Click for next page ( 95


The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 94
STORM AND SURF MICROSEISMS F. W. van Straten Office of the Chief of Naval Operations The practical meteorologist is interested in the so-called microseismic phenomenon be- cause it may provide a potential for giving early warning of the existence of large de- structive storms and because it may permit a type of direction-finding which will track such storms once they have formed. If microseismic techniques can realize one or both of these potentialities, the meteorologist will have a means of coping with the apparent unpredictable nature of typhoons and hurri- canes and will thereby provide the storm in- formation necessary to reduce hazard and dam- age to a minimum. Storm information at the present time is obtained only at a great cost and considerable risk. Meteorological art is advancing to the point of permitting the fore- casting of typhoon-prone conditions. But not all such conditions develop into typhoons. Conversely, an occasional typhoon develops in an unclassical situation which gives no clue as to an incipient storm development. The track which a storm will take, once developed, is also still shrouded in mystery and the most effective method of tracking still remains the purely visual one employing aircraft or radar to pene- trate the eye of the storm, or both. As scientists, meteorologists are interested in the theory of microseismic generation and propagation. That interest, however, is a "pure" interest to be contrasted with their interest in what microseisms can do for mete- orologists. Theirs is essentially the pragmatic approach. This paper deals entirely with the prag- matic approach and while, necessarily, the the- ory of microseismic generation must be touched on its consideration is limited to how the vari- ous theories affect the potential usefulness of microseisms as a meteorological tool. The history of microseismic research in this country and abroad has been so thoroughly covered that any repetition is sheer redun- dancy. Let it suffice to say that during the middle years of World War II, the Navy estab- lished a microseismic network to exploit the possibilities of storm detection and tracking in the Caribbean. In so doing, the Navy was accepting the theory that microseisms origi- nated within the storm and that they proceeded instrument-ward not through water but through the crust of the earth at the ocean bottom. The Microseismic Research Program proceeded through the years with just enough success to warrant its continuation but without sufficiently clear-cut results to establish it on a firm operational basis. Under Mr. Gilmore's direction, the network in both the Atlantic and the Pacific has been expanded and results for numerous storms have been tabulated, studied and published. As a separate endeavor, the Naval Re- search Laboratory instituted a program to improve the instrumentation used in microseis- mic research. Under the direction of Drs. Rammer and Dinger a method was devised which much simplified the methods of calibra- tion and interpretation of microseismic records. In testing out their equipment, the Naval Re- search Laboratory scientists also recorded the progress of storms in the Western Atlantic and unlike Mr. Gilmore, reached the conclusion that in these cases coastal action was responsible for the tremors which affected their microseis- mic installations. As far as the operating forces of the Navy are concerned, the question of the value of microseismic research was thrown wide-open. If microseisms originate within or near a storm, the possibility of early warning and tracking remains real. If microseisms are the result of a local coastal effect, an observer on the coast watching the incoming surf might prove a reasonable substitute for a microseis- mic network. At best, if the latter theory is the correct one, an oceanographic tool was be- ing developed. Many published papers were studied in an effort to decide between storm and surf micro- seisms. Before lining up the evidence on each side, it might be well to define in what sense the words "storm" and "surf" are being used here. By storm microseisms, I mean a crustal distur- bance which is produced in the vicinity of the storm, transmitted downward through the water to the ocean bottom and then transmitted through solid matter to the land block on which the seismometer rests. The definition of surf microseisms is somewhat less clear-cut. In the * In view of the limited scope of this paper, no attempt has been made to cite fundamental or supporting in- vestigations. 94

OCR for page 94
STORM AND SURF MICROSEISMS 95 vicinity of the storm, an oceanographic distur- bance is established which radiates through the ocean in the form of ocean swell until it reaches a continental boundary—be it continental shelf or actual coast. There, the disturbance is transmitted to the solid continental block on which is erected the microseismic recorder. It should be noted that this definition of surf microseisms in no way adopts the usual defini- tion of surf—the breaking of the sea against the coast. The mechanism by which the micro- seisms are generated is undefined in both cases. The distinction is merely whether the initial generation is near the storm or at a consider- able distance from the storm. Let us examine some of the evidence favor- ing the one theory or the other in much the same way that we in Naval Operations ex- amined it. For a number of years, Gilmore used the tripartite method for storm tracking. The fact that he was unable to track certain storms he attributed to the presence of geological bar- riers in the ocean bottom which refracted and reflected the microseismic disturbance to the extent that the direction from which the micro- seisms apparently came bore no simple rela- tionship to the direction of original propaga- tion. On scientific and theoretical grounds, I am not in a position to contest this theory—as indeed I am not in a position to contradict any of the prevalent theories. Two things seem apparent, however, in reviewing the earlier Gilmore work. First, if the ocean bottom is really as discontinuous as is indicated, success- ful tracking of typhoons and hurricanes would appear to be improbable. A half-century of storm tracks indicate that those are so non- reproducible that it would be rare for two to follow the exact same path. Newly apparent barriers would be appearing all the time and no large confidence factor could be given to micro- seismic storm tracking. The second thought concerning the results must probably be labelled a more-or-less philo- sophical one. Time and again it has been demonstrated in the history of science that when an accepted theory results in practice in more exceptions to the rule than cases which follow the rule, the theory has been inadequate or incorrect. Certain geological barriers are well established but as more and more unsus- pected ocean barriers appear, the possibility that some additional factor not taken into ac- count by Mr. Gilmore becomes more plausible. It seems possible that the unknown factor might be surf microseisms. On the other hand, of course, the ocean barrier theory may be entirely sound. The work at the Lament Geological Ob- servatory would indicate also that microseis- mic disturbances are produced only in the vicin- ity of the storm. Three dominant arguments are presented. The first is a number of instan- ces when microseismic level was high and ob- served swell was low. The second is the reverse of this picture—observed swell was high but microseismic response was low. The third in- volves cold frontal passages off the east coast of the United States when the seismometers did not record the relatively strong winds pre- ceding the passage of the front off the coast but began to respond only when the front itself progressed over water. Lower wind velocities behind the front did not apparently affect the microseisms adversely. Kammer and Dinger, on the other hand, indicate that microseisms occur when the storm affects shallow water and that the magnitude of response to such "surf action" is such as to mask any direct storm response. An individual not directly involved in mic- roseismic research must accept all of the pub- lished data as valid. He must make a choice between two courses of action. He must either abandon further consideration of the problem until the experts reach agreement or he must attempt to derive some logical explanation which will provide consistency in apparently divergent findings. Abandoning the field to the experts would undoubtedly be more dis- creet. Unfortunately, as administrators of the Navy Microseismic Program, that would im- mediately involve withdrawal of Naval Aero- logical support. A less wise but more practical solution is to attempt to resolve the differences. As a starting point, let us adopt the as- sumption that microseisms can only be pro- duced when two trains of swell intersect and produce stationary or quasi-stationary waves. This phenomenon may occur in the vicinity of the storm or along a coast. Simple surf or shore pounding would, therefore, not produce microseisms. The absence of microseisms in some of the cases when surf measurements showed high waves would thus be explained. The conventional cold front is preceded by southwesterly winds which over a water area would be persistent enough to produce apprecia- ble swell. Following the cold front, the winds are northwesterly. The interaction of the swells produced by these winds and those produced ahead of the front may well produce micro- seisms. The cold front case might be ex- plained in this way. Hurricanes paralleling the eastern coast of the United States could be expected to produce large areas of swell which would first reach the continental shelf and then secondarily produce coastal surf. The time lag observed by Donn might conceivably be the result of the time lag between the primary effect and the secondary effect. Thus microseismic activity would build up before the observed waves at the beach. Bonn's observations are mainly those taken

OCR for page 94
96 SYMPOSIUM ON MICROSEISMS along the New England coast when the storms are traveling rapidly north and east. The Dinger-Kammer observations are made furth- er south where most of the storms are traveling slowly first west and then north. Under these conditions, sea and storm are able to approach the coast at an almost simultaneous rate, par- ticularly when the slowing down resulting from the recurvature of the storm is taken into ac- . count. While nothing definitive is established by this analysis, it would appear that the question of what produces the microseismic phenomenon is still completely unresolved and that some hope still remains that the meteorologist may find a tool for studying hurricanes. It would seem that the first test for de- termining whether microseisms originate in or near a storm is to seek a case when it is well- established that no surf action is occurring but when microseismic activity is marked. In find- ing such a case, it is obvious that East Coast hurricanes are unsatisfactory. The numerous land masses of the Caribbean and the long coast of the United States make it difficult to prove the absence of interfering coastal action any- where along the storm's path. The microseis- mic installation on Guam seemed more appro- priate in establishing the test case. Except for the chain of islands of the Marianas them- selves, Guam is at least 800 miles removed from the nearest appreciably land mass—the Philip- pines to the west and New Guinea to the south. If no surf were observed at Guam, it would be reasonable to assume that microseismic activity was not produced by surf action only. It is not a simple matter to determine whether the island of Guam was under the in- fluence of surf action at a given time or not. There have been no instrumental records avail- able. Observers, at one time or another, have recorded visual observations. Due to the diffi- culty of making an accurate visual estimate and also due to location difficulty—observations were made at fixed points—individual records must be viewed with some question and simul- taneous measurements do not agree. Table 1 shows two sets of data recorded by different observers at different places on Guam at the same time. The lack of agreement is obvious. The somewhat more detailed rec- ord made by Observer B lends more credence to his observations. Using approved sea swell forecasting techniques, a hindcast of the swell reaching Guam was made. The values ob- tained are in last column of the table and com- pare quite favorably with those of Observer B. Table 2 shows a similar comparison of ob- served swell and calculated swell reaching Guam in connection with two other storms. While I am not willing to defend the relative merits of either the observed values or the cal- culated values, it is worth noting that the cal- culated values tend to be higher than the ob- served. Thus any deduction made concerning- the effects of swell would presumably be biased in the direction of exaggeration. Table I Typhoon Marge -- August 1951 Date Observed Observed Calculated Time Swell (A) Swell (B) Swell (GCT) (ft) (ft) (ft) 111200 7 Moderate sea 3-5 .. 121200 6 Rough sea 12 131200 3 Confused 6-8, 3-5 -- 141200 I Moderate sea 12 13 151200 2 Heavy 5-6 9 161200 2 Moderate 8-12 9 Table 2 Typhoon Ruth Date Time (GCT) Observed Swell (ft) Calculated Swell Micro Amplitude (mm) (ft) 071200 1 4.5 11 081200 3-4 6 36 091200 3 7 64 101200 3 7. 169 111200 2 8.5 144 121200 2-3 4 147 131200 1-2 3 180 141200 2 3 141 Typhoon Nora 281200 3 3 17 291200 2 3 20 301200 1 3 29 Only those storms were studied which pre- sented a relatively simple meteorological pic- ture. This choice was made by necessity since the procedure for forecasting swell became too involved when more than one storm appeared on the map of the western Pacific. In all, ten storms were found which met the requirement of simplicity. For each of these, the swell was calculated corresponding to the time of micro- seismic observation, and the development of the storm in a meteorological sense was recon- structed from post-analyzed maps, pilot reports, etc.

OCR for page 94
STORM AND SURF MICROSEISMS 97 Four of the storms developed far enough away from Guam as to produce insignificant swell at Guam. The swell, microseismic re- sponse and center wind force of these storms are illustrated in figures 1, 2, 3, and 4 and Ta- bles 3, 4, 5, and 6. Table 3 Typhoon Ruby -- October 1950 Table 6 Typhoon Ruth -- October 1951 Date Distance Center Swell at Micro Time from Intensity Guam Amplitude (GCT) Guam (mi) (kts) (ft) (mm) 271200 900 40 8 20 280000 900 55 5 32 281200 890 70 5 42 290000 740 90 4 25 291200 720 100 3 40 300000 720 110 3 55 301200 840 100 4 48 310000 1140 65 5 43 Date Time (GCT) Distance from Gu am (mi) Center Intensity (kts) Swell at Guam Micro Amplitude (mm) (ft) 070000 15 3 11 071200 20 4.5 11 080000 30 4.5 15 081200 40 6.0 36 090000 50 6.5 57 091200 260 75 7 64 100000 85 7 80 101200 540 100 7 169 110000 110 9 158 111200 750 110 8.5 144 120000 120 5.0 134 121200 870 120 4 147 130000 110 3 150 131200 1080 120 3 180 140000 130 3 169 141200 1170 110 1 141 Table Typhoon Iris -- April-May 1951 Date Distance Center Swell at Micro Time from Intensity Guam Ampl itude (GCT) Guam (mi) (kts) (ft) (mm) 291200 390 40 3 39 300000 420 55 3 48 301200 450 60 3 45 010000 470 70 5 42 011200 540 95 5 45 020000 600 70 5 36 021200 700 65 5 38 030000 810 100 5 53 031200 850 130 5 65 040000 930 130 4 80 041200 1020 110 4 85 Table 5 Typhoon Nora -- August 1951 Date Distance Center Swell at Micro Time from Intensity Guam Amplitude (GCT) Guam (mi) (kts) (ft) (mm) 280000 270 30 3 19 281200 3 17 290000 545 40 3 20 291200 3 20 300000 800 45 3 27 301200 3 29 310000 1050 75 3 28 Each of these storms illustrate the point that microseismic activity can be significantly above noise level even when no oceanographic activity reaches the coast. The first three of these storms were almost completely unat- tended by surf effects and the microseismic ac- tivity reached a maximum of 85. It would appear that the first crucial test was passed. Of the remaining storms which did pass close enough to Guam to produce significant oceanographic effects, two are presented here. The first illustrates the case when center inten- sification and swell arrival did not coincide, Figure 5. The second shows Typhoon Allyn which passed directly over Guam with both the storm and swell reaching maximum amplitude simultaneously, Figure 6. See also Tables 7 and 8. Table 7 Typhoon Allyn -- November 1949 Date Time (GCT) Distance Center Intensity (kts) Swell at Guam (ft) Micro from Guam (mi) Amplitude (mm) 131200 960 35 3 30 140000 850 60 3 30 141200 750 65 3 38 150000 645 70 4 44 151200 540 105 8 70 160000 480 100 9 77 161200 420 110 13 99 162000 115 14 164 162200 120 22 186

OCR for page 94
98 SYMPOSIUM ON MICROSEISMS Table 8 Typhoon Doris -- November 1949 Date Time (GCT) Distance from Guam (mi) Center Intensity (kts) Swell at Guam (ft) Micro Amplitude (mm) 061200 450 60 3 30 070000 360 70 3 40 071200 300 80 10 50 080000 290 85 22 70 081200 150 90 27 90 090000 100 110 27 140 091200 180 115 27 135 100000 270 120 17 100 101200 300 115 5 90 From these cases, it would appear that the microseismic phenomenon usually recorded, is essentially a dual one, with part of the effect being the result of generation within the storm and part as the result of secondary surf effects. From the order of magnitude of response ob- served on Guam, it appears further that the surf effect is the dominant one having a magni- tude at least twice that of the storm-induced microseism. Despite the paucity of data available, it seemed worthwhile to attempt to find some sort of relationship between the various figures available. One such attempt involved plotting central wind intensity against microseismic response for the three storms which showed WIND IN CENTER (KTS) RANGE CIRCLES CENTERED ON GUAM Figure 1

OCR for page 94
STORM AND SURF MICROSEISMS 99 no swell at Guam. The composite graph, ignor- ing distance, as well as the individual graphs plotting the same values in various distance categories are presented as Figure 7. This rather pleasing result is somewhat soured by the fact that an attempt to subtract the amplitude of the microseism produced by the storm as shown by Figure 7 from the total effect on a storm-surf microseismic record did not produce any relationship between swell am- plitude and microseismic activity. This may be the result of incorrect hypothesis, poor swell forecasting or the result of the fact that the microseismic amplitude as measured is not a simple additive function of the two disturb- ances. Microseismic records seem to indicate that at least two phenomena are being measured simultaneously. Figure 8 shows a construc- tion which seems to bear this out. A simple disturbance having a period of 4 units and a crest-to-trough amplitude of 18 units was CENTER RANGE CIRCLES CENTERED ON OUAM Figure 2

OCR for page 94
100 SYMPOSIUM ON MICROSEISMS added to another disturbance with a period of 5.5 and an amplitude of 30. The individual disturbances are represented by (A) and (B) in the figure while the sum of the disturbances is represented by (c). A sample microseismic record is included as (d). Although the values taken for the simple cases were selected at ran- dom a considerable parallelism seems to exist between the composite record and the micro- seismic record. Another interesting deduction can be made from this construction. Although the original periods were 4 and 5.5, the most marked period on the composite is 5—representing neither the shorter nor the longer wave. Moreover, the amplitude, depending upon where and how it is measured could be recorded as a value between 39 and 43. It would seem from this, that if the reason- ing has been valid to this point, much of the microseismic data collected thus far are in- consistent due to the fact that they do not represent analysis into the component parts. WIND IN CENTER (KTS) RANGE CIRCLES . CENTERED ON GUAM Figure 3

OCR for page 94
STORM AND SURF MICROSEISMS 101 These speculations have led to the develop- ment of the 1952 program for microseismic re- search for the Navy. Wave recorders have been installed around Guam. The amplitude and period data are being so recorded that machine subtraction from the microseismic re- cord will be possible. Perhaps by a frequency analysis the question of where microseisms originate will be solved. To amplify the wave recorder data, further visual observations are being made. Aerologists aboard weather re- connaissance planes are using their drift me- ters to measure the wave length of the swell approaching the coast as well as that leaving the storm. Discussion B. GUTENBERG California Institute of Technology Dr. van Straten in her careful investi- gation came to the conclusion that hurricane microseisms are partly generated within the storm area, partly by surf effects, and that the latter are dominant. As a result of a recent investigation of microseisms connected with non-tropical storms approaching the Pacific coast of the United States, the present author has reached the similar conclusion that these microseisms derived their energy mainly from RANGE CIRCLES CENTERED ON GUAM Figure 4

OCR for page 94
102 SYMPOSIUM ON MICROSEISMS high ocean waves near the recording stations. Sources of energy near the storm center ap- peared to play at best a minor role and even that only as long as the storm was over the ocean and not too far from the recording sta- tion. More detailed results of this investiga- tion which was sponsored by the Geophysical Research Division of the Air Force Cambridge Research Center, are to be published in the Transactions of the American Geophysical Union. In several microseismic storms recorded at stations in California, the State of Washington and in British Columbia during November- December, 1951, the increase and decrease of the microseismic amplitudes were more and more delayed with increasing distance of the recording station from the storm center. In Southern California the time of the largest microseismic amplitudes lagged the time at which the storm center passed the coast (usu- ally in British Columbia or in the State of WIND IN CENTER (KTS) RANGE CIRCLES CENTERED ON GUAM

OCR for page 94
STORM AND SURF MICROSEISMS 103 Washington) by about two days, but they usually coincided approximately with the high- est breakers and ocean waves recorded at or observed at three points in Southern California. For an example, see Fig. 1A. Microseisms traveling along the mountain chains and fault systems which follow the Pacific coast, decreased rather rapidly, along their paths. In some instances these "bar- riers" reduce large microseismic maxima re- corded at some stations to practically normal size while the seismic waves travel for a few hundred miles in the geologically disturbed area. The periods of the ocean waves and those of the microseisms usually showed no parallel- ism (Fig. IB); to the contrary, the microseisms usually reached their largest periods at about the time of the maximum amplitudes, while the ocean waves frequently exhibited relatively short periods when their amplitudes were large. This is a consequence of the well-known fact that the periods of ocean waves usually increase with the distance from the source. CENTER RANGE CIRCLES CENTERED 01 Figure 6

OCR for page 94
104 SYMPOSIUM ON MICROSEISMS The fact that in Southern California the microseisms usually reach their maximum at a time when the storm center has moved rather far inland and when its intensity is decreasing makes it easier to investigate their correlation with the meteorological effects there than at the east coast of North America where the storm is usually moving out to sea and fre- quently intensifying at the time of the micro- seismic maximum. It also seems to be easier at the west coast to distinguish between the rather regular microseisms with periods from 4 to 10 or more seconds (Fig. 2B) and the irregu- lar microseisms with periods near 4 seconds (Fig. 2A). While microseisms of the regular type usually increase and decrease rather slowly and, during their maxima, have periods of at least 6 seconds, the amplitudes of the ir- regular microseisms frequently increase from small values to large maxima within a few hours with the periods remaining close to 4 seconds. This type of irregular microseisms seems to be correlated with strong local winds, especially after passage of cold fronts. Micro- seisms with periods of about 2 seconds are frequently superposed (Fig. 2A), especially during local rain. In the author's opinion, the fact that the irregular microseisms with peri- ods of about 4 seconds are frequently confused with the regular microseisms (they do not al- ways differ as much as the examples in Fig. 2) is the main reason for the failure to reach a conclusion as to the source of energy for these microseisms in spite of extensive investigations for about 50 years. Discussion J. JOSEPH LYNCH, S.J. Fordham University Dr. van Straten has presented a thought provoking paper. She has courageously at- tempted to reconcile two schools of thought on the origin of group microseisms—the school that holds that microseisms originate at the center of a storm and the school that holds that microseisms cannot originate at the center of a storm over deep water but rather at some dis- tance from the center in shallow water, as a Figure 7

OCR for page 94
STORM AND SURF MICROSEISMS 105 result of intermediary surf action very broadly understood. She moreover clearly points out that it is only the practical aspects of the problem that she is considering. Much as I agree with her paper as a whole, I feel obliged as devil's advocate to disagree with or at least to challenge her on some minor points. As the first proponent of the storm center school Dr. van Straten cites the work of Marion Gilmore. She does not mention the work of Fr. Ramirez, presumably because she felt that all in the field are aware that Gilmore's work was based on that of Ramirez and because Gilmore has provided the most extensive appli- cation of the theory of Ramirez. In criticizing Gilmore's work Dr. van Straten refers to the indifferent results that he obtained. It is not quite clear whether she is classifying all of Gilmore's results as indiffer- ent or whether she is criticizing those which he admitted were not too successful. Either way, Gilmore did successfully track some hurricanes following their storm center. This successful part of his work must be kept in view in any evaluation of his results. In trying to track others he was unsuccessful and in an attempt to account for his lack of success he ventured a possible explanation suggested by Dr. Guten- berg—namely geological ocean barriers. Dr. van Straten objects to this explanation, (1) that since the paths of hurricanes differ widely from year to year, the ocean barriers encountered would differ widely from year to year and hence "No large confidence factor could be given to such storm tracking," and (2) that as a theory which in the words of Shakespeare is "more honoured in the breach than in the observance" it should be discarded as inadequate or incorrect. I disagree both with the logic and the wis- dom of Dr. van Straten's criticism. When a marksman scores more misses than hits he should not be dissuaded from trying to hit the Cc) Figure 8

OCR for page 94
106 SYMPOSIUM ON MICROSEISMS target but should be encouraged to find the reasons for his misses. When Sir Isaac Newton first measured the velocity of sound in air by measuring the pres- sure and density, he obtained a value ridicu- lously far from the accepted value. He erroneously attributed his erratic result to the fact that air is not a perfect gas. Since no actual gases are perfect gases, he should, fol- lowing Dr. van Straten's logic have given up the method since it would fail in more cases than it would succeed. Fortunately for Phy- sics, further investigations were conducted. These finally led LaPlace to point out that not the method but Newton's explanation of his failure was incorrect. This may well be true in Gilmore's case. Newton's erratic result arose from his treating sound as an isothermal process whereas actually it is an adiabatic pro- cess. With this correction the method has been used successfully to measure the velocity of sound in all gases. The analogy is far from perfect but applying the same reasoning here, I would disagree with Dr. van Straten's criti- cism of Gilmore's work "that it does not permit a decision one way or another" and say rather that since it has proved successful in some cases, it represents progress and should be con- tinued until the reason for its failure in other cases is definitely established beyond question. As the second proponent of the storm center theory, the work of the Lamont Observa- tory is cited—presumably the work of Wm. L. Donn. Since Donn's work differs from Gil- more's chiefly in that Dpnn assigns a definite theory of the origin of microseisms, I shall pass over this section of Dr. van Straten's paper. I should like to mention however that it is un- fortunate that Ramirez, Gilmore and Donn all used average values of time intervals in deter- mining the direction of the microseismic source. It would have been more satisfying if direction had been obtained from time intervals of indivi- dual waves. The resulting directions could then be grouped into the most prominent ones and the presence of more than one seismic source would at once become apparent. 24 I I I i I I I I I MICROSEISMS. SANTA CLARA, AMPLITUDES WAVE HEIGHTS. ELLWOOD WAVE PERIODS, CAMP PENDLETON WAVE PERIODS.ELLWOOD MICROSEISMS, SANTA CLARA, PERIODS I l I i I 27 29 357 19 21 23 25 27 29 31 I 4 1951 NOV. 25. Oh. DEC.I.Oh. DEC. 17. Oh. 1952 JAN. 2, Oh. Figure 1. (A) Amplitudes of regular microseisms with periods of 5-8 seconds recorded at Santa Clara University, California, during November-December, 1951, and highest recorded ocean waves at Ellwood, California. (B) Periods of microseisms at Santa Clara and periods of ocean waves recorded at Ellwood and at Camp Pendleton, California.

OCR for page 94
STORM AND SURF MICROSEISMS 107 Passing now to the proponents of the surf center theory loosely so called, the work of Kammer and Dinger is cited. Quite properly Dr. van Straten states that she must accept all of the published data as valid. It is one thing to accept the data as valid—it is quite another thing to accept author's conclusions from data as valid. Dr. van Straten states that "Kammer and Dinger demonstrate that microseisms occur only if surf action is appreciable and that the magnitude of response to surf action is such as to mask any direct storm response." I think that is a fair statement of the conclusion of Kammer and Dinger, but I emphasize that it is their conclusion from their data and not merely a statement of their data. Since it is Kammer and Dinger who are being criticized, it seems fair to quote them verbatim. Their conclusion in NRL Memoran- dum Report No. 3 reads: "No microseisms can be identified as being propagated through the earth from the storm center when the storm is over deep water. Therefore the early warning value of microseisms in the Atlantic and Carib- bean seems to be non existent!" I cannot agree that this necessarily follows from their data. The old logicians had a say- ing "QUI NIMIS PROBAT, NIHIL PROBAT!" He who proves too much proves nothing. That Kammer and Dinger did not get any bearings on the center of the storm is an ex- tremely interesting and important fact. That their comparatively meagre experiments dem- onstrate that no bearings can be obtained from the center of the storm is too sweeping an assertion. Gilmore did obtain bearings on the center of the storm in deep water. The inter- esting problem to be solved now is—"Why did Kammer and Dinger fail to get bearings while Gilmore succeeded?" Gilmore too failed on oc- casion. What are the conditions that cause failure ? The problem of the two conflicting views on microseisms is very much like the problem of the nature of light that confronted physicists during the past few decades. Is light a corpus- ONE MINUTE Figure 2. Microseisms recorded by Benioff vertical seismograph at Pasadena.

OCR for page 94
108 SYMPOSIUM ON MICROSEISMS cular or a wave phenomenon? The physicist now finds it is both! Not enough data is avail- able at present for settling this microseism problem. I hope that both Gilmore and Dinger will continue their hurricane tracking inde- pendently so that with increased data an ef- fort can be made to solve the baffling problem. Dr. van Straten's suggestion that we take as a starting point that microseisms are gen- erated by interfering swells is an excellent one and should be the object of intense experi- mental work. Our work with cold front micro- seisms has convinced us that these frontal microseisms originate in the Great Lakes, pos- sibly, as Dr. van Straten suggests as the result of interfering swells. Further and more pro- nounced activity is then noticed when the cold front enters the Atlantic. This may well be the masking action that Kammer and Dinger observed in hurricane micros. Our reason for the conviction that the Great Lakes are a source of micros is the following: When we moved our tripartite station to Poughkeepsie to get away from New York traffic, we noticed a persistent source of two second micros to the West. We recorded from many directions, but there was a persistent source to the West. Since the Hudson River is a sizable body of water to the West we felt it necessary to test this as a pos- sible source of local micros. We set up a sta- tion at West Park on the West side of the Hudson, opposite Poughkeepsie. The source of the micros was still to the West, eliminating the Hudson River as a possible origin. More- over, there were two persistent time intervals indicating two persistent directions—one due West and one North West. Lake Erie was due West of our Station and Lake Ontario, North West. This threw suspicion on the Lakes as the source of our micros. As a first test we set up a third station in the Western part of North Carolina, at Hot Springs. This is due South of Lake Erie and as far South of the Lake as Poughkeepsie is to the East of the Lake. At Hot Springs the source of our micros was due North. Since Lake Erie is due North of Hot Springs and due West of Poughkeepsie we are now convinced that Lake Erie is the source of our persistent two second cold front micro- seisms. We propose in the near future to record both water and ground activity on the shores of one of the Great Lakes. In this con- nection it might be worth mentioning in pas- sing that while working on frontal microseisms in Fort Schuyler, one instrument set up on a concrete pier in the Sound recorded the propel- ler pattern of each passing tug. We agree with Dr. van Straten that most micros are complex waves. Manly has given a simple method of analyzing such wave trains by inspection. He uses the fact that when two waves of different periods combine, the result- ing wave assumes the period of the wave of greater amplitude, the amplitude oscillating, as in ordinary beats between the sum and differ- ence of the amplitudes of the combining waves. In the resultant wave, if the separation of peaks at the maxima is greater than the separa- tion of peaks at the minima, the frequency of the component of greater amplitude is greater than the frequency of the component of less amplitude, otherwise the reverse is true. Sup- pose we have a microseismic wave train in which the time interval between two group maxima is one minute. Suppose there are 12 peaks between the maxima. Then the fre- quency of the component of greater amplitude is 12 per minute and the period of this compon- ent is therefore 5 seconds. If the separation of peaks at the maxima is less than the separa- tion of peaks at the minima, the frequency of the lesser component is greater than that of the major—namely 13 per minute. Hence the period of the lesser component is 4.6 seconds. Suppose the amplitude of the maxima is 11.4 mm and that of the minima is 2.2 mm. Since these are respectively the sum and differ- ence of the constituent amplitudes, the con- stituent amplitudes are 6.8 and 4.6 mm respec- tively. Hence our microseismic wave can be analyzed into two waves of periods 5 and 4.6 seconds and of amplitudes, 6.8 and 3.6 mm. When three waves combine the treatment becomes more complicated, but it is given by Manley (1945). In conclusion I wish to thank Dr. van Straten for the privilege of having been able to pre-digest her very masterful resume of storm and surf microseisms. REFERENCES MANLY, R. G., Wave Form Analysis, New York, John Wiley and Sons, 1945.