Download (2008) Stress-forecasting not predicting earthquakes: A paradigm shift

Stress-forecasting (not predicting) earthquakes: A paradigm shift?
Stuart Crampin* School of GeoSciences, University of Edinburgh, Edinburgh EH9 3JW, UK
Yuan Gao Institute of Earthquake Science, China Earthquake Administration, 100036 Beijing, China
Sheila Peacock AWE Blacknest, Brimpton, Reading RG7 4RS, UK
ABSTRACT
After 120 years of unsuccessful endeavor, a paradigm shift is required before earthquakes
can be predicted. The most sensitive diagnostic of low-level changes of stress in in situ rock,
variations in microcrack geometry, can be monitored by analyzing shear-wave splitting.
The suggested paradigm shift is that, instead of investigating the source zone, we monitor
stress accumulation before earthquakes at, possibly, substantial distances from the source.
Characteristic temporal variations of shear-wave time delays have been observed in retrospect before 14 earthquakes worldwide. On one occasion, when changes were recognized
early enough, the time, magnitude, and fault break of an M = 5 earthquake in southwest Iceland were successfully stress-forecast in a narrow time-magnitude window. Such stress accumulation can be theoretically modeled and is believed to be at least partially understood.
When sufficient shear-wave source earthquakes are available, increasing time delays also
show an abrupt decrease shortly before the impending earthquake occurs. This is not fully
understood but is thought to be caused by stress relaxation as microcracks coalesce onto the
eventual fault break. The new result confirming these ideas, and justifying the paradigm
shift, is that logarithms of the durations of both increases and decreases in time delays are
found to be proportional (self-similar) to the magnitudes of impending earthquakes.
σh
σH
O
σV
Keywords: stress accumulation, paradigm shift, shear-wave splitting, stress-forecasting earthquakes.
INTRODUCTION
In a comprehensive review of earthquake prediction, Geller (1997) concludes that a paradigm
shift is required before times, magnitudes, and
epicenters of large earthquakes can be predicted.
Previous attempts at earthquake prediction investigated the earthquake source (and associated precursors) with singular lack of success. We suggest
that the paradigm shift is to neglect the earthquake
source zone, which is locally heterogeneous and
probably overly complicated by initial conditions
(Geller, 1997). Instead, with a new understanding
of fluid-rock deformation (Crampin, 1999, 2006;
Crampin and Peacock, 2005), seismic shear-wave
splitting is used to monitor the buildup of strain
energy in the rock mass remote from the source.
We refer to increasing strain energy as “stress
accumulation.” Since rock is weak to shear stress,
stress must accumulate over enormous volumes
of rock in order to amass sufficient energy for
release by large earthquakes. This allows stress
accumulation to be recognized at substantial distances from the eventual source zone. We report
new results showing that the time, magnitude,
and in some cases location of earthquakes can be
stress-forecast by monitoring swarm activity with
shear-wave splitting.
MONITORING IN SITU STRESS WITH
SHEAR-WAVE SPLITTING
Azimuthally aligned shear-wave splitting
(seismic birefringence) is widely observed
throughout Earth’s crust in almost all geologi*E-mail: [email protected]
cal and tectonic regimes. The polarizations of
the faster split shear waves are typically parallel
to the direction of maximum horizontal stress
(Crampin, 1981, 1994, 2006; Alford, 1986).
Shear-wave splitting is diagnostic of some
form of seismic anisotropy, and the only anisotropic symmetry system that has such parallel
polarizations for propagation within ~30° of
the vertical is hexagonal symmetry (transverse
isotropy) with a horizontal axis of symmetry
(Crampin, 1981). Moreover, the only geological configurations common to almost all in situ
igneous, metamorphic, and sedimentary rocks
that have such symmetry are distributions of
fluid-saturated microcracks that are aligned,
like hydraulic fractures in the oil industry
(Hubbert and Rubey, 1959), perpendicular to
the direction of minimum horizontal stress
(Crampin, 1981, 1994, 2006).
Figure 1 gives a schematic illustration of
shear-wave splitting in stress-aligned microcracked rock. Fluid-saturated microcracks are
the most compliant elements of in situ rocks
and are the only geological phenomenon that
can display the occasionally rapid temporal
changes in time delays listed in Table 1. No
other source of anisotropy has such sensitivity, and characteristic patterns of temporal
changes in time delays between split shear
waves are observed before earthquakes (15
examples listed in Table 1A, nine in Table 1B),
before volcanic eruptions (three examples in
Table 1C), during distant low-level seismicity
at 70 km distance (with energy equivalent to
an M ≈ 3.5 earthquake) (Table 1D), and during
Figure 1. Schematic illustration of shearwave splitting through parallel vertical fluidsaturated microcracks aligned normal to
the direction of minimum horizontal stress,
σh , below the depth at which the increasing
vertical stress, σV , equals σH . σH is the maximum horizontal stress.
both high- and low-pressure CO2 injections in
a carbonate reservoir (Table 1E).
Generally, changes in time delays before
earthquakes have been observed only in retrospect. However, in October 1999, it was recognized that time delays at two seismic stations in
southwest Iceland were increasing at rates similar to those before an M = 5.1 earthquake some
four months earlier. Consequently, we issued
a preliminary forecast and, on 10 November
1998, made the final (successful) stress forecast
(Crampin et al., 1999).
Texts of e-mail messages between the University of Edinburgh (EU) and the Iceland Meteorological Office (IMO) (Crampin et al., 1999):
10 November 1998—EU to IMO: “…the last
plot…(of shear-wave time delays)…is already
very close to 10 ms/km. This means that an
event could occur any time between now
(M ≥ 5) and end of February (M ≥ 6).”
13 November 1998—IMO to EU: “… there
was a magnitude 5 earthquake just near BJA,
preliminary epicenter 2 km west of BJA this
morning 10 38 GMT.”
© 2008 The Geological Society of America. For permission to copy, contact Copyright Permissions, GSA, or [email protected].
GEOLOGY,
May
2008
Geology,
May
2008;
v. 36; no. 5; p. 427–430; doi: 10.1130/G24643A.1; 3 figures; 1 table; Data Repository item 2008102.
427
In response to an initial less-specific stress
forecast a few days earlier, Ragnar Stefánsson,
IMO, correctly inferred that the impending
earthquake would be associated with the fault of
an earlier M ≥ 5.1 earthquake where low-level
seismicity was still continuing. This allowed
time, magnitude, and fault plane to be accurately stress-forecast (Crampin et al., 1999).
The earthquake was 2 km from station BJA, but
similar effects have been observed with hindsight up to 70 km from small-scale seismicity,
and 245 km from volcanic eruptions (Table 1).
Note the following:
1. Time delays are normalized to ms/km
(Crampin, 1999).
2. A time delay of 10 ms/km corresponds to
levels of fracture criticality estimated from previous earthquakes (Volti and Crampin, 2003).
Fracture criticality is the level of cracking when
cracks lose shear strength and earthquakes occur
(Crampin and Peacock, 2005).
3. This procedure is called “stress forecasting” to distinguish it from source-zone–based
earthquake prediction.
4. There are few examples in Table 1 because
combinations of persistent swarms of small
earthquakes, seismic three-component recorders
immediately above the swarms, and suitable
large earthquakes nearby, are extremely scarce
(Crampin, 1993).
MODELING FLUID-ROCK
DEFORMATION
The deformation of fluid-saturated microcracked rocks for small changes of stress can be
modeled by anisotropic poro-elasticity (APE)
(Zatsepin and Crampin, 1997; Crampin and
Zatsepin, 1997; Crampin, 2006). Fluid-saturated
microcracks are the most compliant elements of
in situ rocks, and APE models the mechanism
for low-level deformation before fracturing
occurs. Deformation is by fluid movement along
pressure gradients between neighboring microcracks at different orientations to the stress field.
These microcracks are grain-boundary cracks
in crystalline rocks and aligned pores and pore
throats in higher-permeability sedimentary
rocks. APE approximately matches a large
range of phenomena involving cracks, stress,
and shear-wave splitting (Crampin and Chastin,
2003; Crampin and Peacock, 2005). Note that
the match is only approximate because of the
difficulty of getting accurate in situ measurements of subsurface microcracks.
APE shows that the most diagnostic effect of
small increases (or decreases) of stress on microcrack geometry is to increase (or decrease) the
aspect ratio of vertical cracks striking parallel
to the direction of maximum horizontal stress.
These increase (or decrease) the average time
delays in Band-1 directions to the free surface
(Crampin, 1999). Band-1 is the double-leafed
428
TABLE 1. OBSERVATIONS OF CHANGES IN SHEAR-WAVE SPLITTING
No.
Earthquake (or volcano)
location
Approximate
distance
(km)
Magnitude
Approximate
duration
(days)
Reference*
(A) Observations of increasing shear-wave time delays before earthquakes
1
2
3
4
5
6
7
8§
9§
10§
11§
12
13
14†
15
Swarm at BRE, northern Iceland
Swarm at BRE, northern Iceland
Southwest Iceland
Dongfang, Hainan, China
Enola Swarm, Arkansas
Southwest Iceland
Parkfield, California
Southwest Iceland
Southwest Iceland
Grímsey Lineament, Iceland
Southwest Iceland
(successful stress forecast)
Shidian, Yunnan, China
Northern Palm Springs, California
Southwest Iceland
Chi-Chi earthquake, Taiwan
7
7
10
9
3
14
14
10, 43
10, 43
50, 92, 96
2, 36
M† 1.7
M 2.5
M 3.4
ML 3.6
ML 3.8
M 3.8
ML 4.0
M 4.4
M 4.7
M 4.9
M 5.0#
≥0.055
≥0.210
47
21
≥4.5
40
≥220
83, 77
123, 106
147, 163, –
127, 121
1
1
2
3
4
2
5
2
2
6
1, 2, 7
35
33
3, 46
55
Ms 5.9
Ms 6.0
M 5.6/Ms 6.6
Ms 7.7
400
1100
75, 151
598
1
1, 8
2, 9
10
(B) Observations of decreasing shear-wave time delays immediately before earthquakes in (A)
Swarm at BRE, northern Iceland
7
M 1.7
0.0306
1
Swarm at BRE, northern Iceland
7
M 2.5
0.0465
1
Enola Swarm, Arkansas
3
ML 3.8
0.123
4
Grímsey Lineament, Iceland
50
M 4.9
24
6
4.4
1, 2, 7
Southwest Iceland
2
M 5.0#
(successful stress forecast)
6
Shidian, Yunnan, China
35
Ms 5.3
38
1
7 (13)
North Palm Springs, California
33
Ms 6.0
69
1, 8
Southwest Iceland
3, 46
Ms 6.6/M 5.6
38, 21
2, 9
8§ (14)
9 (15)
Chi-Chi earthquake, Taiwan
55
Ms 7.7
131
10
1 (1)
2 (2)
3 (5)
4 (10)
5 (11)
(C) Observations of changes before volcanic eruptions
1§
2§
3§
Gjàlp, Vatnajökull, Iceland
230 S, 240 SW Large fissure
(increasing time delays)
245 WSW
eruption
Mount Etna, Sicily (increasing and
1, 5
Minor
eruption
decreasing time delays, 90° flips‡)
Mount Ruapehu, New Zealand
2–15
Minor
(90° flips)
eruption
120
2
66**
11
–
12
(D) Observations during seismic swarm at prototype stress-monitoring site, Húsavík, Iceland
1
Grímsey Lineament, Iceland
M ≡ 3.5
~70
–
13
(E) Observations during high- and low-pressure CO2 injections in carbonate reservoir
1
Vacuum Field, New Mexico
0 (vertical)
–
–
14
*1—Gao and Crampin (2004); 2—Volti and Crampin (2003); 3—Gao et al. (1998); 4—Booth et al. (1990);
5—Liu et al. (1997); 6—Gao and Crampin (2006); 7—Crampin et al. (1999); 8—Peacock et al. (1988);
9—Wu et al. (2006); 10—Crampin and Gao (2005); 11—Bianco et al. (2006); 12—Miller and Savage (2001);
13—Crampin et al. (2003); 14—Angerer et al. (2002).
†
Iceland seismic catalog magnitude M ≈ mb.
§
Observed at more than one seismic station.
#
Older magnitude value compatible with other listed magnitudes, now M = 4.9 in current catalog.
##
As interpreted by this study.
‡
90° flips, where faster and slower split shear waves exchange polarizations due to microcrack
realignments for critically high pore-fluid pressures (Angerer et al., 2002; Crampin, 2006).
solid angle of ray-path directions with an incidence 15°–45° either side of the plane of vertical cracks (geometry in Fig. DR1 of the GSA
Data Repository1). Band-2 directions, 15° to
either side of the plane of the cracks, are sensitive principally to crack density, and crack
density does not have a simple relationship with
changes of stress (Crampin, 1999).
Recently, it has been noticed that whenever
there are enough shear-wave ray paths in Band-1
1
GSA Data Repository item 2008102, geometry for Band-1 and Band-2 directions, and further
examples of characteristic changes in time delays
before earthquakes, is available online at www.
geosociety.org/pubs/ft2008.htm, or on request from
[email protected] or Documents Secretary,
GSA, P.O. Box 9140, Boulder, CO 80301, USA.
directions, the increase of time delays abruptly
starts to decrease shortly before the time of
the impending earthquake (Gao and Crampin,
2004). Figure 2 shows a typical example (see
also Fig. DR2). Table 1B lists nine earthquakes
where such decreases have been observed in the
15 earthquakes showing increases.
The new result is that durations of both
increases and decreases are found to scale logarithmically with magnitudes of the impending
events for three to four units of earthquake magnitude in Figures 3A and 3B. This self-similarity
directly links variations in time delays with
earthquake magnitudes, analogous to the wellknown Gutenberg-Richter relationship.
Since APE modeling shows that increasing
stress causes increases in average Band-1 time
GEOLOGY, May 2008
Southwest Iceland, Station BJA
A
B
M=5
M=5
8.0
20
10
0
Jul
Sep
Nov
Months in 1998
0
5
Days
10
Figure 2. A: Increase in normalized time
delays (ms/km) in Band-1 of the shear-wave
window (Booth and Crampin, 1985) for about
four months before an M = 5 earthquake in
southwest Iceland (Crampin et al., 1999).
B: Enlarged time scale for dotted box in A,
with dashed line showing decrease in time
delays starting about four days before the
earthquake (Gao and Crampin, 2004). Error
bars are location errors.
Magnitude (various scales)
Normalized
time delays
(ms/km)
30
Duration of increasing time delays
A
Icelandic Bulletin magnitude
Other magnitude
15
7.0
6.0
14
13
12
11
5.0
10
9
8
6
4.0
4
3
3.0
M = (2.16 ± 0.37) log10(days) + 0.37 ± 0.82
Correlation coefficient = 0.89
2.0
10.0
100.0
1000.0
Duration (days)
GEOLOGY, May 2008
Duration of decreasing time delays
B
8.0
7.0
Magnitude (various scales)
delays, the observed increase is interpreted as
monitoring the effects of stress accumulation
on microcrack geometry at large distances from
impending earthquakes. In any particular region,
the overall stress can be expected to increase at a
comparatively uniform rate over the recurrence
cycles of large earthquakes (decades to thousands of years), driven by the movement of tectonic plates, which typically converge or diverge
at some 2–10 cm/yr. If stress accumulates over
a small volume in a heterogeneous Earth, the
rate of increase will be rapid, and the level of
fracture criticality at which earthquakes occur
will be reached within the small volume after a
short time, so that the total energy accumulated
and the impending stress-releasing earthquake
will be small. If stress accumulates over a larger
volume, the rate of accumulation will be slower
and the duration longer, but the impending
earthquake will be larger, as in Table 1A and
Figure 3A. The slope of the straight line in Figure 3A depends on the rate of increase of stress.
This rate appears to be similar for most data
in Figure 3A, which, except for earthquake 15
(Ms 7.7), plots smaller earthquakes where stress
accumulation is presumed to be wholly within
the crust. Stress accumulation before larger
earthquakes, such as earthquake 15, involves
the whole crust, and the increase will be over
two dimensions, rather than the accumulation
over three dimensions for smaller earthquakes.
Consequently, the relationship may not be linear
for larger earthquakes. Note that we have not
attempted to unify the various magnitude scales,
which are typically unreliable over such a range
of dissimilar data, and this no doubt contributes
to the scatter in Figure 3A.
The decrease shown in Figure 3B is not
wholly understood but is interpreted as monitoring stress relaxation as microcracks coalesce
onto the eventual fault plane (similar phenomena occur in laboratory stress cells; see Figs.
Icelandic Bulletin magnitude
Other magnitude
9
6.0
8
6
5
5.0
7
4
4.0
3
Figure 3. Least-squares
lines through impending
earthquake magnitudes
plotted against logarithms of the duration in
days of increasing (A)
and decreasing (B) time
delays. Numbered data
points refer to Tables
1A and 1B, respectively.
Where earthquakes have
observations at more
than one seismic station (earthquakes 8, 9,
10, 11, and 14, in Table
1A, and earthquake 8 in
Table 1B), duration values from stations closest
to the impending earthquakes are plotted. Note
that onsets of increases
of time delays are sometimes uncertain because
of overlays with other
earthquakes in the seismograms (earthquakes
1, 2, 5, and 7 in Table 1A),
and these data points are
also not plotted.
3.0
2
2.0
1
M = (1.17 ± 0.17) log10(days) + 4.01 ± 0.20
Correlation coefficient = 0.93
1.0
0.0
0.01
0.1
1.0
10.0
100.0
Duration (days)
DR2F and DR2G). As APE suggests, “crack
coalescence” is probably a function of crack
geometry rather than the properties of the geology or tectonics of the rock matrix.
Figures 3A and 3B suggest that, if patterns
of increasing and decreasing time delays can
be recognized before impending earthquakes,
the times and magnitudes of larger earthquakes
can be stress-forecast, as the M = 5 earthquake
in southwest Iceland was successfully stressforecast three days before it occurred (Crampin
et al., 1999). Shear-wave splitting does not
appear to contain direct information about the
location of the impending event, but knowing a
larger earthquake is approaching allows other
information to be interpreted realistically, as
occurred with the successfully forecast earthquake (Crampin et al., 1999). The data in
Table 1 show that these changes in time delays
can be recognized at considerable distances
from the impending epicenters.
THE PARADIGM SHIFT
The suggested paradigm shift is that instead of
attempting to predict earthquakes by investigating source zones and source-zone precursors, the
time and magnitude and in some circumstances
location of impending larger earthquakes can
be stress-forecast by monitoring the approach
to fracture criticality anywhere within an extensive volume surrounding the eventual epicenter.
This bypasses the essential deterministic unpredictability of the earthquake source. Rock is
so weak to shear stress that the strain energy
released by large earthquakes necessarily accumulates over very large volumes of rock, as indicated by distances at which changes in shearwave splitting are observed (Table 1). There is
an enormous variety and quantity of evidence
supporting this new understanding of fluidrock deformation (Crampin and Chastin, 2003;
Crampin and Peacock, 2005) with no contrary
observations. Consequently, we suggest that
429
large earthquakes can be stress-forecast by using
shear-wave splitting to monitor stress-induced
changes to microcrack geometry remote from
impending epicenters.
Unfortunately, variations in shear-wave splitting above small earthquakes cannot be used to
routinely stress-forecast earthquakes, because
of the scarcity (Crampin, 1993) and irregularity
(Wu et al., 2006) of suitable swarms of small
earthquakes. Reliable routine stress forecasting requires the controlled-source cross-hole
seismics of a stress-monitoring site (SMS)
(Crampin, 2001), where a borehole source
transmits shear waves in Band-1 directions to
recorders in adjacent boreholes at depths below
the level of the near-surface stress release and
weathering anomalies. The prototype SMS at
Húsavík, northern Iceland, used existing boreholes drilled for geothermal purposes adjacent
to a transform fault of the Mid-Atlantic Ridge.
The SMS was not in optimal source-stress
geometry, but showed spectacular sensitivity
to very low-level seismic swarm activity (seismic energy equivalent to one M ≈ 3.5 earthquake at ~70 km distance; Crampin et al., 2003)
(Table 1D), confirming the science, technology,
and sensitivity of SMSs for monitoring stress
accumulation and stress-forecasting impending
large earthquakes. SMSs would allow variations
in tectonic strain rates between different regions
to be cataloged routinely, so that the accuracy of
forecast magnitudes would be improved.
ACKNOWLEDGMENTS
Gao was partially supported by National Natural
Science Foundation of China project 40774022. The
authors thank Francesca Bianco, David Booth, and
an anonymous reviewer whose comments helped to
improve the manuscript.
REFERENCES CITED
Alford, R.M., 1986, Shear data in the presence of azimuthal anisotropy, Dilley, Texas: 56th Annual
International Meeting, Society of Exploration Geophysicists, Expanded Abstracts,
p. 476–479, doi: 10.1190/1.1893036.
Angerer, E., Crampin, S., Li, X.-Y., and Davis, T.L.,
2002, Processing, modelling, and predicting time-lapse effects of overpressured fluidinjection in a fractured reservoir: Geophysical
Journal International, v. 149, p. 267–280, doi:
10.1046/j.1365-246X.2002.01607.x.
Bianco, F., Scarfi, L., Del Pezzo, E., and Patanè,
D., 2006, Shear wave splitting changes
associated with the 2001 volcanic eruption
on Mt. Etna: Geophysical Journal International, v. 167, p. 959–967, doi: 10.1111/
j.1365-246X.2006.03152.x.
Booth, D.C., and Crampin, S., 1985, Shear-wave
polarizations on a curved wavefront at an iso-
430
tropic free-surface: Geophysical Journal of the
Royal Astronomical Society, v. 83, p. 31–45.
Booth, D.C., Crampin, S., Lovell, J.H., and Chiu,
J.-M., 1990, Temporal changes in shear wave
splitting during an earthquake swarm in Arkansas: Journal of Geophysical Research, v. 95,
p. 11,151–11,164.
Crampin, S., 1981, A review of wave motion
in anisotropic and cracked elastic-media:
Wave Motion, v. 3, p. 343–391, doi: 10.1016/
0165-2125(81)90026-3.
Crampin, S., 1993, Do you know of an isolated
swarm of small earthquakes?: Eos (Transactions, American Geophysical Union), v. 74,
p. 451 and 460.
Crampin, S., 1994, The fracture criticality of crustal
rocks: Geophysical Journal International,
v. 118, p. 428–438, doi: 10.1111/j.1365-246X.
1994.tb03974.x.
Crampin, S., 1999, Calculable fluid-rock interactions:
Geological Society [London] Journal, v. 156,
p. 501–514, doi: 10.1144/gsjgs.156.3.0501.
Crampin, S., 2001, Developing stress-monitoring sites
using cross-hole seismology to stress-forecast
the times and magnitudes of future earthquakes: Tectonophysics, v. 338, p. 233–245,
doi: 10.1016/S0040-1951(01)00079-8.
Crampin, S., 2006, The new geophysics: A new
understanding of fluid-rock deformation, in
Van Cotthem, A., et al., eds., Eurock 2006:
Multiphysics coupling and long-term behaviour
in rock mechanics: London, Taylor and Francis,
p. 539–544.
Crampin, S., and Chastin, S., 2003, A review of
shear-wave splitting in the crack-critical
crust: Geophysical Journal International,
v. 155, p. 221–240, doi: 10.1046/j.1365-246X.
2003.02037.x.
Crampin, S., and Gao, Y., 2005, Comment on “Systematic analysis of shear-wave splitting in the
aftershock zone of the 1999 Chi-Chi, Taiwan,
earthquake: Shallow crustal anisotropy and
lack of precursory changes,” by Liu, Teng, and
Ben-Zion: Temporal variations confirmed: Bulletin of the Seismological Society of America,
v. 95, p. 354–360, doi: 10.1785/0120040092.
Crampin, S., and Peacock, S., 2005, A review of
shear-wave splitting in the compliant crackcritical anisotropic Earth: Wave Motion, v. 41,
p. 59–77, doi: 10.1016/j.wavemoti.2004.05.006.
Crampin, S., and Zatsepin, S.V., 1997, Modelling the
compliance of crustal rock, II: Response to temporal changes before earthquakes: Geophysical
Journal International, v. 129, p. 495–506, doi:
10.1111/j.1365-246X.1997.tb04489.x.
Crampin, S., Volti, T., and Stefánsson, R., 1999, A
successfully stress-forecast earthquake: Geophysical Journal International, v. 138, p. F1–F5,
doi: 10.1046/j.1365-246x.1999.00891.x.
Crampin, S., Chastin, S., and Gao, Y., 2003, Shearwave splitting in a critical crust, III: Preliminary report of multi-variable measurements
in active tectonics: Journal of Applied Geophysics, v. 54, p. 265–277, doi: 10.1016/
j.jappgeo.2003.01.001.
Gao, Y., and Crampin, S., 2004, Observations of stress
relaxation before earthquakes: Geophysical
Journal International, v. 157, p. 578–582, doi:
10.1111/j.1365-246X.2004.02207.x.
Gao, Y., and Crampin, S., 2006, A further stressforecast earthquake (with hindsight), where
migration of source earthquakes causes
anomalies in shear-wave polarisations: Tectonophysics, v. 426, p. 253–262, doi: 10.1016/
j.tecto.2006.07.013.
Gao, Y., Wang, P., Zheng, S., Wang, M., Chen,
Y.-T., and Zhou, H., 1998, Temporal changes
in shear-wave splitting at an isolated swarm
of small earthquakes in 1992 near Dongfang,
Hainan Island, southern China: Geophysical
Journal International, v. 135, p. 102–112, doi:
10.1046/j.1365-246X.1998.00606.x.
Geller, R.J., 1997, Earthquake prediction: A critical review: Geophysical Journal International,
v. 131, p. 425–450, doi: 10.1111/j.1365-246X.
1997.tb06588.x.
Hubbert, M.K., and Rubey, W.W., 1959, Role of
fluid pressure in mechanics of overthrust faulting, 1: Mechanics of fluid-filled porous solids
and its application to overthrust faulting: Geological Society of America Bulletin, v. 70,
p. 115–166, doi: 10.1130/0016-7606(1959)70
[115:ROFPIM]2.0.CO;2.
Liu, Y., Crampin, S., and Main, I., 1997, Shearwave anisotropy: Spatial and temporal variations in time delays at Parkfield, central
California: Geophysical Journal International,
v. 130, p. 771–785, doi: 10.1111/j.1365-246X.
1997.tb01872.x.
Miller, V., and Savage, M., 2001, Changes in seismic anisotropy after volcanic eruptions: Evidence from Mount Ruapehu: Science, v. 293,
p. 2231–2235, doi: 10.1126/science.1063463.
Peacock, S., Crampin, S., Booth, D.C., and Fletcher,
J.B., 1988, Shear-wave splitting in the Anza
seismic gap, southern California: Temporal
variations as possible precursors: Journal of
Geophysical Research, v. 93, p. 3339–3356.
Volti, T., and Crampin, S., 2003, A four-year study
of shear-wave splitting in Iceland, 2: Temporal
changes before earthquakes and volcanic eruptions, in Nieuwland, D.A., ed., New insights
into structural interpretation and modelling:
Geological Society [London] Special Publication 212, p. 135–149.
Wu, J., Crampin, S., Gao, Y., Hao, P., Volti, T., and
Chen, Y.-T., 2006, Smaller source earthquakes
and improved measuring techniques allow the
largest earthquakes in Iceland to be stressforecast (with hindsight): Geophysical Journal International, v. 166, p. 1293–1298, doi:
10.1111/j.1365-246X.2006.03054.x.
Zatsepin, S.V., and Crampin, S., 1997, Modelling
the compliance of crustal rock, I: Response
of shear-wave splitting to differential stress:
Geophysical Journal International, v. 129,
p. 477–494, doi: 10.1111/j.1365-246X.1997.
tb04488.x.
Manuscript received 26 November 2007
Revised manuscript received 30 January 2008
Manuscript accepted 6 February 2008
Printed in USA
GEOLOGY, May 2008