Tecronophyms. Elsevier
144 (1987) 207-214
Science Publishers
207
B.V., Amsterdam
- Printed
in The Netherlands
Coseismic strain steps observed by three-component borehole strainmeters SEIICHI SHIMADA, SHOJI SAKATA and SHIN’ICHI NOGUCHI National Research Center/or
Disaster Preoention,
Tennodm 3. Chome, Sakura-mura,
Ibaraki-ken. (Received
January
Niihuri-gun,
Tsukubu Science City.
305 (Japan)
2, 1986: revised version accepted
May 8, 1986)
Abstract Shimada,
S., Sakata.
strainmeters. Thirteen Japan.
S. and
coseismic
strain
motions
or long-period
amplitudes
theoretical seismic
demonstrate
the
high
borehole
to determine
Coseismic
M&hanics
were‘installed
by 3-component
60 m apart,
waves.
The calculated
reliability
strainmeters
of the
strain
including
aseismic
event.
by long-period
Even mechanism surface
strain
in the Kanto
area,
steps were in good agreement
at Yasato
with
determined
steps agree with observations strainmeter
by P-wave initial
within
a factor
of 2 for
can detect steps over 0.3 nanostrain.
strainmeter.
deformation
slight
de-
When
the network
steps are observed
at several
These
of stations
stations,
of
it will be
of the earthquake.
difference
step amplitudes investigated
can only evaluate the “total including aseismic deformation
Publishers
strainmeters
borehole
144: 207-214.
existed
in the direction of printo calculated
cipal axes, and the ratios of observed
or body waves
0 1987 Elsevier Science
three-component
of less than 1000 km agree fairly well with those calculated based on the fault model. However, a
Comparisons of observed coseismic strain steps to the deformation calculated from focal mechanisms have been done for several earthquakes (for instance, Kasahara, 1974; Okada, 1980; McGarr et al., 1982; Furuya, 1984). Kasahara (1974) studied strain steps from the Off Nemuro 0040-1951/87/$03.50
by
Tectonophysics,
Peninsula earthquake of 1973 (M = 7.4) (Japan Meteorological Agency (JMA)) and concluded that the strain steps observed at the epicentral distance
cannot detect aseismic motion, if it occurs, accompanying seismic faulting. The measurement of static deformation fault mechanism” of an earthquake.
observed
from local mechanisms
and coseismic
The focal mechanism of an earthquake is determined by two methods: P-wave initial motions and long-period seismic waves. When the source process is complicated, the mechanism determined by P-wave initial motions represents only the motermination
borehole
three-component
is constructed
fault mechanisms,
steps Faulting.
and the observed
The three-component
Introduction
tion of the triggering
strain
of Earthquake
strain steps were calculated
and within 10 o for directions.
three-component possible
S.. 1987.
(Editor),
steps were observed
Two sets of strainmeters
each other. Corresponding
results
Noguchi,
In: R.L. Wesson
strain
and
from 2 to 5. Okada tilt steps
(1980)
from six earth-
quakes in and around Japan and found no empirical relationships between the observed and the calculated steps, and that the polarities of the steps were nearly random. He concluded that there remain unknown miscellaneous factors involved in strain or tilt measurements. McGarr et al. (1982) studied the coseismic strain changes ranging from 5 x lo-‘” to values exceeding 5 X lo-’ which were observed with three Sacks-Evertson borehole dilatometers for hundreds of mine tremors in the magnitude
B.V.
ranged
range
-1-3.7
and at hypocentral
dis-
208
tances
of 50 m to about
that the amplitudes strain
2 km. They
and polarities
steps were generally
with theoretical dislocation
Furuya
(1984)
with borehole
belonging
to the Japan
that the polarities
those calculated were in proportion
studied
volume
the absolute
strain
strainmeters
Meteorological
Agency differential
did not agree with
from the source
at some stations,
agreement
based on point-source
and
found
of the coseismic
in excellent
expectations
theory.
steps observed
concluded
model,
values
to the theoretically
transformer
but that
of the steps predicted
values. The biggest problem in coseismic strain step measurements is the coupling between the crust and instruments. There are many troublesome factors involved with the observations in vaults (Shichi
and Okada,
1979). Weathered
a vault loosens the coupling tensometers used for strain
section of detection
port
rock around
with the crust. Exmeasurements in a
vault have a mechanically complex detector and supporters, and this complexity also affects the coupling with the ground. In contrast, borehole measurements have some advantages concerning
Fig. 1. Schematic illustration of the three-component strainmeter.
the coupling with the crust. It is easier to find a site and to reach the depth of fresh and massive rocks. The three-component borehole strainmeters
method used to detect section area change is essentially the same as that in the Sacks-Evertson volumetric strainmeter (Sacks et al., 1971). The observed section area increases AS,, AS, and AS, are obtained from the horizontal prin-
have a very simple structure and are well coupled to the surrounding rocks. In this paper we present the strain steps observed with the three-compo-
transmitted
to the differential
cipal strains
extension B-see al., 1982a):
the first time.
ei + ez = (AS, + AS, + A&)/3/f
Observation Instrument
transformer.
c,, e2 and the direction
nent borehole strainmeter at Yasato, central Japan, where this kind of strainmeter was installed for
following
borehole
equations
The
of maximum (Sakata
et
C, - C2 -+[(A&-AS?)z+(AQA&)* + (AS, - AS,)*] li2/3B tan 2B = a(
AS, - AS,)/(2AS,
- AS, - AS,)
The concept of the instrument was introduced by Sakata et al. (1982a). Figure 1 shows a schematic illustration of the three-component strainmeter. The detection part of the strainmeter is equally divided into three independent sections by
(1) In this formulation, the three sections are numbered counterclockwise in the topview cross section, and B is the angle, defined counterclockwise, from the axis of c1 to the boundary axis between
rigid partition walls. Each section is filled with silicon oil and is connected to the bellows with a pipe. When the crustal strain field around the strainmeter changes, the volumetric change induced by the section area change is detected as the displacement of a bellows, and the displacement is
section 1 and section 3. A and B are the constants determined by elastic constants and the thicknesses of the detection cylinder metal, expansive mortar and surrounding rock. Detailed descriptions of the instrument (1987).
are given by Sakata
et al.
209
Kanto
Asian
observed,
District
the
original
shows an example
of the strain
earthquake
occurred
area
Plate
including which
was
observation steps.
For
Figure
about
35 km
point.
The record
the thirteen
3
step caused by the
in the central
Kanto
27, 1983 (M = 6.0). The
on February
center
case.
to the
south
shows
coseismic
epi-
of the
very sharp
steps
that
we
observed,
there seems to be no case in which the
coseismic
step
was followed
by creeping
move-
ments. Table Fig. 2. Location Research
Center
of installed
strainmeters
for Disaster
and
Prevention
the National
1 lists the observed
steps at Yasato,
(NRCDP).
where extension
tive. The absolute strains In 1982, two sets of strainmeters 60 m apart at Yasato
station
(Fig.
of the boreholes
2). The depths
principal
0.1 and 100 nanostrain. changes
The
are in good agreement
with each other, mostly within a factor of 2 for the amplitude and within 10” for the direction. Table 2 shows epicentral coordinates and magnitudes of the earthquakes in Table 1. The epicentral coordinates are determined by the JMA and the National Research Center for Disaster (NRCDP), except for the Off Akita (No. 5 in Table 2). For this earthquake two-fault models determined Mori (1983) and coordinates
Coseismic strain steps were originally detected from the Off Ibaraki earthquake (Sakata et al.,
Observed
are between
Tsukuba
Strain steps
TABLE
is taken as posi-
are 160 m
near Mount
(No. 1) and 165 m (No. 2) respectively. The surrounding rock is massive granite. The data were telemetered to the National Research Center for Disaster Prevention (Shimada et al., 1983).
1982b). So far, 13 coseismic
of the strain
values of the observed
two sets of strain
were installed
value
Prevention earthquake we use the
by Shimazaki and are given at the
central points of the deeper horizontal edges of the faults in Table 2. Shimazaki and Mori (1983)
strain steps have been
1 values of coseismic
No.
Origin
strain
steps at Yasato,
time (JST)
Observed
with origin principal
times of the earthquakes
strain (nanostrain) No. 2
No. 1 max.
direct.
min.
max. 68
- 35
93”
-0.5
0.7
60”
-0.5
1.2
3.7
82’=
- 3.2
2.6
-0.2 -0.7
0.3
0.1 -0.3 -0.5 - 0.6
0.2
loo
- 0.4
0.9
64O
-0.4
- 1.0 - 1.3
1.4
75O
- 1.0
1.0
45”
-1.4
84” 99O
-38 - 0.5
Aug. 14 1982
6:14
0.6
60”
-0.3
4
Feb. 27 1983
21:14
6.1
82”
5 6
May 26 1983 Jul. 2 1983
11:59 7:03
3.3 0.6
2O 23”
7
Aug. 8 1983
12:48
0.1
-15”
8
Oct. 28 1983
lo:50
0.3
-30°
9
Dec. 30 1983
11:30
0.3
-20”
10
Jan. 1 1984
18:03
0.3
I0
11
Jan. 17 1984
20:13
1.1
61°
12
Jan. 18 1984
0:31
1.5
7o”
13
Sep. 19 1984
2:02
1.1
39O
3
* “Max.” maximum
and
“min.”
principal
indicate strain
values
as measured
of the maximum clockwise
and
minimum
from the north.
min.
85”
84 1.4
Jul. 23 1982 Jul. 24 1982
direct.
1.3
23:24 2154
1 2
*
principal
-5O 20”
1.5 - 3.1 0.0
-I0
-0.6
0.5
-23”
-0.3
0.4
-30°
- 0.5
0.2
strain
and
“direct.”
indicates
the direction
of
210
Fig. 3. The records occurring
of coseismic
in the central
Kanto
strain
step observed
by the three-component
area (M = 6.0) on February
borehole
strainmeter
(No.
1) for the earthquake
27, 1983.
did not determine the depths of the faults, SO we presume reasonable values for these depths. Fig-
Focal
mechanisms
reported
by Dziewonski
et al.
earthquakes whose coseismic strain steps are observed. The epicentral distances ranged from 20 to
(1983a, b; 1984a, b, c; 1985) were determined by the centroid-moment tensor method, using the records of long-period seismometers. The mechanism of the Off Ibaraki earthquake (No. 1 in
480 km.
Table
Theoretical
termined using long-period body waves. Shimazaki and Mori (1983) used P-wave initial motions,
ure
4 shows
the
epicentral
distribution
of the
calculation
3) by Kikuchi
and
Sudo
(1985)
Focul mechanisms
long-period
Table 3 shows the focal mechanisms and the seismic moments of the earthquakes in Table 1.
termine the mechanisms of the Off Akita earthquake (No. 5). They did not determine dip angle and slip angle of the second event in the earth-
TABLE Origin
No.
body waves and surface
was de-
waves to de-
2 times. empirical Origin
region,
time (JST)
1
Jul. 23 1982
2
Jul. 24 1982
23:24 2154
epicentral
coordinates
Region
Epicentral
according
to the JMA of the earthquakes
listed in Table 1
coordinate
M
lat.
long.
depth
(north)
(east)
(km)
Off Ibaraki
36.18”
141.95 o
30
JMA
7.0
Off Ibaraki
36.12”
142.05 o
30
JMA
6.2
3
Aug. 14 1982
6:14
4
Feb. 27 1983
21:14
Southwest
5
May 26 1983 11:59
Off Akita
7:03
and magnitudes
Off Ibaraki
reference
36.48 o
141.22O
40
JMA
5.4
35.94O
140.16 o
72
JMA
6.0
40.4O
139.3O
15
Shimazaki
7.1
40.9 o
139.4 o
15
Off Fukushima
36.91 o
141.19O
54
JMA
5.8
Southeast
35.54O
139.05 o
18.3
NRCDP
6.0 5.1
Ibaraki
6
Jul. 2 1983
7
Aug. 8 1983
12:48
8
Oct. 28 1983
lo:50
9
Dec. 30 1983 11:30
Near Choshi
10
Jan. 1 1984
18:03
Off Kii Penin.
Southwest
Yamanashi Ibaraki City
and Mori (1983)
36.23”
139.99O
51.6
NRCDP
35.73O
140.73O
45.5
NRCDP
5.3
33.62O
136.84”
388
JMA
7.3 5.6
11
Jan. 17 1984
20:13
Off Ibaraki
36.45 o
141.25 o
43
JMA
12
Jan. 18 1984
0:31
Off Ibaraki
36.45 o
141.27”
43
JMA
5.9
13
Sep. 19 1984
2:02
Off Boso Penin.
34.05 o
141.55O
13
JMA
6.6
211
TABLE
3
Focal mechanisms No.
and seismic moments
of the earthquakes
listed in Table 1 *
Focal mechanism strike
dip
Seismic moment slip
( x
reference
Reference
70 26 dyne-cm)
1
190”
9”
80”
2.8
Kikuchi
2
231°
10”
110”
Dziewonski
et al. (1983a)
0.15
Dziewonski
et al. (1983a)
3
18X0
26O
84”
Dziewonski
et al. (1983a)
0.011
Dzrewonski
et al. (1983a)
4
191.1”
26.4”
0.078
Dziewonski
et al. (lY83b)
5
15O
25”
Shimazaki
and Mori (1983)
Kikuchi
100.0 o
and Sudo (1984)
EPID. JMA (1983)
85O
Shimazaki
and Mori (1983)
25
and Sudo (1985)
~ IO0
25”
85”
6
33O
85”
119O
Dziewonski
0.034
Dziewonski
et al. (1984a)
7
268O
76”
165”
lmoto et al. (1984)
0.037
Dziewonski
et al.
8
70”
14O
lmoto (1984)
0.0045
9
0.018
Dziewonskr
et al. (1984b)
6.0
Dziewonski
et al. (1984~) et al. (1984~)
33
YOO
(1984a)
231.4O
30.3O
10
163”
74”
73”
11
206O
43”
112”
EPID, JMA (1984)
0.028
Dziewonski
12
217O
48”
119”
EPID, JMA (1984)
0.045
Dziewonski
et al. (1984~)
13
137O
73O
Dziewonski
2.1
Dziewonski
et al. (1985)
* Azimuth
quake,
132.7 o
et al. (1984a)
-69’
of the strike is measured
so we suppose
Imoto (1984) Dziewonski
that
clockwise
these
et al. (1984~)
et al. (1985)
from the north
angles
had
the
same values as those in the first event. The seismic moment of the Off Akita earthquake was recalculated using the value of 6 x 10” dyne/cm2
for the
and slip angle is positive
rigidity
for reverse faulting.
of the crust instead
of 3 X 10” dyne/cm2,
as adopted by Shimazaki and Mori (1983) because we calculated theoretical strain steps at Yasato station using the value of 6 X 1O1’ dyne/cm2. Also, we assumed that the fault lengths and widths of the first and the second events of the Off Akita earthquake to be 35 x 30 km and 60 x 35 km respectively, based on the distribution of its aftershocks.
Focal
mechanisms
from Imoto
(1984), Imoto et al. (1984) and the Earthquake Prediction Information Division, JMA (1983, 1984) were determined by P-wave initial motions recorded with short period seismometers. We used the seismic moments calculated with the focal mechanisms in the same reports for the abovementioned earthquakes. For the earthquakes with no estimation of the seismic moments in the reports, we used the seismic moments determined by centroid-moment tensor method (1983b, 1984a, b, c). For the southwestern Ibaraki earthquake (No. 8) the seismic moment has not yet been determined, so we estimated the moment using the empirical relationship between magnitude M and seismic moment M,,, given by Kasahara (1975): log M,, = 1.5M + 16.0 Fig.
4.
coseismic
Epicentral strain
distribution steps
(YST) is also plotted.
of
are observed.
the
earthquakes
Location
whose
of strainmeter
Here, M as determined the calculation.
(2) by the JMA
is used for
212
82.
off
7.23. lbaraki
82.
off
7.24. lbaraki
82.
off
8.14. lbaraki
83. SW
83.
2.27.
off
lbaraki
5.26.
83.
kit.3
7.
2.
Fukushima SE
off
83.
8.
8.
Yamanashl
I No.1
-
Fig. 5. Comparison
Calculated
between
the observed
strain step
and the calculated
strain
with the observed of
the
From the focal mechanisms and seismic moments shown in Table 3, we calculated the strain steps at the Yasato station. For the calculation we
calculate
assumed a point source except for the Off Akita earthquake (No. 5) and a uniform semi-infinite body for the medium. For the Off Akita earthquake, we used the formula for a finite fault given
servation
steps.
steps, we discuss
measurements
and
the theoretical
the instrumental sensitivity
the
models
steps. Firstly,
constants,
of the instrument,
the reliability used
to
we examine
the azimuth
and
the
which vary with ob-
point.
by Sato and Matsu’ura (1974). For both cases, the P- and S-wave velocities were assumed to be 6
The azimuths of the instruments are determined during installation using a magnetic compass. The measurement error is within 3”. The difference in the directions of the maximum extensions between the observed and the calculated
km/s and 3.5 km/s for the medium, respectively. Figure 5 shows the comparison between the ob-
steps shows the mean values and standard deviations of 1” + 93 and lo JI 12” for No. 1 and No. 2
served and the calculated strain steps. There is no case in which the observations are not in harmony with the calculation, although the agreements between the observed steps of instrument No. 1 and No. 2 are better than those between the calculated and observed steps. Discussion
Reliability
of meusurements
Based on the observations of strain steps, and the comparison of the theoretically predicted steps
respectively.
From
these values,
it seems that the
azimuth measurements using a magnetic compass do not have any systematic errors larger than 10 O. On the other hand, the sensitivity of the instruments is a function of the elastic constants of the surrounding rock, the mortar and the instrument, and the geometry of the detection part of the instrument. Although physical constants are measured by well logging and core sample tests, an error might arise from estimations of the elastic constants of surrounding rock, because the surrounding rock usually has 3-dimensional inhomogeneity. The detection part of the strainmeter is
213
very slender
and the sensitivity
two-dimensional
by this geometrical to or less than sensitivity. and
From
strains ments
percent
of the observed
may be equal of the value
the comparison steps,
on a
The error caused
approximation
several
the calculated
ratios
is calculated
approximation.
of
of the observed
the mean
values
to the calculated
of the
principal
are 1.08 i 0.61 and 1.00 f 0.87 for instruNo. 1 and No. 2 respectively.
indicate
that the estimation
These values
amplitude
those steps,
and
the
are
observed
probably
earthquake,
if coseismic
nanostrain
are available
the three-component observation a network several
points
of stations
1 x 1O’4 dyne-cm,
the
inho-
instruments or azimuths
of the instruments. of models
In the theoretical calculation of strain steps, we assumed the medium to be a uniform semi-infinite body, neglecting the effect of layered structure and sphericity. Those effects should increase in
depth
although
occurring of more
the resolution
on the mechanism
and
the
of the earthquake.
(1) These instruments
simultaneously
strain steps for 13 earthquakes, from 5.1 to 7.7. The coseismic tained
from
the
two
straincentral recorded
with M ranging strain steps ob-
independent
strainmeters
showed a good agreement for all the earthquakes. The differences in magnitude and direction of the horizontal maximum strain are mostly within a factor of 2’ and 10” respectively. (2) The coseismic strain steps observed were found to be in satisfactory agreement with those theoretically computed from the focal mechanism and
served and the calculated steps does not seem to become larger for the earthquakes with long epi-
bility
for instance the southeast Yamanashi earthquake (Imoto et al., 1984) the discrepancy between the seismic and strain observations seems to become larger than in other earthquakes (Fig. 5). The comparison between the observed and the calculated strain steps suggests that it would be possi-
on a scale of
the focal mechanisms
Two sets of three-component borehole meters were installed 60 m apart at Yasato, Japan in 1982.
gated.
that caused by other uncertainties. Although it is difficult to evaluate the focal models from the strain steps observed at only 1 point, the observed steps are reasonably consistent with the focal mechanisms as shown in Fig. 5. For the earthquakes with a complicated focal process,
When
Conclusion
proportion to the epicentral distances for shallow earthquakes, but the discrepancy between the ob-
central distances. So it seems that the error caused by the approximation of the medium is less than
of
positioned.
with seismic moments
slightly
an 0.3
and the
for the earthquakes
the network
with than
network,
will be estimated
depend
by
larger
is constructed
tens of kilometers,
than
the calculated
steps
are properly
may
and
including
associated
from at least 2 stations
strain
caused
mechanism,
strainmeter
of the observed
mogeneity of the rock around the rather than by the errors of sensitivities
Reliubility
the fault deformation
has a
between the observed strain steps of No. 1 and No. 2 are smaller than
between
aseismic
of the sensitivity
steps ranges between 100 and 0.3 nanostrain. The unit of analogue/digital conversion is 50 picostrain for a dilatation change and it is difficult to detect the steps smaller than 0.3 nanostrain. The differences instruments
any
within
high reliability. The maximum
ble to determine
seismic
(3) These of the
moment results
of the earthquakes demonstrate
three-component
investi-
the high strainmeter.
measurement of static deformation by strainmeter will make it possible to evaluate
reliaThe this the
“ total fault mechanism” including aseismic deformation, whereas the focal mechanism determined by seismic waves cannot detect any aseismic motion associated with seismic faulting. Acknowledgements The authors wish to express our gratitude to Dr. Yoshimitsu Okada for his permission to use the computer program to calculate static deformations by faulting. We also like to thank Dr. Masakazu Ohtake for his critical reading of the manuscript.
214
References
17, 1973.
In: Y. Sakai
Earthquake Dziewonski,
A.M., Friedman,
J.H..
Global
1983a.
tensor
solutions
Inter.,
33: 76-90.
Dziewonski,
A., Giardini,
seismicity
for 308 earthquakes.
A.M., Friedman,
Earth
A. and Woodhouse,
tensor
1983. Phys. Earth
solutions
Planet.
Inter.,
A.M., Franzen,
Centroid-moment
for
for
tensor
solutions
Planet.
Inter., 34: 129-136.
A.M., Franzen,
Phya. Earth Dziewonski,
Planet.
A.M.,
1984. Phys. Earth Agency,
Ibaraki
Prefecture,
for Earthquake Earthquake
1983.
ture on January Earthquake Furuya,
J.H.,
Division,
30: 68-69
Notes
Nat.
Japanese.
Res. Cent.
for
S., Okada,
M.. 1984. The Yamanashi, August Cent.
Comm.
8. 1983 and related for Disaster
Prefecfor
Disaster
volume
strainme-
thesis.
Univ.
area revealed
earthquakes. Prev.,
Res.
66: l-78
Japan,
seismicity.
Prev., 57: 1-13
(in
K. and Ohtake. earthquake
of
Res. Notes Nat. Res.
(in Japanese.
Prog. Abstr. Kasahara.
Seismol.
with Engl.
value of fault
parameters.
ated with the earthquake
distribution
of strain-steps
off the Nemuro
associated
with
Peninsula,
associJune
the
D.W. and Yamaguchi.
strainmeter;
Geophys.,
strain
with Engl. abstr.).
results
its installation
concerning
strain
Y..
in Japan steps. Pap.
22: 195-208. S. and Noguchi,
observation
S.. 1982a.
in Kanto
District,
three-component
Continuous a triple junc-
borehole
strainmeMay
7715. 1982, pp. 226-232. S., Shimada,
of new-type Proc. quake Sakata,
Panel
Prediction
S., Shimada.
of UJNR
strainmeters.
In:
Panel on Earth-
Geographical
Survey
In-
S., 1987. Development
of
pp. 234-248.
S. and Noguchi,
three-component
R. and
surface
Meeting
Tsukuba,
S., 1982b. Development
borehole
Technology.
of Japan,
a borehole Sato,
S. and Noguchi,
three-component
3rd Joint
strainmeter
and some results
(in prep.).
Matsu’ura.
M., 1974. Strains
of a semi-infinite
medium.
and
J. Phys.
tilts on the Earth,
22:
213-221. R. and
vault.
Okada,
J. Geod.
Y., 1979. Strain
Sot. Jpn..
25: loll134
measurement
in the
(in Japanese,
with
Engl. abstr.). Shimada,
S., Sakata.
observation
S. and Noguchi,
and data strainmeters.
Prev., 31: 101lllO Shimazaki,
Sot. Jpn., 1: 8 (in Japanese).
M.. 1974. Spatial
tilt-steps
of 1974 and Im-Oshima-Kin-
S., Evertson,
S.. Shimada,
borehole K.. 1975. The standard
in a
Sot., 70:
of 1978. Zisin (J. Seismol. Sot. Jpn.) II. 33:
(in Japanese,
Meteorol.
and
earthquake
1971. Sacks-Evertson
Shichi, Y., Kasahara, Central
Strain-
IS., Suyehiro,
stitute
abstr.). Kasahara.
Sacks,
Sakata, MeteoroComm.
Ph.D.
of small-sized
J. R. Astron.
ters. In: Proc. of the Gen. Meet. of the IAG. Tokyo,
with Engl. abstr.).
M., Shimada.
Y., 1980.
IzuHanto-Oki
of observation
mechanisms
Geophys.
strain
strainmeters
tion area, by new-type
Tokyo. focal
Sacks-Evertson
southern
Japan
68:
S.M. and
and other short-term
Meteoro-
(in Japanese).
of borehole
A.T., Spottiswoode,
Japan
off Ibaraki
to seismology.
with
deep mine, South Africa.
Sakata,
(in Japanese).
Division,
of the July
Monthly,
in the
17 and 18, 1984. Rep. Coord. 32: 96-97
recorded
crustal
27, 1983. Rep. Coord.
1984. The earthquakes
Predict.,
1985.
July-September
M., 1984. Stress field in the Kanto-Tokai
from
Imoto,
for
earthquake
I., 1984. Characteristics
Tokyo,
1984.
38: 2033213.
Information
ter and its application Imoto,
The
February
Predict.,
Agency.
Inter.,
Information
Prediction
logical
solutions
I.S., Linde,
and some preliminary
J.E. and Woodhouse.
Planet.
Prediction
logical
J.H., 1984~.
for January-March
The Earth
R.W.E.. 1982. Coseismic
kai earthquake
Inter., 34: 2099219. tensor
A., Sacks,
5255539
J.E. and Woodhouse,
Franzen,
Centroid-moment Earthquake
J.H.. 1984b.
for OctoberrDecember
tensor solutions
McGarr,
Okada,
1983. Phys. Earth
earthquake.
717-740.
July-September
Inter., 34: 1-8. J.E. and Woodhouse,
on the
(in Japanese).
Green,
J.H., 1984a.
Report
June 17, 1973. pp.
with Engl. abstr.).
23, 1982 off Ibaraki
changes
Centroid-moment
Centroid-moment
J.H., 1983b.
33: 71-75.
solutions
Planet.
A.M., Franzen,
Dziewonski,
Planet.
January-March.
J.E. and Woodhouse,
tensor
1983. Phys. Earth Dziewonski,
Phys.
(in Japanese,
General
Peninsula,
M. and Sudo, K., 1985. The focal process
Kikuchi,
centroid-moment
72-78
Centroid-moment Dziewonski,
36-62
D. and Woodhouse.
of 1982:
(Editor),
off the Nemuro
processing Rep.
S.. 1983. The methods
Nat.
(in Japanese,
Res. Cent,
for Disaster
with Engl. abstr.).
K. and Mori. J., 1983. The focal mechanism
May 26, 1983 Japan Sot. Jpn., 2: 15.
of
for the three-component
Sea earthquake.
Prog. Abstr.
of the Seismol.