouse fault zone, Tectonophysics 52, 585-597.
Shimazaki, K. (1986). Small and large earthquakes: the effects of the thickness of the
seismogenic layer and the free surface, in Earthquake Source Mechanics, M. Ewing (Editor),
Geophysical Monograph 37, Am. Geophys. Union 6, 209-216.
Sibson, R. H. (1974). Frictional constraints on thrust, wrench and normal faults, Nature 249, 542-544.
Sipkin, Stuart A. and Arthur L. Lerner-Lam (1992). Pulse-shape
distortion introduced by broadband deconvolution, Bull.
Seism. Soc. Am. 82, 238-258.
Slemmons, D. B. (1966). Cenozoic volcanism of the central Sierra Nevada, California, Calif. Div. Mines Geol.
Bull. 190, 199-208.
Slemmons, D. B., A. E. Jones, and J. I. Gimlett (1965). Catalog
of Nevada earthquakes 1852-1960, Bull. Seism. Soc. Am., 55, 537-583.
Speed, R. C. (1979). Collided Paleozoic microplate in the western United States, J. Geology 87, 279-292.
Stewart, J. H. (1988). Tectonics of the Walker Lane belt, western Great Basin: Mesozoic and
Cenozoic deformation in a zone of shear, in Metamorphism and Crustal Evolution of the Western
United States, W. G. Ernst (Editor), Ruby Vol. VII, Prentice-Hall, Englewood Cliffs, N. J., 683-713.
Thompson, G. A. (1964). Regional geology of the Steamboat Springs area, Washoe Valley, Nevada,
U.S. Geol. Surv. Profess. Pap. 458A.
Toppozada, T. R., C. R. Real, and D. L. Parke (1981). Preparation of isoseismal maps and summaries
of reported effects of pre-1900 California earthquakes, Calif. Div. Mines Geol. Open File Rept 81-11SAC.
Tsai, Yi-Ben and K. Aki (1970). Source mechanism of the Truckee,
California, earthquake of September 12, 1966, Bull. Seism. Soc. Am. 60, 1199-1208.
Vidale, J. E. (1989). Influence of focal mechanism on peak ground accelerations of strong motions of the
Whittier Narrows, California, earthquake and an aftershock, J. Geophys. Res. 94, 9607-9613.
Wallace, T. C. (1987). Grouping and migration of surface faulting
and variations in slip rates on faults in the Great Basin
Province, Bull. Seism. Soc. Am. 77, 868-876.
Zeng, Y. and J. G. Anderson (1995). A method of direct computation of the
differential seismogram with respect to the velocity changes in a layered elastic
solid, Bulletin Seismological Soceity of America 85, 300-307.