Earthquakes in the West Indian region

At the 1939 meeting of the Eastern Section of the Seismological Society of America, the author presented a paper on earthquakes occurring about 15° to 30° ...
276KB taille 26 téléchargements 340 vues
REPORTS AND PAPERS, SEISMOLOGY--1940

229

EARTHQUAKES IN THE WEST INDIAN REGION Daniel Linehan At the 1939 meeting of the Eastern Section of the Seismological Society of America, the author presented a paper on earthquakes occurring about 15° to 30° distance from Weston, Massa­ chusetts. The difficulty of locating these quakes was stressed in the paper, due to the strange characteristics of the recordings. Observers of many stations thought of them as a series of "locals". Using the records from Williamstown, Harvard, Ottawa, and Fordham along with those of Weston, the author located these disturbances just north of the West Indies. He afterwards learned that one or two of them had been felt at Puerto Rico. Earthquake-history of the area--Seismic activity in the West Indies has been well known

Fig. 1—West Indian Area; shaded portion shows area where majority of epicenters mentioned were located even prior to any instrumental investigation. Even today many shocks are of sufficient strength to be felt. The number of shocks has not been large and does not compare with the number oc­ curring in the East Indies. Our reason for making this comparison comes from the many apparent tectonic similarities existing in both places, especially the design of the negative gravityanomaly arcs. A comparison of the seismic activity from 1918 to 1930 demonstrates that the major activity in the East Indies is about 12 times greater than in the West Indies, and we may imagine that minor activity would hold about the same proportion. Importance of seismic studies in West Indies--Theories have been proposed concerning the tectonics producing the great gravity-arc that encircles the Caribbean Sea. It seems likely to many that mechanical motion of the crust must be involved, and If this is so we should expect seismic effects of this motion. The only sort of flow without seismic motion would be a liquid flow and this is most unlikely in the case of the Earth's crust. Some have stated there is no evidence of seismic motion from this area near the gravity-arc. We feel, however, that this lack of evidence has been due entirely to the inadequacy of the seismometers in use. As will be stated below, instruments of high sensitivity and short period Cone second) are necessary. Current seismic activity--For a three-year period, 1937 to 1939, we have instrumental evi­ dence of about 50 quakes happening near the West Indies. Unfortunately, the station-records we have studied are all from New England and Canada, and as the azimuths of these stations from the region under study are about the same there is lack of good control over the arc-distance inter­ sections. Even a preliminary examination, however, shows that the great majority are just slightly north of the gravity-arc and southeast of the Bahama Islands. None to date has been located east of the Lesser Antilles, although locations of the International Seismological Sum­ mary for the period from 1918 to 1930 place four out of the 12 West Indian quakes there, while only two have been located north of the arc. One recently located offered good control frQm the readings of the station at -Kingston, Jamaica. This again was north of the gravity-arc. The hope of more accurate readings for the future is increased now that the station at Kingston has begun operation.

230

TRANSACTIONS, AMERICAN GEOPHYSICAL UNION

Characteristics of the seismograms--As has been mentioned, the records from this area are quite different from those we have recorded from other locations. The predominant characteristic is the multiplicity of phases and their extremely short period. The Earth-period recorded at New England from these quakes is rarely more than one second, which seems a little odd for quakes of this distance. If we keep the same distance and change the azimuth for quakes from Mexico, the record takes on a more normal appearance. Whether this difference is due to the mechanics at the source, the geological location of the focus, or the wave-path followed, we can not say. The so-called P-group may last as long as three or four minutes due to the multiplicity mentioned and frequently runs into the S-group. The second group lasts about the same length of time as the former. About 20 minutes after the P-group is a third unidentified group which we have labelled as T. These are definitely transverse in character and also short-period. In the

MARCH, 7, 19 39 22 HR. £ (IK.

I

DECEMBER. 1,1*38 23

HR.

9

3

10

IB

SCALE 10 V

OF SECONDS 80 H 40

4(

W

U

(0

Fig. 2—P-groups of three different quakes recorded on short-period Benioffs at Weston L

"P'

10

&0

SCALE OF SECONDS SO 40

GROUP

$0

Fig. 3—P-group for quake of March 5, 1939, recorded at Weston on short-period Benioffs at 15* 1

REPORTS AND PAPERS, SEISMOLOGY--1940

231

S AUGUST

-V—v-

MARCH

2 2 , 1937

5. 1939

SEPTEMBER

1,1938

SEPTEMBER

11,1938

NOVEMBER

MARCH

10. 1 9 3 6

I I , 1938

JANUARY

MARCH

2 9 , 1939

7, 1939

DECEMBER

APRIL

2 3 , 1939

APRIL

17,1939

SEPTEMBER

DECEMBER

-v— _§

SCALE

-fr*—

1,1938

6, 1937

1 2 , 1936

OF MINUTES

16

17

19

20

21

22

23

24

Fig. 4—Diagrammatic arrangement of the P-, S-, and T-groups according to Increase in (S-P)-interval stronger quakes there is a fourth group, also transverse, that follow about four minutes after the third group. In the 50 odd quakes studied there has been only one quake that demonstrated surface-waves. The P-group from three quakes recorded by the vertical short-period component at Weston are placed together in Figure 2. Each represents a different distance but the characteristics are similar. The first one of March 5, 1939, Is 17?2 distance, the second of March 17, 1939, is 21?3 distance, and the third of December 1, 1938, is 23?5 distance. Note also on the third quake that it is led by an impulse about 35 seconds before the first group. This impulse belongs to an entirely different quake of some 80° distance. At first sight one might think that the instrument recording these quakes was underdamped, but actually the damping is critical at all times. Figure 3 demonstrates the three components of the P-group for the quake of March 5, 1939. During the first nine seconds the records show that the vertical and the north-south components are identical and show clearly the epicenter to the south. The amplitudes are approximately the same, although the actual instrumental magnification of the horizontal component is weaker, demonstrating that the angle of emergence is small. The group on the east-west component is about six seconds late, as would be expected in the present azimuth. Throughout the recording of the group this component is always the weakest. Although an illustration of the S-group is not presented here, an examination of the orig­ inal grams shows definite polarization. Again the azimuth is well demonstrated by the late ar­ rival of this phase on the north-south component with respect to the east-west. The transverse character of the wave is also emphasized. The sustained greater amplitude on the transverse component is also brought out. This latter characteristic is always present. Figure 4 shows a diagrammatic arrangement of the P-, S-, and the T-groups. Approximately ten minutes are omitted by the break in the diagram. The quakes are arranged according to the Increase in the (S-P)-interval. They do not appear to be a function of this interval. If there Is any functional quality, it appears to be inverse, such as many of the core-waves whether re­ flected or refracted, but the discontinuity of this function as appears about half-way down Figure 4 is quite amazing. In a recent private communication mith the Honorable R. C. Hayes of the Dominion Observatory, Wellington, New Zealand, he states his belief that these quakes are of deep focus and that the .third group are of the ScS-type, and the fourth group mentioned above would be sScS, or the former reflected from the Earth's crust before being reflected from the core. Unfortunately, I could not give M m all the other data, such as low angle of emergence, etc., before he made tills statement. I feel that such would eliminate their being of deep focus. In Bulletin No. 107 of

232

TRANSACTIONS, AMERICAN GEOPHYSICAL UNION

the Dominion Observatory he has determined the depths of several quakes using the (sScS-ScS)interval. The azimuth of his station with regard to the quakes he studied is identical with the azimuth of Weston to the West Indian Region, and many similarities exist in the records. How­ ever, this third or T-group arrives too late to be classified as ScS, and if we used the interval mentioned for depth we would find many of the depths very great. For example, the quake of June 12, 1939, with an interval for these two groups amounting to 240 seconds would give a depth of about 500 km. One reason why it should not be considered deep is because none of the many sta­ tions which reported it--we have the readings from over 40--even up to distances of 154° refers to any deep-focus characteristics. Most of the travel-time tables examined are bare of any phases occurring at this instant and it is probable that we shall have to deal with a new phase before the studies are completed. As was mentioned above, only one of the quakes studied presented surface-waves, namely, the quake of June 12, 1939. This one did present such waves with periods up to 25 seconds, but the amplitude of these was by no means what one would expect from the energy shown in the preliminary phases. The half amplitude of the P-phases in this quake were over 50 millimeters on the Weston records, while the surface-waves were less than half that amount. Quake of June 12. 1939--It seems quite necessary that a set of travel-times should be con­ structed for this area and the quake of June 12, 1939, offers fine opportunity for this work. The San Juan station is perhaps closest to the epicenter with a distance of 270 km and we have other fine readings of P- and P up to 154°. Many of the station-records have already been gathered by the author, but because of other pressing duties, this study has not been completed. 1

Summary--It is the hope of the author that other stations with short-period instruments of high magnification may be established nearer to the West Indian Region than the present group at New England, to make a better location of the very many earthquakes that are taking place there. He feels it is important in the light of recent interest in tectonics and believes a study of the possible movements along the gravity-arc near the West Indies will offer great aid to this study. Weston College, Weston, Massachusetts

THE VIEQUES, PUERTO RICO, SEISMOGRAPHIC RECORD OF THE DESTRUCTIVE EARTHQUAKE OF 1918 AS A CLUE TO LOCAL CRUSTAL STRUCTURE Frank Neumann (Abstract) An (S-P)-interval of 17 seconds is recorded for the main shock and a numoer of aftershocks. According to recent ideas concerning travel-times in continental structure, this would correspond to a distance of 135 to 160 km depending upon focal depth. But we know from Reid and Taber's very thorough investigation that the epicentral distance from Vieques must be 195 km and possibly more. This calls for velocities of P and S which would exist only if a very large part of the continental structure were missing, that is, a structure somewhat like the Pacific A feature of the records is that the surface-waves have the (slow) velocity to be expected in granite and are quite strong, showing that while a layer of granite is probably present, It is too thin to appreciably affect the travel-times of the interior waves as it does in normal continental struc­ ture. U. S. Coast and Geodetic Survey, Washington, D. C.

SOME POINTS OF GEOPHYSICAL INTEREST REVEALED IN THE MASS-ANALYSIS OF TELESEISMIC RECORDS R. R. Bodle Mass-analysis of seismograms may at first glance sound uninteresting. As a matter of fact to the casual observer of the subject this is no doubt quite true. When one is in continuous contact with the interpretation of seismograms the reverse is very apt to be true. The picture that one gets under these circumstances might well be compared to the old picture-puzzle with