Long-term Earthquake Prediction Along the Seismic Zone of the Solomon Islands and New Hebrides Based...

Natural Hazards 11: 17-43, 1995. 17 1995 Kluwer Academic Publishers. Printed in the Netherlands.

Long-Term Earthquake Prediction along the Seismic Zone of the Solomon Islands and New Hebrides Based on the Time- and Magnitude- Predictable Model

D. G. PANAGIOTOPOULOS Geophysical Laboratory, University of Thessaloniki, GR-54006 Thessaloniki, Macedonia, Greece

(Received: 28 July 1993; in final form: 16 February 1994)

Abstract. Instrumental and historical information on strong main-shocks for 13 seismogenic sources along the seismic zone of the Solomon Islands and New Hebrides has been used to show that the interevent time, Tt (in years), between two strong earthquakes and the magnitude, Mf, of the following mainshock are given by the relations

log Tt = 0.17Mmin + 0.31Mp - 0.33 log Mo + 6.36,

Mf = 0.51Mmin - 0.12Mp + 0.541 log 3)10 - 9.44,

where Mmln is the surface wave magnitude of the smallest main-shock considered, M v is the magnitude of the preceding mainshock, and Mo is the moment rate per year in each source. On the basis of these relations, the probability for the occurrence of a mainshock during the decade 1993-2002 as well as the magnitude of this expected mainshock in each seismogenic source has been calculated. The highest probability value (P10 = 0.79) was estimated for the seismogenic sources of Santa Cruz-Ndeni Islands (H1) and Tana Island (Hs) for the occurrence of a large or great earthquake with expected magnitudes Mf = 7.7 and 7.5, respectively.

Key words: Long-term earthquake prediction, time- and magnitude-predictable model, the Solomon Islands, New Hebrides

1. Int roduct ion

In the last decade many projects and ideas have been appl ied in the field of

seismology and geophysics towards understanding the occurrence of large earthquakes over the world along simple plate boundar ies. To est imate the long term

probabi l i t ies for the generat ion of strong earthquakes on single faults, the t ime-

predictable model seems to be more plausible than the s l ip-predictable model

(Wesnousky et al., 1984; Ast iz and Kanamor i , 1984; Nishenko and Buland, 1987).

This mode l also holds if the seismic source includes other small faults where smal ler

mainshocks occur besides the main fault where the characterist ic ear thquake is

generated (Papazachos, 1989).

Accord ing to the t ime-predictab le model , the occurr ing t ime of a future earth-

quake in a certain seismogenic source depends on the size of the last ear thquake in the source. On the other hand, according to the s l ip-predictable model , the size

18 D. G. PANAGIOTOPOULOS

of a future earthquake depends on the time elapsed since the last earthquake. !t means that we can predict, in principle, the size of a future earthquake by the slip-predictable model or the time of its occurrence by the time-predictable model (Bufe el al., 1977; Shimazaki and Nakata, 1980).

The comparison of the data on the inland seismic activity of western Japan (Mogi, 1985) showed that the time-predictable model holds very well and differs significantly from what was expected from slip-predictable model. The data on the geometry, seismic moment and repeat time of large earthquakes of both the strike- slip and convergent types, on the other hand, fit very well with the time-predictable model of earthquake recurrence in several areas over the world (Sykes and Quit- meyer, 1981)

Recently, Papazachos (1988a, b, 1989, 1991, 1992, 1993) concluded that the time-predictable model holds very well for the strong shallow earthquakes which occurred in seismogenic sources in Greece. He proposed two formulas where the interevent time, Tt, as well as the magnitude, M/, of the following mainshock were related to the magnitude, Mmin, Of the smallest mainshock considered and to the magnitude, Mp, of the preceding mainshock in each seismogenic source. Very recently, Papazachos and Papaioannou (1993) improved this idea by including a new term in both of these formulas, which depends on the yearly moment rate, Mo, in each seismogenic source. This is called the time- and magnitude- predictable model. They used this model to estimate the probability of occurrence of the next mainshock during the next decade and the magnitude of this shock for each seismogenie source in the Aegean area.

An attempt is made in the present study to test the applicability of the time- and magnitude-predictable model in the seismic zone of the Solomon Islands and New Hebrides with data from shallow earthquakes. Figure 1 shows the studied area along with its main tectonic characteristics.

It is generally known that the Pacific Plate spreading from the East Pacific Rise is subducted beneath the Eurasian Plate at the Aleutian-Kurile-Japan-Mariana Trench system. The area in the present study is the northern boundary between the Pacific Plate and the Indian Plate. Along this boundary, the Indian Plate is subducted beneath the Pacific Plate to form the Solomon Islands and New Heb- rides ensimatic arcs (Mitchell and Warden, 1971; Karig and Mammerickx, 1972), In the central part, the Pacific Plate is subducted beneath the Indian Plate to form the Tonga and Kermadec ensimatic arcs (Oliver and Isacks, 1967; Karig, 1970, 1971). The pole of the convergence is given by Walcott (1978) as 60 S, 180 and by Chase (1978) as 62 S, 174 E.

The main tectonic features of the studied area are the S. Solomon Trench, the New Hebrides Trench, the Torres Trench, the Polkington Trough, the Loyalty Rise, the North Fiji Basin, the South Fiji Basin, the Vityaz Trench and the Hunter Fracture Zone (Wellman and McCracken, 1979; Cole, 1984; Packham, 1982; Neef, 1982; Davey, 1982; Isacks et al., 1981) (see Figure 1).

Because of the frequent occurrence of large doublet events that are temporally

LONG-TERM EARTHQUAKE PREDICTION 19

0 o

20OS

CAROLINE PLATE

AUSTRALIA

150OE

I

? ,~0 ~

INDIAN PLATE

180 I

. ~ ACTIVE DIVERGING BOUNDARY

ACTIVE CONVERGING BOUNDARY

TRANSFORM OR TRANSCURRENT FAULTS

RELATIVE MOVEMENTS AT PLATE BOUNDARIES

PACIFIC PLATE

~ ~ ~K- FIJI

rtUNxtR

0 o

20os

150OE 180

Fig. 1. The studied area with the main tectonic features (modified from Cole, 1984; Wellman and McCracken, 1979).

and spatially linked, Lay and Kanamori (1981) grouped the seismic zone of the Solomon Islands in the category '2' of the subduction zones which have demon- strated temporal variation in rupture extent, with occasional ruptures 500 km long. Category '3' of this grouping is characterised by repeated ruptures over limited zones with no great earthquakes because of a large component of aseismic slip or subducting ridges. The area of New Hebrides with small ruptures and a strong tendency for earthquakes to cluster in time and space, has to be considered intermediate between the categories '2' and '3'.

2. Method

Using the interevent times of strong mainshocks in the seismogenic sources in Greece, Papazachos and Papaioannou (1993) proposed two relations of the following forms

20 D.O. PANAGIOTOPOULOS

log T~ = bMrnin + cMp + d log 3;/0 + t, (1)

Mf = BMmin + CMp + J log/9/0 + m, (2)

where Tt is the interevent time (in years), Mmin the surface wave magnitude of the smallest mainshock considered, Mp the magnitude of the preceding mainshock, M r the magnitude of the following mainshock and 3;/o the moment rate in each source per year. The parameters (b, c, d, t, B, C, D, m) of (1) and (2) are calculated using the available seismicity data for all sources in the studied area.

The moment rate 3;/o for a seismogenic source can be reliably calculated if enough data are available for the source. In the present study, the values of the moment rate /f/o were determined by applying a method suggested by Molnar (1979). Using this method, the moment rate was calculated on the basis of the maximum magnitude, Mmax, and of the parameters a and b' of the Gutenberg and Richter (1944) relation

log N = a - b 'M, (3)

normalized for one year, as well as of the parameters r, k of the moment- magnitude relation

log Mo = rM + k (4)

which for the studied area have been taken as r = 1.5 and k = 16,1 (Kanamori, 1977). On the basis of a well known technique (Draper and Smith, 1966; Weisberg, 1980), which has been used in strong motion attenuation studies (McGuire, 1978; Joyner and Boore, 1981; Dahle et al., 1990), the coefficients b, c, d and t of (1), and the corresponding coefficients of (2) were determined by use of all data of the 13 seismogenic sources (using a computer program written by C. Papazachos).

3. Seismogenic Sources and Data

An important step for this work was to define the seismogenic sources of the shallow earthquakes in the area under study. Thirteen seismogenic sources were finally defined in this area on the basis of tectonic features (Figure 1), rupture zones (Figure 2) and clustering of epicentres of the mainshocks (Figure 3). The intermediate depth earthquakes at a depth larger than 40 km in the area (Isacks et al., 1981; Ripper, 1982; Cooper and Taylor, 1987) were excluded.

Along the northern part of the tectonic line of the Solomon Trench near the Islands of Bougainville and Choiseul, two seismogenic sources were defined (sources $1 and $2 in Figure 3) on the basis of the spatial distribution of the epicentres of the large event which occurred on 30 January 1939 (Ms = 7,6) in the seismogenic source $1 and the large event on 19 April 1936 (M, = 7.2) in the seimogenic source $2. The position of the seismogenic source $2 between the seismically active Bougainville and inactive Woodlark segments of the Solomon arc suggests that it is a transitional region (Nishenko, 1991).

LONG-TERM EARTHQUAKE PREDICTION

165OE i ' i i , l r , i

175OE I

/ PACIF IC OCEAN

l~S

20oS

.2\

\ l

CORAL SEA e~

FIJI ISLANDS

~. . .~ ......

1t 8

. '~ 4

21

0 106 212 Fan - II 7

, I ~ I , J 165OE 1750E 180OE

Fig. 2. Estimated rupture zones in the area under study since 1934 (modified from Kelleher et al., !974; Isacks et al., 1981).

The region near New Georgia Island, which is of extensional tectonics, has not experienced large earthquakes since the turn of the century and is considered to have a low seismic potential for such earthquakes (Nishenko, 1991). According to McCann et al. (1979), this region is a large gap which occupies that segment of the arc that intersects the Woodlark rise, a series of spreading centres that forms the boundary between the Solomon Sea and Indian Plate. On this basis, the New Georgia region is considered as one seismogenic source (source $3 in Figure 3).

The segment of the Solomon Plate boundary near the western part of Guadal- canal Island exhibits a clear cluster for the epicentres of the mainshocks on 25 May 1901 (Ms = 7.1), 3 February 1939 (Ms = 7.1) and their aftershocks sequences. Lithosphere in that area appears to be too weak to store the energy to generate


a o

, |

0 o

i ,~ ,o: r

~t

"S J N

e

i oo

0

) z~

) o

o

g


large earthquakes (McCann et al., 1979). According to these, this area is considered as one seismogenic source (source $4 in Figure 3).

The region of San Cristobal has been ruptured in some great mainshocks, on 3 October 1931 (Ms = 7.7), on 15 June 1966 (3//, = 7.7) and on 7 February 1984 (M, = 7.5). According to Nishenko (1991), this region has seen two episodes of high earthquake activity, 1926-1935, 1977-1988 and the cumulative seismic moment release in the interval 1926-1935 is approximately 2.3 x 1028 dyn cm, while the moment release in the sequence from 1977 to 1988 is about 33% of the earlier (i.e. 7.5 x 1027 dyn cm). Nishenko (1991) based on this point considered that the end of the current sequence may still lie in the future. Because of the spatial distribution of the epicentres of the mainshocks and their aftershocks and of the bathymetric features, the region of San Cristobal is considered as one seismogenic source (source $5 in Figure 3).

The region between 162 E and 164 E, which is the western part of the Santa Cruz Strait, defines the transition zone between the Solomon Trench and the New Hebrides Trench and has been aseismic for large earthquakes with Ms > 7.1 (Nishenko, 1991). On this basis, the zone is considered as one seismogenic source (source $6 in Figure 3).

In the northern part of the New Hebrides Trench from the Ndeni Island to Torres Islands, between 10S and 13.0S, the New Hebrides Plate boundary exhibits two clear clusters for the great earthquakes which occurred on 18 July 1934 (214, = 7.9) and on 31 December 1966 (Ms = 7.9) and their aftershocks. The great earthquakes on 18 July 1934 (Ms = 7.9), on 31 December 1966 (Ms = 7.9) and on 17 July 1980 (Ms = 7.7) have ruptured similar segments of the arc in the New Hebrides Trench, but exhibit some different details in the rupture process (Nishenko, 1991). The 1980 earthquake ruptured the Vanikoro asperity that acted as a barrier for the southern expansion of the 1966 rupture (Tajima et al., 1990). On the other hand, according to McCann (1980) the aftershock area of the 1934 mainshock overlapped both the 1966 and 1980 aftershocks areas. Based on these clusters and their estimated ruptures zones in Figure 2 (Kelleher et al., 1974), the region of Santa Cruz-Ndeni Island is considered as a seismogenic source (source H1 in Figures 2 and 3).

Along the tectonic line of New Hebrides Trench from 13.5S to 15.5S at the northern part of the collision's boundary between the Pacific Plate at the New Hebrides Trench, three large earthquakes occurred on 11 August 1970 (Ms = 7.0), 28 December 1973 (Ms = 7.3), and 10 January 1974 (Ms = 7.0) and its rupture zones were estimated by Kelleher et al. (1974) and Isacks et aI. (198l) (Figure 2). According to this, the region of Banks Islands and Espiritu Santo Islands is considered as a seismogenic source (source H2 in Figures 2 and 3).

In the central part of New Hebrides Trench, from 15.0S to 17.5 S, at the region of Malekula Island a clear rupture zone of two large earthquakes on 11 August 1965 with magnitudes Ms = 7.1, and Ms = 7.3 and another one on 13 August 1965 with Ms = 7.1 have been estimated by Kelleher et al. (1974) (Figure

24 D.G. PANAGIOTOPOULOS

2). On this basis the region of Malekula Island is considered as the seismogenic source H3 in Figures 2 and 3.

The feature of Efate Island's region between 17.5S and 19~0S is that there are fewer large shocks than the rest of the plate boundary at New Hebrides Trench (Nishenko, 1991). On the other hand, the rupture zone of a large earthquake on 23 July 1961 with 34, = 7.3 has been estimated by Kelleher et al: (1974) (Figure 2). On this basis, the region of Elate Island is considered as the seismogenic source H4 in Figures 2 and 3.

To the south from 19S to 21S in the region of Tana Island Kelleher et aL (1974) estimated two rupture zones (Figure 2). In the first rupture zone, four large earthquakes occurred in 1950 with Ms = 6.8-7.2 and in the second a large event occurred on 2 November 1972 with M~ = 7.0. In addition, the great earthquake on 20 September 1920 (Ms = 7.7), which is located near the axis of the trench, and generated a tsunami observed on Samoa (Iida et aI., 1967), exhibits a clear cluster at this region. On this basis and the spatial distribution of the strong and large earthquakes, the region of Tana Island is considered as the seismogenic source Hs in Figures 2 and 3.

In contrast to these seismogenic sources (H4, gs), the region of Matthew Islands (source H6 in Figure 2) exhibits frequent large mainshocks (Vidale and Kanamori, 1983). Additionally, Kelleher et al. (1974) estimated the rupture zone of a large event on 14 September 1943 (Ms = 7.2) (Figue 2). On this basis and the clear cluster of the large earthquakes (Figure 3), the region of Matthew Island is considered as the seismogenic source H6 in Figures 2 and 3.

The region of the Hunter Islands along the Hunter fracture zone, from 171E to 176 E, is considered as one seismogenic source (source H7 in Figure 3), because there is no history of great earthquakes in this segment and only two large mainshocks with Ms ~> 7.0 occurred on 6 July 1981 (Ms = 7.0) and on 3 March 1990 (Ms = 7.4).

These 13 seismogenic sources are shown in Figure 3, along with the epicentres of the complete data of the shallow earthquakes for which data are used in the present study. Black circles show epicentres of the mainshocks. Open circles show epicentres of the foreshocks and aftershocks in the broad sense, that is, earthquakes which may occur up to several years before or after the main shock. The terms 'foreshocks' and 'aftershocks' are used in their broadest sense because a model that can predict the mainshocks is required, that is, the strong earthquakes which occur at the beginning and the end of each seismic cycle and not smaller earthquakes which occur during the preseismic and postseismic activations. This concept of 'foreshocks' is in accordance with Mogi's (1985) suggestion that the seismic activity over a wide area would increase through a rise in crustal stress and that these events are foreshocks in the broadest sense. The same author also suggested that according to the time-predictable model, the preseismic activity is constant, while the aftershock activity depends on the magnitude of the preceding mainshock. On the other hand, Karakaisis et al. (1991) found that the last phase


Table 1. Basic parameters which are used for every source. The code number and the names of the sources are given in the first and second columns. The constants for the Gutenberg-Richter relation are given in the next two columns and the maximum magnitude and the logarithm of the moment rate are given in the last two columns

25

Seismogenic sources a b' mmax log IVlo

Source $1 Bongainville Isl. 6.88 1.2 7.6 25.96 Source $2 Choiseul Isl. 6.95 1.2 7.2 25.91 Source $3 New Georgia Isl. 6.67 1.2 7.2 25.63 Source $4 Guadalcanal Isl. 6.81 1.2 7.1 25.74 Source $5 San Cristobal Isl. 7.48 1.2 7.7 26.59 Source $6 Santa Cruz Strait 6.78 1.2 7.1 25.71 Source H1 Santa Cruz-Ndeni Isl. 7.58 1.2 7.9 26.75 Source H2 Espiritu Santo Isl. 6.93 1.2 7.3 25.92 Source H3 Malekula Isl. 7.18 1.2 7.3 26.17 Source H4 Elate Isl. 6.72 1.2 7.3 25.71 Source H5 Tana Isl, 7.09 1.2 7.7 26.20 Source H 6 Matthew Isl. 7.13 1.2 7.8 26.27 Source H7 Hunter Isl. 6.88 1.2 7,4 25.90

of the seismic cycle is characterized by an accelerated activity period with an average duration of 2.7 years, which culminates before the second main shock and does not depend on the magnitudes of the first and second mainshocks. For this reason, the foreshock period is considered constant and equal to three years, As aftershocks, earthquakes have been considered which had followed the mainshocks in a time period Ti (years) which is related to the magnitude Ms of the mainshock by the relation:

log(T/) = 0.06 + 0.133//,. (5)

This relation has been derived by the use of a very large sample of data (Papa- zachos, 1993 personal communication). Information on the magnitudes and on the epicentres of the earthquakes plotted in Figure 3 were taken from the catalogue of Pacheco and Sykes (1992) for events which occurred during the present century with M~ ~> 7.0. The magnitudes of historical events and for events not listed in the previous catalogue as well as for events with magnitudes less than 7.0 have been obtained from Abe (1981), Abe and Noguchi (1983), and Tsapanos et al. (1990). In addition, the more recent events for the period 1986-1992 were taken from ISC bulletins.

The values of the parameters b', a, Mmax and log 21;/o for each seismogenic source are listed in Table I. The name and the number for the seismogenic sources, corresponding to the sources of Figure 3, are listed in the first two columns of this table. A common value for the parameter b' (1.20) was estimated for all seismogenic sources for the area under study, by applying a least-squares method, in the range of the completeness from all available data. The value of Mmax by consider-


ing all available historical and instrumental data for each seismogenic source were estimated.

Table II lists all the information on the data used in the present study. A name is written for each seismogenic source in the first column of this taNe. The date, epicentre and surface wave magnitude for the shocks, which satisfy the completeness condition, are given in the second, third, and fourth columns of this table. The fifth column shows the cumulative magnitude, M, of each sequence, that is, the magnitude which corresponds to the total moment released by the major shocks (mainshocks and large foreshocks and aftershocks) of each seismic sequence according to the relation suggested by Kanamori (1977). In this study, these cumulative magnitudes instead of the magnitudes of the mainshocks were used.

The magnitude Mm~n of the smallest mainshock, the magnitude Mp of the preceding mainshock, the magnitude Mf of the following mainshock, the interevent (repeat) time T between the two shocks, the year tp of occurrence of the preceding mainshock, and the year t I of occurrence of the following mainshock, as they were derived from the data of Table II for each seismogenic source, are given in the columns of Table III for each source. In each seismogenic source, after considering the first minimum magnitude, Mmin, of all mainshocks, the interevent times, T, between succesive main shocks with magnitudes equal to or larger than Mmin are calculated. Then, the second minimum, Mmin, is considered and new interevent times between successive mainshocks with magnitudes equal to or larger than the second Mmin are calculated. This is continued until the last Mmin will be considered.

To understand the way in which these values were derived from Table II, the procedure for the first source will be explained. The procedure starts with the smallest mainshock which occurred in 1953 and the magnitude of which is Mmin = 6.0. The first repeat time (T = 4.48 years), which holds for this minimum, in the range of the completeness is the interevent time between the earthquake which occurred on 18 June 1953 (Mp = 6.0, 6 - 1953) and the mainshock which occurred on 10 December 1957 (M I = 7.0, tf = 1957) and is written in the first line of Table III with the corresponding values of Mmin, Mp, Mr, tp and tf. The second repeat time (T = 12.79 years), which also holds for Mmin = 6.0 is the interevent time between the mainshock which occurred on 10 December 1957 (Mp---7.0, rp = 1957) and the mainshock which occurred on 23 September 1970 (Mj = 6~7, t~ = 1970) and is written in the second line of Table III. In the same way the third line of Table III was derived.

In the next step the Mmin = 6.6 as it comes out from the fifth column of Table II, and the results corresponding to this minimum magnitude are shown in the fourth, fifth, sixth and seventh lines of Table III. The same procedure is followed until the last Mini, (7.0) is considered.


Table II. Basic information on the data used for each seismogenic source (f: foreshocks, a: aftershocks in their broad sense). All known shallow earthquakes that occurred during the following time periods and were equal to or larger than certain cut-off magnitudes were considered: (1) 1897-1992, Ms I> 7.0; (2) 1931-1992, Ms 1> 6.3; (3) 1953-1992, Ms ~> 6.0

Seismogenic source Date Epicenter Ms M Ref.*

Source $I Bougainville 6 4 1931 -7.0 155.0 6.7 f 5 24 4 1931 -6.5 155.0 6.9 f 5 29 1 1932 -6.0 155.0 7.1 7.3 2 29 1 1932 -6.5 155.0 6.5 a 5 30 1 1939 -6.5 155.5 7.6 7.6 1 8 3 1939 -6.0 155.0 6.7 a 5

18 6 1953 -6.5 155.0 6.0 6.0 5 11 9 1955 -6.9 155.2 6.0 f 5 24 6 1956 -7.0 154.9 6.2 f 5 10 12 1957 -6.6 154.8 6.8 7.0 5 20 9 1958 -6.5 154.9 6.3 a 5 29 6 1959 -7.3 155.2 6.1 a 5 10 9 1959 -6.4 154.9 6.0 a 5 23 2 1960 -6.0 154,5 6.1 a 5 6 2 1961 -6.8 155.3 6.6 a 5 6 7 1964 -6.3 154.7 6.4 a 5

23 9 1970 -6.5 154.6 6.5 6.7 4 13 8 1971 -6.2 154.1 6.2 a 4 13 8 1971 -6.2 154.1 6.2 a 4 23 3 1983 -6.5 154.6 6.2 f 4 23 3 1983 -6.6 154.6 6.0 f 4 5 7 1984 -6.1 154.4 6.5 6.6 4

Source $2 Choiseul Isl. 20 3 1935 -7.7 156.0 6.5 f 5 19 4 1936 -7.5 156.0 7.2 7.2 1 11 8 1946 -8.0 156.0 6.7 6.7 5 6 12 1952 -8.0 156.5 7.2 7.4 1 8 9 1955 -6.9 155.7 6.5 a 5

19 5 1956 -6.9 155.7 6.4 a 5 10 11 1957 -7.2 155.6 6.5 a 5 10 11 1957 -7.5 155.5 6.0 a 5 30 1 1958 -7.2 155.7 6.5 a 5 17 8 1959 -7.8 156.4 7.0 a 1 24 2 1960 -7.5 156.0 6.6 a 5 31 1 1974 -7.5 155.9 7.0 7.4 1 1 2 1974 -7.8 155.6 7.0 f 1 9 3 1974 -7.5 156.0 6.6 a 4

22 7 1975 -7.2 155.7 6.1 a 4 29 7 1977 -8.0 155.6 7.0 a I 1 9 1978 -7.5 156.5 6.2 a 4

15 10 1983 -8.1 156.4 6.7 6.7 4

Source $3 New Georgia Isl. 25 1 1926 -9.0 158.0 7.2 7.3 1 27 3 1926 -9.0 157.0 7.0 a 1 24 7 1963 -9.0 158.2 6.0 f 5 22 8 1963 -9.4 158.0 6.7 6.8 5 19 1 1968 -9.3 158.5 6.3 a 5 6 9 1969 -8.9 158.0 6.1 a 4

18 11 1976 -8.8 157.0 6.5 6.6 4 24 4 1983 -8.7 157.5 6.1 a 4


Table II. Continued

Seismogenic source Date Epicenter Ms M Ref. ~

Source $4 Guadalcanal Isl.

Source $5 San Cristobal

24 2 1986 -9.0 156.8 6.0 6.0 4 14 10 1991 -9.1 158.4 7.1 7.1 4 14 10 1991 -9.1 158.6 6.4 a 4 14 10 1991 -9.0 158.4 6.1 a 4

25 5 1901 -10.0 160.0 7,1 7.3 1 19 2 1906 -10.0 160.0 7.0 a 1 3 2 1939 -10.5 159.0 7.1 7.1 2 8 11 1950 -10.0 159.5 7.0 7.0 1

29 4 1953 -9.7 159.6 6.3 a 5 17 7 1965 -9.7 159.8 6.1 f 5 27 11 1965 -9.7 159.7 6.6 6.6 5 21 4 1977 -9.7 160.2 6.3 6.4 5 6 11 1979 -9.5 159.3 6.1 a 4

27 9 1985 -9.8 159.9 6.9 7.1 4 9 2 1991 -9.9 159.1 6.9 a 4 9 2 1991 -9.9 159.2 6.4 a 4

12 4 1926 -10.0 161.0 7.2 7.2 1 3 10 1931 -10.5 161.7 7.7 8.0 1 3 10 1931 -11.0 161.5 6.7 a 5 3 10 1931 -11.0 161.5 7.0 a 2

10 10 1931 -10.0 161.0 7.6 a 1 10 10 1931 -10.0 161.0 6.7 a 5 10 10 1931 -10.0 161.0 6.7 a 5 10 10 1931 -10.0 161.0 6.7 a 5 20 11 1931 -10.5 161.7 6,7 a 5 24 3 1934 -10.0 161.5 7.1 a 2 15 12 1935 -9.7 161.0 7.4 a 1 26 5 1955 -10.0 160.7 7.0 f 5 13 10 1955 -10.0 160.7 7.3 7.5 5 19 8 1957 -10.3 160.8 6.5 a 5 20 8 1957 -10.2 160.8 6.0 a 5 20 8 1957 -10.4 161.1 6.5 a 5 30 1 1959 -9.8 160.9 6.2 a 5 24 8 1959 -10.7 161.4 6.7 a 5 1 8 1961 -9.9 160.5 6.5 a 5

13 2 1963 -9.9 160.8 6.7 a 5 15 6 1966 -10.4 160.9 7.7 7.9 1 15 6 1966 -10.2 161.0 7.2 a 1 4 8 1972 -11.2 162.1 6.0 a 4 4 8 1972 -11.1 162.1 6.1 a 4

20 4 1977 -9.9 160.4 6.7 a 4 20 4 1977 -10.0 161.0 6.1 a 5 20 4 1977 -9.9 160.4 7.3 a 1 20 4 1977 -9.8 160.7 7.2 a t 21 4 1977 -10.0 160.8 7.6 a 1 21 4 1977 -10.3 160.7 6.0 a 4 4 11 1978 -11.2 162.1 6.0 f 4

23 10 1979 -10.7 161.4 7,1 7.1 4 23 10 1979 -10.7 161.3 6.0 a 4 7 2 1984 -10.0 160.5 7.5 7.7 1 8 2 1984 -9.8 160.3 6.4 a 4

LONG-TERM EARTHQUAKE PREDICT ION

Table II. Continued

29

Seismogenic source Date Epicenter M, M Ref.*

Source $6 Santa Cruz Strait

Source H1 Santa Cruz Isl.

10 8 1988 -10 .4 160.8 7.4 a 1 14 2 1989 -10.5 161.4 6.4 a 4 17 8 1990 -11 .2 162.0 6.8 a 4

3 10 1931 -11 .0 163.0 7.0 7,0 2 25 1 1937 - I0 .0 163.0 7.1 7,1 2

1 12 1970 -11.0 163.4 6.1 f 4 2 12 1970 -11 .0 163.3 7.0 7.1 4 4 8 1972 -11 .2 162.1 6.0 a 4 4 8 1972 -11.1 162.1 6.1 a 4 5 8 1972 -11.3 162.2 6.1 a 4 5 8 1972 -11 .2 162.2 6.1 a 4 6 8 1972 -11.1 162.2 6.1 a 5 4 11 1978 -11 .2 162.1 6.0 f 4 5 11 1978 -11.1 162.2 7.1 7.2 4 7 11 1978 -11.0 162.2 6.1 a 4

21 12 1978 -11.3 162.7 6.3 a 4 19 10 1982 -11 .4 163.1 6.0 a 4 23 6 1985 -11 .0 163.6 6.5 a 4 23 6 1985 -11.0 163.7 6.0 a 4 23 6 1985 -10.9 163.8 6.0 a 4 27 10 1989 -11 .0 162.4 7.0 7.0 1 6 11 1989 -11.3 162.4 6,0 a 4

29 7 1900 -10 .0 165.0 7.5 f 1 29 12 1900 -10 .0 165.0 7.6 7.8 3 13 3 1934 -11 .0 164.0 6.7 f 5 18 7 1934 -11 .7 166.5 7.9 8.0 1 19 7 1934 -12 .0 166.0 6.7 a 5 19 7 1934 -13.0 166.5 6.5 a 5 19 7 1934 -13 .0 166.0 6.9 a 5 21 7 1934 -11 .0 165.7 7.0 a 1 7 8 1934 -12.7 166.7 6.9 a 5

18 10 1934 -10 .7 165.0 6.5 a 5 17 5 1941 -10 .0 166.2 7.1 a 1 16 11 1944 -12.5 167.0 7.1 a 1 26 2 1953 -11.0 164.2 7.0 f 1 4 11 1953 -13 .0 166.5 7.4 7.8 1 4 11 1953 -13 .0 166.4 6.5 a 5

13 11 1953 -13.0 166.4 6.7 a 5 3 10 1954 -10.7 165.5 6.9 a 5 7 10 1954 -10.9 166.2 6.7 a 5

26 12 1956 -10 .0 166.0 6.0 a 5 18 5 1958 -13.1 166.5 6.3 a 5 18 5 1958 -13 .2 166.3 6.2 a 5 28 11 1959 -13.1 167.1 6.0 a 5 27 3 1960 -13 .4 165.9 6.2 a 5 27 3 1960 -13.3 165.8 6.3 a 5 28 3 1960 -13.5 165.9 6.0 a 5 30 3 1960 -13.5 165.9 6.0 a 5 15 4 1960 -13.5 166.0 6.3 a 5 20 10 1960 -11.1 164.9 6.2 a 5 22 1 1961 -12.3 166.2 6.8 a 5


Table II. Continued

Seismogenic source Date Epicenter M, M Ref. ~

Source H2 Espiritn Santo

6 11 1961 -13.3 166.5 6.0 a 5 15 9 1963 -10.4 165.6 7.2 a i 17 9 1963 -10 ,2 165.4 7.4 a 1 3 12 1963 -12 .2 166.0 6.0 a 5

22 1 1964 -13 .7 166.0 6.3 f 5 10 1 1965 -13.5 166.6 6.8 f 5 1 8 1965 -13.5 165.8 6.0 f 5

10 12 1965 -11.3 166.2 6.0 f 5 31 12 1966 -11.9 166.4 7.9 8.0 1 31 12 1966 -12.1 165.7 7.1 a 1 17 6 1968 -12.3 166.7 6.1 a 4 6 1 1969 -10 .6 164.5 6.8 a 4

16 4 1969 -13.6 166.2 6.1 a 5 25 10 1971 -13.1 166.4 6.2 a 4 28 10 1971 -13 .4 166.4 6.7 a 4 23 1 1972 -13.1 166.3 7.1 a 4 24 1 1972 -13.1 166.4 6.2 a 4 2 4 1972 -13.1 166.2 6.6 a 5

19 12 1975 -11.7 164.8 6.1 a 5 8 3 1976 -10 .6 165.1 6.3 a 5

14 4 1980 -10.9 164.2 6.1 f 4 8 7 1980 -12.4 166.4 7.3 f 1 9 7 1980 -12.7 165.8 6.7 f 4

17 7 1980 -12.5 165.9 7.7 7.8 1 2 8 1980 -11 . i 165.4 6.3 a 4

27 10 1980 -13.0 166.2 6.2 a 4 24 4 1981 -13 .4 166.4 6.9 a 4 20 2 1982 -10.9 166.0 6.8 a 4 5 8 1982 -12.6 165.9 7,1 a 1

26 5 1984 -10.8 164.2 6.2 a 4 23 i0 1986 -11.0 165.2 6.5 a 4 23 10 1986 -11.1 165.5 6.4 a 4 12 6 1988 -10.8 165.2 6.4 a 4 27 5 1992 -11.1 165.2 7.0 7.0 4

26 4 1934 -14.0 166.0 6.5 6.8 5 27 4 1940 -15.0 167.0 6.5 a 5 27 4 1940 -15.0 167.0 6.5 a 5 15 2 1955 -13.8 167.3 6.0 f 5 26 10 1956 -13.7 166.9 6.5 6.8 5 12 12 1957 -13.6 166.7 6.0 a 5 20 8 1958 -14.3 166.3 6.3 a 5 21 7 1959 -14.3 167.8 6.1 a 5 29 6 1961 -13.9 166.0 6.3 a 5 20 5 1965 -14.6 167.4 7.1 7.3 1 17 8 1965 - i5 .2 166.6 6.0 a 5 t0 12 1969 -14.9 167.0 6.3 a 4 11 8 1970 -14.1 166.6 7.0 a 4 6 1 1973 - i4 .7 166.4 6.0 f 4 9 10 1973 -14.3 167.1 6.4 f 4

28 12 1973 -14.5 166.6 7.3 7.4 i 30 12 1973 -15.4 166.5 6.6 a 4


Table II. Continued

31

Seismogenic source Date Epicenter M~ M Ref. ~

Source H3 Malekula Isl.

10 1 1974 -14.4 166.9 7.0 a i 11 1 1974 -14.2 166.5 6.2 a 4 4 9 1977 -13.7 166.7 6.5 a 4

12 5 1980 -14.5 167.9 6.1 a 4 28 11 1985 -14.0 166.2 7.0 7.2 1 7 12 1987 -13.6 167.5 6.3 a 4

19 2 1990 -15.5 166.4 6.7 a 4 14 8 1991 -13.6 167.6 6.6 a 4 15 10 1992 -14.5 166.6 6.8 a 4 4 11 1992 -14.2 167.5 6.2 a 4

17 11 1898 -16.5 168.5 7.2 7.3 3 13 5 1903 -17.0 168.0 7.0 a 3 24 1 1927 -16.5 167.5 7.1 7.1 2 24 1 1932 -17.2 167.7 6.5 a 5 20 1 1946 -17.5 167.5 6.7 6.8 5 21 7 1950 -16.5 169.0 6.5 a 5 5 1 1955 -16.3 167.1 6.9 7.1 5 5 1 1955 -16.1 167.6 6.8 a 5 6 1 1955 -16.0 167.3 6.0 a 5

21 9 1955 -17.3 168.2 6.0 f 5 11 9 1956 -17.0 169.0 6.0 f 5 22 8 1957 -15.3 167.7 6.0 f 5 24 10 1957 -14.9 167.9 6.2 f 5 28 11 1957 -14.9 167.9 6.2 f 5 31 5 1958 -15.4 168.5 7.0 7.2 1 3 6 1958 -15.6 168.2 6.5 a 5

22 10 1958 -14.9 167.8 6.3 a 5 29 3 1960 -16.9 167.2 6.7 a 5 1 9 1960 -16.5 167.6 6.0 a 5

11 12 1960 -15.7 167.0 6.2 a 5 6 i0 1962 -17.4 167.7 6.6 f 5 6 10 1962 -17.2 168.0 6.0 f 5 6 10 1962 -17.5 167.5 6.0 f 5

29 11 1962 -17.3 168.5 6.1 f 5 11 8 1965 -15.5 166.9 7.1 f 1 11 8 1965 -15.6 167.0 6.7 f 5 11 8 1965 -15.8 167.1 7.3 7.6 1 12 8 1965 -15.9 167.4 6.6 a 5 12 8 1965 -15.9 167.4 6.0 a 5 13 8 1965 -15.9 166.8 7.1 a i 13 8 1965 -16.6 167.6 6.5 a 5 18 8 1965 -16.1 166.9 6.3 a 5 29 8 1965 -15.7 167.6 6.0 a 5 27 i0 1971 -15.6 167.2 7.1 a 4 8 4 1973 -15.8 167.2 6.4 a 4 5 6 1973 -17.2 167.8 6.1 a 4

26 8 1979 -17.6 167.6 6.0 f 4 15 7 1981 -17.3 167.6 7.0 7.1 1 3 7 1985 -17.2 167.8 6.4 a 4 3 1 1987 -15.0 167.9 6.5 a 4

26 11 1987 -16.4 168.1 6.3 a 4

32

Table II. Continued

D. G. PANAGIOTOPOULOS

Seismogenic source Date Epicenter M~, M Ref,*

Source H4 Elate Isl.

Source H5 Tana Isl.

27 11 1987 -16.4 168.1 6.4 a 4 13 2 1992 -15.9 166.4 6.8 6.8 4

24 7 1957 -18.0 169.1 6.5 6.6 5 11 8 1957 -17.9 169.0 6.3 a 5 30 9 1959 -18.1 168.0 6.3 f 5 9 6 1960 -18.0 169.0 6.2 f 5

23 7 1961 -18.6 168.2 6.2 f 5 23 7 1961 -18.3 168.2 7.3 7.3 1 28 7 1961 -18.7 167.7 6.0 a 5 16 2 1966 -17.7 168.0 6.4 a 5 26 5 1974 -17.7 167.8 6,0 6.0 4 27 1 1979 -18.5 168.2 6.3 6,4 4 17 8 1979 -17.7 167.8 6.1 a 4 25 10 1986 -17.7 168.1 6.0 6.0 4 28 9 1987 -18.5 168.1 6.8 f 4 28 9 1987 -18.5 168.1 6.5 f 4 5 3 1990 -18.3 168.1 7.0 7.2 4

23 9 1990 -17.7 167.6 6.2 a 4

20 9 1920 -20.0 168.0 7.7 7.7 1 17 1 1935 -20.2 169.5 6.5 6.5 5 I3 11 1943 -19.0 170.0 7.1 7.1 1 17 5 1950 -21.0 169.0 6.6 f 5 26 5 1950 -20.2 169.2 6.9 f 5 21 6 1950 -20.2 169.2 6.8 f 5 24 6 1950 -20.5 169.5 6.8 f 5 2 12 1950 -20.0 169.5 7.2 7.5 1

26 2 1953 -19.9 169.0 6.0 a 5 3 3 1953 -20.6 t69.0 6.7 a 5

30 4 1953 -20.6 169.0 6.7 a 5 16 11 1953 -21,2 168.5 6.0 a 5 31 I0 1954 -18.9 169.9 6.1 a 5 7 3 1955 -18.9 168.8 6.0 a 5 1 2 1956 -20.0 169.2 6.3 a 5

16 8 1959 -21.0 169.1 6.3 a 5 29 7 1960 -19.5 170.5 6.5 a 5 6 9 1960 -20.4 169.4 6.2 a 5 6 7 1961 -20.6 169.4 6.7 6.8 5 8 7 1961 -20.2 168.7 6.0 a 5 8 7 1961 -20.1 168.7 6.2 a 5 9 8 1961 -19.2 168.8 6.0 a 5 9 3 1970 -19.1 168.6 6.5 f 4

10 9 1972 -20.2 168.8 6.2 f 4 13 9 1972 -20.4 168.9 6,6 f 4 27 10 1972 -19.9 168.9 6.2 f 4 2 I1 1972 -20.0 168.9 7.0 7.2 4 4 11 1972 -20.2 168.9 6.1 a 4 9 12 1973 -19.9 169.7 6.8 a 4 9 12 1973 -19.8 169.8 6.3 a 4 3 3 1974 -20.0 169.8 6.1 a 4

28 7 1976 -20.2 170.1 6.0 a 4


Table IL Continued

33

Seismogenic source Date Epicenter Ms M Ref.*

Source H 6 Matthew Isl.

Source H7 Hunter Isl,

6 9 1978 -20.2 168.8 6.0 a 4 17 4 1983 -20.7 169.2 6.2 6.2 4 17 10 1992 -19.3 169.5 6.5 6.5 4

9 8 1901 -22.0 170.0 7.8 7.8 1 16 3 1928 -22.0 170.5 7.4 7.4 1 14 3 1943 -22.0 169.5 6.8 f 5 15 3 1943 -22.0 169.5 6.9 f 5 14 9 1943 -22.0 171.0 7.2 7.4 1 19 4 1945 -21.0 169.5 6.7 a 5 11 5 1953 -21.7 169.3 6.6 a 5 2 8 1953 -21.6 169.7 6.0 6.0 5

26 11 1956 -21.6 169.1 6.6 6.6 5 5 1 1961 -21.2 169.5 6.7 6.9 5

28 1 1961 -21.3 169.5 6.2 a 5 22 12 1962 -22.0 170.1 6.5 a 5 24 10 1980 -22.0 170.1 6.7 f 4 25 10 1980 -22.1 170.1 6.7 f 4 25 10 1980 -21.9 169.9 7.0 7.3 1 25 10 1980 -22.3 170.4 6.5 a 4 29 10 1980 -21.4 169.5 6.4 a 4 19 2 1981 -21.6 169.5 6.0 a 4 29 7 1981 -21.6 169.7 6.2 a 4 6 9 1981 -21.5 169.6 6.2 a 4

17 9 1981 -22.5 170.5 6.6 a 4 16 11 1981 -22.1 169.5 6.2 a 4 24 11 1981 -22.5 170.6 6.7 a 4 24 11 1981 -22.4 170.6 6.2 a 4 21 4 1987 -22.7 170.2 6.0 a 4

27 4 1934 -22.7 171.2 6.5 6.7 5 4 11 1934 -22.0 174.0 6.5 a 5

20 4 1938 -22.0 175.0 6.5 f 5 5 7 1938 -22.5 171.5 6.7 6.8 5 6 10 1953 -23.3 170.9 6.0 6.0 5 6 7 1965 -22.5 172.9 6.0 f 5

12 9 1966 -23.0 170.6 6.1 6.3 5 23 3 1974 -21.9 173.7 6.1 6.1 4 8 3 1980 -22.7 171.4 6.7 f 4

11 4 1980 -23.0 171.2 6.2 f 4 22 2 1981 -22.1 174.8 6.2 f 4 6 7 1981 -22.3 171.7 7.0 7.1 1

17 7 1982 -21.7 173.2 6.2 a 4 23 7 1988 -22.1 174.9 6.4 f 4 7 11 1988 -22.2 175.0 6.7 f 4 3 3 1990 -22.1 175.2 7.4 7.5 4 3 3 1990 -21.6 175.8 6.2 a 4 3 3 1990 -22.4 174.2 6.9 a 4

11 12 1991 -23.4 171.0 6.5 a 4

* 1. Pacheco and Sykes (1992), 2. Abe (1981), 3. Abe and Noguchi (1983), 4. ISC, NEIC, 5. Tsapanos et al. (1990).

34

Table III. Data used to determine all the parameters of the (7)


empirical relations (6) and

Seismogenic source Mini. Mp M I T tp tf

Source S1 Bougainville Isl.

Source $2 Choiseul Isl.

Source $3 New Georgia Isl.

Source $4 Guadalcanal Isl.

Source $5 San Cristobal Isl.

6.0 6.0 7.0 4.48 1953 1957 6.0 7.0 6,7 12.79 1957 1970 6.0 6.7 6.6 13.78 1970 1984 6.6 7.3 7.6 7.00 1932 1939 6.6 7.6 7.0 18.86 1939 1957 6,6 7.0 6.7 12.79 1957 1970 6.6 6.7 6.6 13.78 1970 1984 6.7 7.3 7.6 7.00 1932 1939 6.7 7.6 7.0 18.86 1939 1957 6.7 7.0 6.7 12.79 1957 1970 7.0 7.3 7.6 7.00 1932 t939 7.0 7.6 7.0 18.86 1939 1957

6.7 7.2 6.7 10.31 1936 1946 6.7 6.7 7.4 6.32 1946 1952 6,7 7.4 7.4 21A5 1952 1974 6.7 7.4 6.7 9.70 1974 1983 7.2 7.2 7.4, 16.63 1936 1952 7.2 7.4 7.4 21,15 1952 1974 7.4 7.4 7.4 21.15 1952 1974

6.0 6.8 6.6 13.24 1963 1976 6.0 6.6 6.0 9.27 1976 1986 6.0 6.0 7.1 5.64 1986 1991 6.6 6.8 6.6 13.24 1963 1976 6.6 6.6 7.1 14,91 1976 1991 6,8 6.8 7.1 28.14 1963 1991

6.4 7.1 7.0 11.76 1939 1950 6.4 7.0 6.6 15,05 1950 1965 6.4 6.6 6.4 11.40 1965 1977 6.4 6.4 7.1 8.43 1977 1985 6.6 7.1 7.0 11.76 1939 1950 6.6 7.0 6.6 15.05 1950 1965 6.6 6.6 7.1 19.83 1965 1985 7.0 7.3 7.1 37.69 1901 1939 7.0 7,1 7.0 11.76 1939 1950 7.0 7.0 7.1 34.89 1950 1985 7.1 7.3 7.1 37.69 1901 1939

7.1 7.2 8.0 5.48 1926 1931 7.1 8.0 7.5 24.03 1931 1955 7.1 7.5 7.9 10.67 1955 1966 7.1 7.9 7.1 13.36 1966 1979 7.1 7.1 7.7 4.29 1979 1984 7.2 7.2 8.0 5.48 1926 1931 7.2 8.0 7.5 24.03 1931 1955 7.2 7.5 7.9 10,67 1955 1966 7.2 7.9 7.7 17,65 1966 1984 7.5 8.0 7.5 24,03 1931 1955 7.5 7.5 7.9 10.67 1955 1966 7.5 7.9 7.7 17.65 1966 1984 7.7 8.0 7.9 34.70 1931 1966


Table III. Continued

Seismogenic source M~n Mp Mf T tp tf

Source $6 Santa Cruz Strait

Source H1 Ndeni-Santz Cruz Islands

Source H2 Espiritu Santo Isl.

Source H3 Malekula Isl.

Source H4 Elate Isl.

Source H5 Tana Isl.

7.7 7.9 7.7 17.65 1966 1984 7.9 8.0 7.9 34.70 1931 1966

7.0 7.1 7.1 33.85 1937 1970 7.0 7.1 7.2 7.93 1970 1978 7.0 7.2 7.0 10.98 1978 1989 7.1 7.1 7.1 33.85 1973 1970 7.1 7.1 7.2 7.93 1970 1978

7.0 7.8 8.0 33.55 1900 1934 7.0 8.0 7.8 19.29 1934 1953 7.0 7.8 8.0 13.16 1953 1966 7.0 8.0 7.8 13.55 1966 1980 7.0 7.8 7.0 11.86 1980 1992 7.8 7.8 8.0 33.55 1900 1934 7.8 8.0 7.8 19.29 1934 1953 7.8 7.8 8.0 13.16 1953 1966 7.8 8.0 7.8 13.55 1966 1980 8.0 8.0 8.0 32.45 1934 1966

6.8 6.8 6.8 22.50 1934 1956 6.8 6.8 7.3 8.56 1956 1965 6.8 7.3 7.4 8.61 1965 1973 6.8 7.4 7.2 11.92 1973 1985 7.2 7.3 7.4 8.61 1965 1973 7.2 7.4 7.2 11.92 1973 1985 7.3 7.3 7.4 8.61 1965 1973

6.8 6.8 7.1 8.96 1946 1955 6.8 7.2 7.6 7.20 1958 1965 6.8 7.6 7.1 15.93 1965 1981 6.8 7.1 6.8 10.58 1981 1992 7. i 7.3 7.1 28.18 1898 1927 7. i 7.1 7.1 27.95 1927 1955 7.1 7.2 7.6 7.20 1958 1965 7.1 7.6 7.1 15.93 1965 1981 7.2 7.2 7,6 7.20 1958 1965

6.0 6.6 7,3 4.00 1957 1961 6.0 7.3 6.0 12.84 1961 1974 6.0 6.0 6.4 4.67 1974 1979 6.0 6.4 6.0 7.74 1979 1986 6.0 6.0 7.2 3.36 1986 1990 6.4 6.6 7.3 4.00 1957 1961 6.4 7.3 6,4 17.51 1961 1979 6.4 6.4 7,2 11.11 1979 1990 6.6 7.3 7,2 28.62 1961 1990 7.2 7.3 7.2 28.62 1961 1990

6.2 6.8 7.2 11.32 1961 1972 6.2 7.2 6.2 10.46 1972 1983 6.2 6.2 6.5 9.50 1983 I992 6.5 6.5 7.1 8.82 1935 1943 6.5 7.1 7.5 7.05 1943 1950 6.5 7.5 6.8 10.60 1950 1961

35

36

Table III. Continued


Seismogenic source Mini, Mp My T tp ty

Source H 6 Matthew Isl.

Source H 7 Hunter Isl.

6.5 6,8 7.2 11.32 1961 1972 6.5 7.2 6,5 19.96 1972 1992 6.8 7.1 7.5 7.05 1943 1950 6,8 7.5 6.8 10.60 t950 1961 6.8 6.8 7.2 11,32 1961 1972 7. I 7.7 7.1 23.15 1920 1943 7.1 7,1 7,5 7.05 1943 1950 7.1 7.5 7.2 21.92 1950 1972 7.2 7.7 7.5 30.20 1920 1950 7.2 7.5 7.2 21,92 1950 1972 7.5 7.7 7,5 30.20 1920 1950

6.0 6.0 6.6 3,32 1953 1956 6.0 6.6 6.9 4.11 1956 1961 6.6 7.4 6.6 13.20 1943 1956 6.6 6.6 6.9 4.11 1956 1961 6.6 6.9 7.3 19.80 1961 1980 6.9 7.4 6.9 17.31 1943 1961 6.9 6.9 7.3 19.80 1961 1980 7.3 7.8 7.4 26.60 1901 1928 7.3 7.4 7.4 15.49 1928 1943 7.3 7.4 7.3 37.11 1943 1980 7.4 7.8 7.4 26.60 I901 1928 7.4 7.4 7.4 15.49 1928 1943

6.0 6.3 6.1 7.53 1966 1974 6.0 6.1 7.1 7.28 1974 1981 6.0 7.1 7.5 8.66 1981 1990 6.1 6.3 6.1 7.53 1966 1974 6.1 6.1 7.1 7.28 1974 1981 6.1 7.1 7.5 8,66 1981 1990 6.3 6.7 6.8 4.t9 1934 1938 6.3 6.3 7.1 14.82 1966 1981 6.3 7,1 7.5 8.66 1981 1990 6.7 6.7 6.8 4,19 1934 1938 6.7 7.1 7.5 8,66 1981 1990 6.8 7.1 7.5 8.66 1981 1990 7.1 7,1 7.5 8.66 1981 1990

4. Empir ical Relations

The data for the 13 seismogenic sources were used in order to estimate the parameters of relation (1). A total number of 134 observations (T, Mmin, Mp, 21)/o) have been used to determine the relation

log T, = 0.17Mmin + 0.31Mp - 0.331 log 3;/o + 6.36 (6)

with a goodness of fit equal to 0.69 and a standard deviation equal to 0.20, while the corresponding standard deviations for the calculated parameters are b = -+0.05, c = -+0.30, d = -+0.08 and t -- -+0.33. The strong correlation between the

LONG-TERM EARTHQUAKE PREDICT ION 37

repeat time and the magnitude Mp indicates that the time-predictable model holds very well. The values of the parameters (b = 0.17, c = 0.31, d = -0.33 and t-- 6.36) in the relation, (6), are in fairly good agreement with the results in the Aegean area (b = 0.24, c = 0.25, d =-0 .36 and t = 7.36) as it was found by Papazachos and Papaioannou (1993). It indicates the validity of this model for several regimes of the world. In a similar way the parameters of relation (2) were determined and the following empirical formula was obtained

My = 0 .51Mmin - 0 ,21/ l / /p + 0.54 log 3)/0 - 9.44 (7)

with a goodness of fit equal to 0.72 and standard deviation equal to 0.32. The standard deviations of the parameters are B = +0.07, C = -+0.50, D = ---0.14 and m = -+0.50. The meaning of the negative value of C (-0.12) is that large mainshocks are followed by small ones and vice versa.

5. Long Term Earthquake Prediction

Papazachos (1988b, 1991, 1993) and Papazachos and Papaioannou (1993) have proposed that the lognormat distribution of the ratio T/Tt, where T is the observed interevent time between successive mainshocks in a certain seismogenic source and Tt is the corresponding theoretical value given by relation (1), provides a better fit than the Gaussian and Weibull distributions. This is in accordance with a similar study by Nishenko and Buland (1987).

In order to apply a distribution on the data the Kolmogorov test was made for the validity of lognormal, Weibull and normal distributions. For the lognormal distribution of T/Tt the significance level of the Kolmogorov test was found equal to 0.83, for the Weibull distribution the significance level of this test is equal to 0.41 and for normal distribution the significance level is equal to 0.13. Taking into account the results of these statistical tests the use of the lognormal distribution on the data is justified.

Figure 4 displays the frequency histogram of log(T/Tt) and the theoretical normal distribution which has a mean equal to zero and standard deviation equal to 0.20. Figure 5 displays the frequency histogram of the difference (MF- Mr) between the observed magnitude, MF, of the following mainshock and the calculated magnitude, Mr, by relation (7) and the theoretical normal distribution which has a mean equal to zero and standard deviation equal to 0.32. Assuming that the lognormal distribution holds for the area under study and taking into account the time of occurrence and the magnitude of the last mainshock, the probabilities of occurrence during the next decade (1993-2002) for earthquakes with magnitudes equal to or larger than 7.0 were calculated.

Table IV gives information on the expected large (Ms/> 7.0) shallow mainshocks based on the model expressed by relations (6) and (7). The first two columns give the code number and the name of the seismogenic sources shown in Figure 3. The last two columns of this table give the probability values, Plo for the occurrence

38

Fig. 4.

30

25

~20 0 C

J 15 D" W L U. 10

B

0


Frequency Histogram

. . . . I . . . . I . . . . I . . . . I . . . . I - -

-1 .5 - -B .B 0 O.B i i .B LOG (T /Tt )

The frequency distribution of the observed repeat times compared to the theoretical one.

Frequency H is togram

20

31

W 3 O"

e L

4

I . . . . I I I I I I

-1 .5 -I -0 .5 0 0 .5 i 1.5 MF-M~

The frequency distribution of the difference M~ - Mf between the observed MF and the Fig. 5. calculated My of the following mainshocks by relation (6).

1B


Table IV. Information on the expected shallow earthquakes with magnitudes, My, and the corresponding probabilities, P10, for the occurrence of very large or great (M~/> 7.0) ones during the period 1993-2002 in the area of Solomon Islands and New Hebrides

Seismogenic source Ms >/7.0

Mf Pao

Source $1 Bougainville Isl. 7.4 0.76 Source $2 Choiseul Isl. 7.3 0.64 Source $3 New Georgia Isl. 7.1 0.16 Source $4 Guadalcanal Isl. 7.2 0.50 Source $5 San Cristobal Isl. 7.6 0.69 Source $6 Santa Cruz Strait 7.2 0.35 Source H1 Santa Cruz-Ndeni Isl. 7.7 0.79 Source H2 Espiritu Santo Isl. 7.3 0.55 Source H3 Malekula Isl. 7.4 0.56 Source H4 Elate Isl. 7.1 0.19 Source H5 Tanas Isl. 7.5 0.79 Source H6 Matthew Isl. 7.4 0.76 Source H7 Hunter Isl. 7.2 0.15

39

of large (Mmjn I> 7.0) shallow earthquakes during the next decade (1993-2002) with the corresponding magnitudes Mf of the expected mainshocks as these magnitudes were calculated by relation (7). It must be noted that the absolute probability values are of relative importance because there is a possibility to have changes to these values if a larger sample of data is used.

The seismogenic sources of Santa Cruz-Ndeni Islands (H~) and Tana Island (Hs) exhibit the highest probability value (P10 = 0.79) among all the seismogenic sources, for the occurrence of a mainshock with Ms >i 7.0. Very high probability values (Pw 1> 0.75) for the occurrence of a mainshock with Ms i> 7.0 were calculated for the seismogenic sources of Bougainville Island (Sa) and Matthew Islands (H6). High probability values (0.75 > Plo i> 0.55) were calculated for the seismogenic sources of Choiseul Island ($2), San Cristobal Island ($5), Espiritu Santo Island (H2) and Malekula Island (H3). Intermediate probability values (0.55 > P~0 ~> 0.30) were calculated for the seismogenic sources of Guadalcanal Island ($4) and Santa Cruz Strait ($6). Low probability values (Plo < 0.30) for Ms >/7.0 were calculated for the seismogenic sources of New Georgia Island ($3), Elate Island (84) and Hunter Islands (H7).

6. Discussion

The data used in the present study fit fairly well the time and magnitude predictable model. The validity of that model was also checked in many other areas (Papaz- achos and Papaioannou, 1993; Papazachos et al., 1993; Papadimitriou, 1993,


1994; Panagiotopoulos, 1993; Karakaisis 1994a,b and Tsapanos and Papazachos, 1993).

The physical explanation of the empirical relations which are used to calculate recurrence times of large earthquakes, such as relation (6), is not an easy problem. The first term of this relation, bMmin, expresses the recurrence law of Gutenberg and Richter (1944) for the case of the mainshocks and depends on the magnitude of the smallest mainshock considered. This is due to the fact that the applied model does not consider a single fault where the characteristic earthquake occurs, but a seismogenic source which includes faults of different sizes. The second term, CMp, represents the time predictable model, which is strongly supported by recent research (Bufe et al., 1977; Shimazaki and Nakata, 1980; Papazachos, 1989; Nish- enko, 1991).

It is necessary, however, to take into account the fact that there are other factors which also affected the time and magnitude of the next earthquake in a seismogenic source which are not considered in this model. Such factors are the interaction between adjacent faults (Scholz, 1988; Cornell et al., 1993) and the long term clustering of earthquakes (Ambraseys, 1971; Ambraseys and Melville, 1982). It is very difficult to quantify these factors (Johnston and Nava, 1985; Zheng and Vere-Jones, 1991) and for this reason the time- and magnitude- predict- able model must be considered as a satisfactory one for purposes of statistical long term prediction at the moment.

In order to check the effect of the zonation on the results of this paper, the area covered by the sources H1, H2, H3 and H4 (Figure 3), where the zonation is more problematic, was separated in seven sources on the basis of bathymetric features and the calculations were repeated. The calculated values of the parameters b, c, d and t of the relation, (1), for the sixteen sources were equal to 0.16, 0.31, -0.36 and 7.23, respectively, that is, very close to the values of these parameters calculated for the 13 seismogenic sources. This indicates that the accurate definition of the seismogenic sources is important but not very critical for the results of this method.

On the basis of the seismic gap concept an attempt was made for long term earthquake prediction by Nishenko (1991) for the Solomon Islands and New Hebrides. The probability values, which were estimated by Nishenko (199I), for some segments along the Solomon Islands and New Hebrides, are in good agreement with the estimated probability values for the occurrence of a great earthquake with Ms I> 7.0 in the present study. Thus Nishenko proposed intermediate probability values for the regions of Guadalcanal Island (45%), San Cristobal Island (45%) and Vanikoro Island (48%) which correspond quite well with the seismogenic sources of Guadalcanal Island (source $4, with P10 = 0.50), San Cristobal Island (source $5, with PiG = 0.69) and Santa Cruz-Ndeni Island (source H~, with PiG = 0.79), in the present study. The aforementioned author estimated high probability value at the region of South Santo Island (60%) which correspond


quite well with the seismogenic source of Espiritu Santo Island (source H2, with Plo = 0.55) in the present study.

It is interesting to note that the occurrence of a large earthquake at the seismogenic source of Santa Cruz-Ndeni Islands (source H1, Figures 2, 3) on 6 March 1993 with Ms = 7.2 (10.95 S, 164.20 E, USGS) is in good agreement with the result of the present study. The probability value which was estimated on the basis of the model expressed by relations (6) and (7) for the Santa Cruz-Ndeni Island (source H1) for the occurrence of an earthquake with Ms i> 7.0 was very high (Pw = 0.79) with expected magnitude Ms = 7.7.

Acknowledgements

The author would like to express his sincere appreciation to Prof. B. C. Papazachos for his creative idea, his suggestions and the critical reading of the manuscript and to the two anonymous reviewers for their comments that improved this work. In addition thanks are also given to E. Scordilis, C. Papazachos and Ch. Papaioannou for kindly providing the computer programs which were necessary to complete the data processing.

References

Abe, K.: 1981, Magnitudes of large earthquakes from 1904 to 1980, Phys. Earth Planet. Interiors 27, 72-93.

Abe, K. and Noguchi, S.: 1983, Revision of magnitudes of large shallow earthquakes, 1897-1912, Phys. Earth Planet. Interiors 33, 1-11.

Ambraseys, N. N.: 1971, Value of historical records of earthquakes, Nature 232, 375-379. Ambraseys, N. N. and Melville, C. P.: 1982, A History of Persian Earthquakes, Cambridge University

Press, Cambridge,219 pp. Astiz, L. and Kanamori, H.: 1984, An earthquake doublet in Ometepee, Guerrero, Mexico, Phys.

Earth Planet. Interiors 34, 24-45. Bufe, C. G., Harsh, P. W., and Burford, R. O.: 1977, Steady-state seismic slip: A precise recurrence

model, Geophys. Res. Lett. 4, 91-94. Chase, C. G.: 1978, Plate Kinematics: the Americas, East Africa and the rest of the world, Earth

Planet. Sci. Lett. 37, 355-368. Cole, J. W.: 1984, Taupo-Rotorua Depression: an ensialic marginal basin of North Island, New

Zealand, in B. P. Kokelaar and M. F. Howells (eds), In Marginal Basin Geology, Blackwell Scientific Publication, Spee. Publ. 16, pp. 109-120.

Cooper, P. and Taylor, B.: 1987, Seismoteetouics of New Guinea: a model for arc reversal following arc-continent collision, Tectonics 60(1), 53-67.

Cornell, C. A., Wu, S., Winterstein, S. T., Dieterich, J. H., and Simpson, R. W.: 1993, Seismic hazard induced by mechanically interactive fault segments, Bull. Seism. Soc. Am. 83(2), 436-449.

Dahle, A., Bungum, H., and Kramme, L.: 1990, Attenuation models inferred from intraplate earthquake recordings, Earthq. Eng. Struct. Dynam. 19, 1125-1141.

Davey, F. J.: 1982, The structure of the South Fiji Basin, Tectonophysics 87, 185-241. Draper, N. R. and Smith, H.: 1966, Applied Regression Analysis, Wiley Publications, New York, 407

PP. Gutenberg, B. and Richter, C. F.: 1944, Frequency of earthquakes in California, Bull. Seism. Soc.

Am. 34, 185-188.


Iida, K., Cox, D., and Pararas-Carayannis, G. : 1967, Preliminary Catalog of Tsunamis Occurrring in the Pacific Ocean, Hawaii Inst. Geophys, Univ. Hawaii, 67-10, 131 pp.

Isacks, B. L., Gardwell, R. K., Chatelain, J. L., Barazangi, M., Marthetot, J. M., Chinn, D., and Louat, R.: 1981, Seismicity and tectonics of the central New Hebrides Island arc, in D. W. Simpson and P. G. Richards (eds), Earthquake Prediction, An Intern. Review, vol. 4, Maurice Ewing Series, Amer. Geophys. Union, pp. 93-116.

Johnston, A. C. and Nava, S.: 1985, Recurrence rates and probability estimates for the New Madrid seismic zone, J. Geophys. Res. 90(B8), 6737-6753.

Joyner, W. B. and Boore, D. M.: 1981, Peak horizontal acceleration and velocity from strong motion records including records from the 1979 Imperial Valley, California earthquake, Bull. Seism. Soc. Am. 71, 2011-2038.

Kanarnori, H.: 1977, The energy released in great earthquakes, J. Geophys. Res. 82, 2981-2987. Karakaisis, G. F.: 1994a, Long term earthquake prediction in Iran based on the time and magnitude

predictable model, Phys. Earth Planet. Interiors 83, 129-145. Karakaisis, G. F.: 1994b, Long term earthquake prediction along the North and the East Anatolian

fault zones based on the time and magnitude predictable model, J. Intern. Geophys. 116, 198-204. Karakaisis, G. F., Kourouzidis, M. C., and Papazachos, B. C.: 1991, Behaviour of seismic activity

during a single seismic cycle, Proc. Intern. Conf. Earthquake Prediction: State-@the-Art, Strasbourg, France, 15-18 Oct. 1991, 47-54.

Karig, D. E.: 1970, Ridges and basins of the Tonga-Kermadec Island arc system, J. Geophys. Res. 75, 239-254.

Karig, D. E.: 1971, Origin and development of marginal basins in the western Pacific, J. Geophys. Res. 76, 2542-2561.

Karig, D. E. and Mammerickx, J.: 1972, Tectonic framework of the New Hebrides Island arc, Marine Geol. 12, 187-205.

Kelleher, J., Savino, J., Rowlett, H., and McCann, W.: 1974, Why and where great thrust earthquakes occur along islands arcs, J. Geophys. Res. 79, 4889-4899.

Lay, T. and Kanamori, H.: 1981, An asperity model of large earthquake sequences, in D. W. Simpson and P. G. Richards (eds), Earthquake Prediction, An International Review, vol. 4, Maurice Ewing Series, Amer. Geophys. Union 1981, pp. 579-592.

McCann, W.R. 1980, Large- and moderate-sized earthquakes: Their relationship to the Tectonics of subduction, PhD Thesis, Columbia Univ., New York, pp. 194.

McCann, W. R., Nishenko, S. P., Sykes, L. R., and Kranse, J.: 1979, Seismic gaps and plate tectonics: Seismic potential for major boundaries, Pageoph 117, 1082-1147.

McGuire, R.: 1978, Seismic ground motion parameter relation, Proc. ASCE J. Geotech. Eng. Div. 104, 481-490.

Mitchel, A. H. G. and Warden, A. J.: 1971, Geological evolution of the New Hebrides Island arc, J. Geol. Soc. London 127, 501-529.

Mogi, K.: 1985, Earthquake Prediction, Academic Press. Molnar, P.: 1979, Earthquake recurrence intervals and plate tectonics, Bull. Seism. Soc. Am. 69, 115-

133. Neef, G.: 1982, Plate tectonic significance of Late Oligocene/Early Miocene deep sea sedimentation

at Maewo, Vanuatu (New Hebrides), Tectonophysics 87, 177-183. Nishenko, S. P.: 1991, Circum-Pacific Seismic Potential: 1989-1999, Pageoph 135(2), 169-259. Nishenko, S. P., and Buland, R.: 1987, A generic recurrence interval distribution for earthquake

forecasting, Bull. Seism. Soc. Am. 77, 1382-1399. Oliver, J. and Isacks, B.: 1967, Deep earthquakes zones, anomalous structure in the upper mantle

and lithosphere, J. Geophys. Res. 72, 4259-4275. Pacheco, J. F., and Sykes, L. R.: 1992, Seismic moment catalog of large shallow earthquakes, 1900

to 1989, Bull. Seism. Soc. Am. 82, 1306-1349. Packham, G. H.: 1982, Foreword to papers on the tectonics of the Southwest Pacific region, Tectono-

physics 87, 1-10. Panagiotopoulos, D. G.: 1993, Application of the time and magnitude predictable model in the

Philippines area, Proc. of the 2nd Congr. Hellenic Geophys. Union, Ftorina, Greece, 5-7 May, 1993, 472-481.


Papadimitriou, E. E.: 1993, Long term earthquake prediction along the western coast of south and central America based on a time predictable model, Pageoph. 140, 301-316.

Papadimitriou, E. E.: 1994, Long term prediction in North Pacific seismic zone based on the time and magnitude predictable model, NaturalHazards 9, 303-321.

Papazachos, B. C.: 1988a, Seismic hazard and long-term earthquake prediction in Greece, European School of Earthquake Sciences, Course on Earthquake Hazard Assessment, Athens 9-16 May 1988, pp. 1-10.

Papazachos, B. C.: 1988b, Long-term prediction of earthquakes in seismogenic sources of Greece, United Nations Seminar on the Prediction of Earthquakes, Lisbon, Portugal, 14-18 November 1988, pp. 77-84.

Papazachos, B. C.: 1989, A time-predictable model for earthquakes in Greece, Bull. Seism. Soe. Am. 79, 77-84.

Papazachos, B. C.: 1991, Long-term earthquake prediction in south Balkan region based on a time predictable model, First General Conf. Balkan Phys. Union, Thessaloniki, 26-28 Sept. 1991, pp. 1- 7.

Papazachos, B. C.: 1992, A time and magnitude predictable model for generation of shallow earthquakes in the Aegean area, Pageoph. 138, 287-308.

Papazachos, B. C.: 1993, Long-term prediction of intermediate depth earthquakes in southern Aegean region based on a time-predictable model, Natural Hazards 7, 211-218.

Papazachos, B. C., Papadimitriou, E. E., Karakaisis, G. F. and Tsapanos, T. M.: 1993, An application of the time and magnitude predictable model for the long term prediction of strong shallow earthquakes in Japan area, Bull. Seism. Soc. Am. 84, 426-437.

Papazachos, B. C., and Papaioannou, Ch. A.: 1993, Long-term earthquake prediction in the Aegean area based on a time and magnitude predictable model, Pageoph 140, 593-612.

Ripper, I. D.: 1982, Seismicity of the Indo-Australian/Solomon sea plate boundary in the southeast Papua region, Tectonophysics 87, 355-369.

Scholz, C. H.: 1988, Mechanisms of seismic quiescences, Pageoph. 126, 589-617. Shimazaki, K. and Nakata, T.: 1980, Time-predictable recurrence model for large earthquake, Geo-

phys. Res. Lett. 7, 279-282. Sykes, L. R. and Quitmeyer, R. C.: 1981, Repeat times of great earthquakes along simple plate

boundaries, in D. W. Simpson and P. G. Richards (eds), Earthquake Prediction, An International Review, vol. 4, Maurice Ewing Series, Amer. Geophys. Union, pp. 217-247.

Tajima, F., Ruff, L. J., Kanamori, H., Zhang, J., and Mogi, K.: 1990, Earthquake source processes and subduction regime in the Santa Cruz Islands region, Physics Earth Planet. Interiors. 61, 269- 290.

Tsapanos, T. M. and Papazachos, B. C.: 1993, Long term earthquake prediction in China, based on the time and magnitude predictable model, J. Earthquake Prediction Res. 3(2), 237-256.

Tsapanos, T. M., Scordilis, E. M., and Papazachos, B. C.: 1990, A global catalogue of strong earthquakes, Publ. Geophys. Lab. Univ. Thessaloniki, 9, 90 pp.

Vidale, J. and Kanamori. H.: 1983, The October 1980 earthquake sequence near the New Hebrides, Geophys. Res. Lett. 10, 1137-1140.

Walcott, R. I. : 1978, Present tectonics and Late Cenozoic evolution of New Zealand, Geophys. J. R. astr. Soc. 52, 137-164.

Weisberg, S.: 1980, Applied Linear Regression, Wiley, New York, 283 pp. Wellman, P. and McCracken, H.M.: 1979, BMR Earth Sclence Atlas: Plate Tectonics of the Australian

Region, Scale 1:20.000.000, Bureau of Mineral Research, Camberra. Wesnousky, S. G., Scholz, C. H., Shimazaki, K., and Matsuda, T.: 1984, Integration of geological

and seismological data for analysis of Seismic Hazard: A case study of Japan, Bull. Seism. Soc. Am. 74, 687-708.

Zheng, X. and Vere-Jones, D.: 1991, Application of stress release models to historical earthquakes from North China, Pageoph 135(4), 559-576.

Long-term Earthquake Prediction Along the Seismic Zone of the Solomon Islands and New Hebrides Based...

Documents

Transcript of Long-term Earthquake Prediction Along the Seismic Zone of the Solomon Islands and New Hebrides Based...