Earthquake Research in Japan

Hideki Shimamura
Professor, Graduate School of Sciences, Hokkaido University

24 September, 2003

First of all, let me briefly explain the life cycle of a plate. Plates are actually sheets of hard rock that cover the earth's surface. A plate forms where ridges are located in the ocean and disappears beneath ocean trenches. There are seven major plates around the world. In addition, there are also smaller plates referred to as "micro-plates."

Let me explain the mechanism of how plates are born and disappear. As shown in Diagram 1, there are ocean ridges. These are the areas where plates are born. These plates move. The speed of movement of the world's plates varies from one centimeter per year for the slowest to 12 centimeters per year for the fastest. Hot magma wells up from deep in the earth from beneath the ocean ridges, and the ocean water on top of the ridges cools the magma and transforms it into rock, resulting in the creation of new plates attached to the rear of the ridges. The thickness of these plates varies from 30 kilometers to 150 kilometers. Thus, plates are born at ocean ridges, and when they collide with the earth below at the bottom of ocean trenches, they pierce it and eventually melt away within. That is the life cycle of a plate. When a plate is submerged into the deeper earth, part of it will melt and form fresh magma, which in turn seeks an outlet in the form of volcanoes. In certain spots around the world, for example the Hawaiian Islands, the magma can well up from as far down as more than 1,000 kilometers. This phenomenon is known as a plume.

Why then do major earthquakes occur so frequently around Japan? The Japanese archipelago is located on top of a plate called a continental plate. At the same time, the Philippine Sea Plate and Pacific Plate, both ocean plates, are pushing in and forcing downward as they collide with the continental plate at the bottom of the trench. Since these plates are in constant motion, the result is that the continental plate on which the Japanese archipelago rides is forced down deeper into the earth through friction. This process cannot continue indefinitely, however, since there are limits to the bearable tension. Specifically, when the plate is forced down about three to six meters deeper, it snaps back fiercely. This rebound takes the form of a large earthquake. When such a major earthquake takes place, it does not mean the end of earthquakes, since the process leading to another one commence immediately.

Generally speaking, Japan is subject to two types of earthquake. First, there is the kind in which an ocean plate sliding into the earth and the continental plate on which the Japanese islands ride collide. This is referred to as an ocean-trench-type earthquake. At the same time, since the Japanese archipelago itself is being constantly pushed by the ocean plates, it is bent and twisted, giving rise to another type of earthquake beneath the archipelago. In other words, the earthquakes plaguing Japan can be divided into two major categories: those due to primary causes, namely the ocean-trench-type earthquake, which occurs at the borders between plates, and those due to secondary causes, namely the direct-hit-type quake, a strong tremor that occurs in relatively shallow areas immediately below the ground through the distorting of the Japanese archipelago by the plates bumping into each other.

Near Japan we have the Pacific Plate, an ocean plate that is moving toward the Japanese archipelago at a rate of approximately 10 centimeters per year. Then there is another ocean plate, the Philippine Sea Plate, which is oriented slightly differently and is moving at a speed of about four centimeters per year. Diagram 2 shows stereographically the four plates near Japan that bump into or submerge under each other. As is apparent from this diagram, the location of the areas where mammoth earthquakes, that is to say, earthquakes of magnitude eight class [M8], hit Japan is well established. Nearly all of them are centered off the Pacific coast. These are ocean-trench-type earthquakes. Since the plates are in constant motion, the probability is quite high that earthquakes will take place repeatedly in the same area every 80 to 150 years.

Changing the subject slightly, I would like to discuss felt earthquakes, that is to say, the kind of earthquakes that we are physically aware of. Many of you probably felt an earthquake for the first time here in Japan. But even within the Japanese archipelago, there are places where we can often feel an earthquake, and others where we seldom do. In some areas people are aware of more than 50 earthquakes a year, while in others they recognize fewer than five. You are now in Tokyo. If you go only about 100 kilometers north of Tokyo, you can feel earthquakes even every week.

Up till now I have been discussing earthquakes that occur off the Pacific coast, but recently we have become aware that such ocean-trench-type quakes also take place outside this Pacific coast zone. We learned this clearly when a major earthquake occurred off the southwest coast of Hokkaido, the Hokkaido Nansei Oki Earthquake. This earthquake occurred back in 1993 and was followed by a tsunami in which more than 230 people died or went missing. When we analyzed this Nansei Oki Earthquake, we came to realize that the seismic event was not limited to Hokkaido but actually was a major global event in geoscientific terms. As Diagram 3 shows, at the top of the ocean ridge on the Atlantic Ocean side there are two plates--namely, the Eurasian Plate and the North American Plate. A gap being created between them causes the two plates on the Pacific Ocean side to come together and collide. This mechanism leads to the occurrence of major earthquakes on the Japanese side.

That being so, there is a possibility that the same mechanism will lead to the occurrence of other quakes off the Japan Sea coast. Many seismologists now believe that the next major earthquake may well take place in an area that until now has not been hit by quakes. Nevertheless, it is extremely difficult to forecast where the next earthquake might occur. One reason is that we have very little detailed evidence concerning, for example, the epicenter of big earthquakes that took place prior to around 1940.

Let us return to the global perspective. The longest mountain range in the world, which is actually to be found on the sea floor, stretches for 75,000 kilometers. This is an "ocean ridge" that I was referring to earlier. My research covers earthquakes occurring in areas near to both the trenches and the ridges and their underground structure. Here I would point out that up to 85 percent of the earthquakes hitting Japan are centered not under the land mass of Japan but rather below the adjoining seabed. In other words, research on the seafloor is extremely important.

We have invented the ocean bottom seismograph, a tool that was not available previously, to measure earthquakes on the ocean bed [see Diagram 4]. The special features of our ocean bottom seismographs are that they are very small but extremely reliable. They are designed in such a way that these seismographs can be positioned and retrieved from any type of ship. That is because when carrying out observations abroad, we never know ahead of time what size or type of vessel we might have to use. Performance-wise, they are capable of carrying out uninterrupted monitoring for two months. Moreover, because in most places the sea bottom is quite deep, unless equipment is capable of withstanding water pressure at 6,000 meters depth, the number of places where seismographs can be positioned is extremely limited.

From 1987 until this year we have taken our ocean bottom seismograph and engaged in joint research with 14 universities and research institutes in 10 countries: the University of Bergen (Norway), the Icelandic Meteorological Office, the University of Paris (IPGP, France), the Geophysical Institute, the Polish Academy of Sciences, Lisbon University (Portugal), the University of Hamburg (Germany), the University of Iceland, Cambridge University (UK), the Icelandic Energy Authority, the Alfred Wegener Institute (Germany), the Argentina Institute of Antarctic, Victoria University (New Zealand), and the Istanbul Technical University (Turkey).

Most of them are located in European countries, and most of the experiments were conducted in the North Atlantic. Diagram 5 provides data on those of our experiments that were carried out in the Atlantic Ocean, with the boxes showing the sea areas where the experiments were carried out for the respective years. Since otherwise there might be some confusion, let me clarify that although they are not included in this diagram, we carried out several other experiments in the region from 2000 to 2003. This monitoring of earthquakes on the seabed had two objectives. First, we wanted to measure regular naturally occurring earthquakes and to identify where they took place and due to what mechanisms. Another goal was to use the ocean bottom seismographs, just like the stethoscopes and X-ray machines used to examine the condition of the human body, to investigate the internal structure of the earth. Of course, in addition to the investigations shown in this diagram, we have carried out experiments in waters close to Japan. We have also done so in the Sea of Marmara in Turkey and the Caribbean Sea.

As shown in Diagram 6, in 2001, in a joint project with universities in New Zealand and Britain, we carried out experiments in New Zealand, installing seismographs both on land and in the ocean so as to research the structure within the earth. As is the case with Japan, the Pacific Plate is burrowing under New Zealand, and the theme of the research was to discover the similarities and differences with Japan.

Diagram 7 shows one of the results of a joint experiment project being carried out in the North Atlantic with Norwegian and other universities. This research is designed to investigate the "history" of the North Atlantic. In the North Atlantic area, 58 million years ago, Norway and Greenland were joined together. As they were split apart, the Atlantic Ocean was born in between and has developed until now. Our research theme is to investigate this history of the Atlantic and to find out why petroleum-laden strata developed, since as you know this area of the ocean has rich oil reserves. In addition, close to the ocean ridge there we continue to be engaged in earthquake monitoring and research with researchers from Iceland and Britain.

Now let us return to the situation in Japan. We counted the number of earthquakes that hit Japan for every 10 years from 600 A.D. until now [see Diagram 8]. As the diagram shows, in recent years the occurrence of damage-causing earthquakes has been increasing drastically. However, I would caution here that an increase in the number of such quakes does not necessarily signify increased earthquake activity as such. What I am trying to say is that earthquakes of the same size as before are now causing more damage. In other words, this increase in the amount of damage caused by earthquakes is a product of the advance of civilization.

Unfortunately, from the perspective of seismologists, the 1995 Great Hanshin-Awaji Earthquake, affecting the southern portion of Hyogo Prefecture, was by no means exceptional, since M7 class tremors invariably take place once or twice a year in Japan and its vicinity. Moreover, for example, Kushiro City in Hokkaido registered a greater degree of acceleration--that is, the strength of the earthquake's sway--in a 1993 earthquake than that measured in Kobe. Yet even though more than 6,400 individuals lost their lives in the Kobe earthquake, in the Kushiro quake only two people died. Actually, the amount of shaking that occurred in Kobe during the quake can be said to have been not all that special. This catastrophe highlighted just how susceptible Japan's urban areas are to damage from earthquakes and just how dangerous these quakes, which would not cause all that much devastation if they hit other areas, can be if they hit major metropolitan centers.

The Great Hanshin-Awaji Earthquake was of the secondary type I referred to earlier, in other words a direct-hit-type tremor. As you will recall, this type of earthquake is caused by the bending and distortion that the Japanese archipelago is constantly undergoing, so there is a possibility that they can occur anywhere throughout the archipelago. Compared with the ocean-trench-type quakes that are limited to the areas off the Pacific and Japan Sea coasts, the direct-hit-type earthquakes can be particularly devastating and are especially fearsome since these secondary-type earthquakes can take place anywhere. In terms of frequency of repetition in the same place, however, whereas ocean-trench-type earthquakes tend to recur once every 80-150 years, the interval between these direct-hit secondary earthquakes taking place in the same area is said to be extremely long, at least about 1,000 years.

Another important point to bear in mind is the kind of structures that collapsed in the Great Hanshin-Awaji Earthquake. The topmost column in Diagram 9, which gives data on the type of buildings that collapsed or did not collapse in this magnitude 7 earthquake, shows structures built prior to 1972; the next column shows those built between 1972 and 1981, and the bottom column shows structures built since 1982. The vertical columns show, from right to left, the breakdown for five degrees of damage incurred--total destruction or collapse, major damage, medium damage, minimal damage or less, and unknown. You can see that the most damage occurred to buildings erected prior to 1972, followed by those built between 1972 and 1981; while the least damaged were those built since 1982. The conclusion to be drawn is that the older the structure, the more dangerous it is likely to be. This phenomenon is proof of the great contribution of the incremental bolstering of building standards that has occurred in response to major earthquakes in various parts of Japan.

Next, I would like to discuss seismic waves. There are both body and surface waves, and the former come in two types: P waves and S waves. If you look at the record of earthquake activity recorded by a seismograph as shown in Diagram 10, you can see that the order of wave arrival is P waves first, then S waves, and finally surface waves. In this case the epicenter was relatively shallow and the recording took place at a spot 500 kilometers distant from the epicenter. As you can see, although the surface waves are the last to arrive, their amplitude is the largest. The distinctive feature of surface waves is that they flow across the surface of an object, such as the surface of the earth. Physically speaking, such waves that are transmitted across a surface decline in force proportionally to the square value of the distance, whereas body waves diminish in proportion to the cube of the distance. Thus, surface waves can be transmitted farther without loss of energy.

Next, I would like to discuss earthquake-resistant designs, or earthquake-resistance standards, in Japan. As can be seen from the damage that occurred in the Great Hanshin-Awaji Earthquake, the older the building, the weaker the structure. This shows how Japan's earthquake resistance standards have incrementally increased over time, but problems still remain. For one thing, records are scanty from past major earthquakes on what designs held up strongest--that is, actual records of strong swaying during earthquakes. The problem is that we have astonishingly little data on exact structural movement during earthquakes, so we cannot take advantage of the lessons from swaying that occurred during past quakes. Another problem is that although here in Japan we conduct simulations when constructing buildings or civil engineering structures, the actual examples of waves--only body waves--that are employed in the calculations are extremely limited in scope. Simulations are not conducted based on data in terms of surface waves.

Not only are adequate simulations not carried out prior to construction, but also we do not have any case in which tall buildings or other high structures have been subjected to surface waves of great amplitude, which do not decline greatly even over considerable distance and have extremely long periodic motion. For example, although the Great Kanto Earthquake of 1923 was the most catastrophic event in the history of Japan, causing a toll of more than 140,000 dead or missing, it was actually a rather shallow earthquake that occurred below the ocean floor of Sagami Bay off Shizuoka Prefecture. If a similar quake were to occur in Sagami Bay in the future, and powerful surface waves were transmitted to Tokyo Bay, many seismologists are concerned about what might happen, for example, if they hit the waterfront subcenter district. It is conjectured that in light of the fact that tall buildings and other structures have very low resonance frequencies, if waves with low resonance frequencies arrived after an earthquake, it would create a phenomenon of resonance with the buildings, causing their amplitude to become even greater.

Another concern is what would happen to nuclear power plants in the event of an earthquake. For example, the Hamaoka Nuclear Power Plant is located within what would likely be the epicenter zone in the case of the so-called "Tokai earthquake," which it is feared could arrive at any time. This facility is designed to withstand an acceleration rate of 450 gals. Actually, Hamaoka has four nuclear reactors, with reactors number one and two being able to withstand up to 450 gals. Reactors number three and four, built later, are a bit stronger, but even so only a portion of their equipment can withstand up to 600 gals. Thus, the emergency core cooling system, a critical mechanism, is still only designed to withstand 450 gals.

Be that as it may, higher acceleration rates than these have been recorded during recent earthquakes. For example, during the Western Tottori Earthquake of 2000, at a location eight kilometers from the epicenter, and moreover on bedrock, acceleration of 575 gals was recorded. We realized that for the first time thanks to the seismographs that have recently been positioned on bedrock beneath the surface in various places. Generally speaking, during an earthquake the swaying is most pronounced on the surface, and that was true in this case as well. But in any event, 575 gals is clearly a much greater value than the assumed level of 450 gals. Therefore, many seismologists are beginning to believe that such data may indicate that we have a significant problem on our hands.

Another problem is the assumed earthquake magnitudes relied upon in the construction of nuclear power plants. Their designs assume maximum exposure to direct-hit-type quakes of M6.5. Since I have not done so already, let me briefly describe how magnitude measurements work. A difference of one level in magnitude represents a difference of a factor of 30 in the energy level of the earthquake in question. Therefore, an M7 earthquake is 30 times more powerful than an M6 quake, and likewise an M8 earthquake is 30 times more powerful than an M7 quake. This means that there is a difference of approximately 1,000 times between the energy let off by M8 quakes versus M6 quakes. So even though M6.5-level earthquakes might be provided for in the design of a plant, recently we have been having a series of direct-hit-type quakes directly below Japan exceeding M6.5 on the scale, such as the M7.3 one in the western part of Tottori Prefecture in 2000 or the M6.7 Geiyo Earthquake of 2001 between Hiroshima and Ehime Prefectures. And remember that the energy released by an M7.3 quake is 20 times larger than that released by an M6.5 quake.

That is the end of my presentation. I regret that because of our lack of time today we could not go into more detail. I would encourage any of you who would like more explanation to ask me questions or to consult a recent book I wrote for the general public titled Jishingaku ga yoku wakaru--Dare mo shiranai chikyu no dorama [Really Understanding Seismology--The Unknown Drama of the Earth], published by Shokokusha in September 2002. Thank you for your kind attention.


Q: Recently experts and members of the mass media have repeatedly speculated that Tokyo may soon be due for a major earthquake. Have you used the seismographs and other equipment you normally employ in the vicinity of Tokyo to conduct any experiments and investigate whether it seems likely that a major earthquake is imminent?
A: We have not conducted any direct monitoring in the Tokyo area. Because Tokyo and its environs are the areas where great caution is required, a number of experts, including those associated with the Meteorological Agency, universities in Tokyo, and the former Science and Technology Agency, are engaged in such monitoring. I do not believe that as yet any of them have detected anything that could be taken as a harbinger of a major earthquake. Considering that a major urban area like Tokyo does not really offer the best conditions for measurements, however, we are not in a position to carry out adequate monitoring.

Q: Of the two major types of earthquakes, the ocean-trench type and the land-based direct-hit type, which poses the greater threat to Japan? In other words, which would be likely to cause the most destruction? Also, what are the special characteristics of the destruction caused by each type?
A: The short answer is that both are terrifying. As I noted earlier, the ocean-trench-type quakes are overwhelmingly greater in sheer energy released, with more than 30 times as much power as an M7 quake occurring directly below the land mass. For that reason, their geographical scope of destruction is extremely broad. In addition, because they occur at the bottom of the ocean, they can generate tsunami. For these two reasons, the ocean-trench-type earthquakes are really very alarming.
At the same time, as I noted earlier, we have no way of knowing where a direct-hit-type quake might strike in Japan. When one occurs, even though it is only of the M7 class--in other words, with roughly only one-thirtieth the released energy of an ocean-trench-type quake--it can cause tremendous damage depending on the point of impact. For this reason, I would say that a direct-hit-type earthquake occurring below a major urban area is a particularly dreaded type of quake.
The "Tokai earthquake" is an ocean-trench-type quake that is likely to occur under very special circumstances. Although I did not have the chance to discuss it in depth today, the mechanism of this particular ocean-trench-type quake is such that roughly half of its seismic focus would extend to land, so that it might cause the very awesome phenomenon of a quake that acts quite like a land-based direct-hit-type quake. The resulting damage could be appalling.

Q: I would like to ask a question concerning budgeting for earthquake research. I believe that one major focus of concern among seismological societies and seismologists is who can secure the most funds, since all parties concerned need money to pursue their own research. Even though much has changed since the enactment of the Large-Scale Earthquake Countermeasures Law in the wake of the Tokai Earthquake in 1978, I think that there still tend to be two major schools of thought: one favoring emphasis on earthquake prediction and the other wanting the bulk of funds to go for post-quake countermeasures because of the difficulties involved with predicting tremors. Which side is winning out as far as budgeting is concerned?
A: That is an extremely complicated question and would take a lot of time to answer properly. Let me just say that earthquake prediction research in Japan has reached a very difficult pass. Just 20 years ago the future of earthquake prediction in Japan looked quite rosy, and it seemed that it was about to start scoring major triumphs. For example, prior to the Izu-Oshima Kinkai Earthquake of 1978, changes in the chemical composition of groundwater were recorded. It was felt that if scientists could get a handle on this phenomenon, they would certainly have found a mechanism for earthquake prediction. As a result, money poured in for the earthquake prediction research system, and research in this area pushed ahead.
Nevertheless, the situation has changed since the Great Hanshin-Awaji Earthquake. In terms of giving a direct answer to your question, I would have to say that as things now stand, it is not the scientists who have triumphed but rather the bureaucrats affiliated with the former Science and Technology Agency. They have gotten a hold of the budget and are using the money to install seismographs in local communities the length and breadth of Japan--in other words, spending vast sums on earthquake monitoring that is "visible to the public." Nevertheless, these seismographs are not the kind of equipment best suited to research.

Q: You spoke of the situation at the Hamaoka Nuclear Power Plant. In that regard I read in the press statements made by academic experts Dr. Kiyoo Mogi, [former chairman of the Earthquake Assessment Committee for Areas under Intensified Measures against Earthquake Disasters], and Dr. Katsuhiko Ishibashi [of Kobe University] at an international conference held in Sapporo this summer [the general assembly of the International Union of Geodesy and Geophysics] to the effect that nowhere else in the world has a nuclear power plant like the one at Hamaoka been built at a location where a major earthquake is forecast, based on such a low earthquake magnitude assumption. Is it true that the situation at Hamaoka is really unique in the world? And in reference to these "assumed" magnitude levels, are these figures that the electric power companies just come up with on their own when they want to construct new nuclear power plants?
A: Let me begin by addressing your last question. These design guidelines are established by the national government, not the individual electric power firms, and they determine how things proceed. The criteria for designing the construction of nuclear power plants [Examination Guide for Seismic Design of Nuclear Power Reactor Facilities] were established in 1978 and, after being partially revised in 1981, have remained unchanged since. In addition, in 1995 after the Great Hanshin-Awaji Earthquake, the then Ministry of International Trade and Industry issued a report entitled "Earthquake-Resistance Safety for Nuclear Power Plants" that affirms that adequate countermeasures for earthquakes are ensured for nuclear power plants from the time of construction through operation. In other words, since the design guidelines have not fundamentally changed at all since 1978, there is criticism to the effect that they do not incorporate the advances made in seismology since then.
In reference to conditions outside Japan, there are other nations that are, like Japan, frequently hit by earthquakes. For example, Taiwan is another country that is earthquake-prone. So in that sense, we can say that the building of nuclear power plants in certain other countries is also risky.
Another thing to remember is that although the European continent is a very stable continental mass subject to few earthquakes, about once every millennium it is directly hit in very mysterious fashion by an exceptional kind of inland direct-hit-type quake. For example, in the fourteenth century [1356] the city of Basel in Switzerland was leveled by an earthquake. For that reason, we have no scientific proof to conclude that such a devastating earthquake will not again hit, say, France or Germany. We do not know why the Basel earthquake occurred where it did at that time, but we can venture to say that there is certainly a possibility that this kind of quake will occur again somewhere on the European continent.

Q: How credible are the forecasts of Yoshio Kushida that have attracted such great attention over the past week or so? [Note: Kushida is an amateur who, in utilizing the back scattering of FM radio waves to monitor the frequency of the appearance of meteors, claimed to have discovered in the data he gathered changes that could be considered to precede the behavior of the earth's crust. Thereafter, he began to gather data allegedly showing a correlation between observed changes in the patterns of broadcast FM radio waves and seismic activity and to conduct public experiments in detecting premonitory symptoms of earthquakes, leading to his forecast that an earthquake would strike soon in the Kanto region.]
Also, based on your remarks, it would seem that earthquake damage to major metropolitan areas is likely to be especially large and that the best way to limit such damage would be to evacuate the major cities. What exactly should individual Japanese people do ahead of time to minimize damage from major quakes?
A: As for your first question, we might say that Kushida's methodology is that of an amateur, that is to say, someone who is not a member of the scientific community. He claims to have observed that abnormalities in the transmission of radio waves can presage earthquake activity. If the predictive indicators that he claims emanate from the epicenters of earthquake areas, they also contain phenomena that would be impossible from the standpoint of physics. It could just be that Kushida has discovered some kind of relationship till now unknown to physicists, but I believe that generally speaking mainstream seismologists have no confidence in his methodology.
As for your other question, I would say that it is indeed preferable not to live in Tokyo. If you have to reside in Tokyo for employment reasons, however, I would recommend living in a relatively new, not-too-tall structure. Here I would note the case of an acquaintance of mine who teaches at Kobe University. Even though prior to the big earthquake no one around there thought that a quake would come, he made sure never to put any chests of drawers or other large furniture in his sleeping quarters. That saved him when the Great Hanshin-Awaji Earthquake struck. Since there is a great danger of being crushed in one's home by heavy furniture, even if the structure itself does not collapse, I would recommend taking such precautions here in Tokyo, too.

Q: I have a question regarding construction technologies. When the Kasumigaseki Building was constructed, I believe that it was considered experimental in many ways, and since then a host of high-rise buildings have gone up in Tokyo. In fact, they seem to be sprouting up one after another recently. Is that because of real advances in earthquake-resistance performance and construction technologies? Is there any justification for this rash of high-rise building?
A: As I pointed out in my remarks, we do not have any records whatsoever concerning the surface waves that would be created during a major earthquake. So in many cases we do not know how much a building would shake if surface waves arrived. For this reason, the truthful answer would be that there is no way to achieve accurate progress in building techniques as far as surface waves are concerned. On the other hand, in regard to body waves, up until a few years ago simulations for building construction in Japan relied on data recorded for two overseas earthquakes as many as 50 years earlier. This data was recorded on strong-motion seismographs set up in the towns of El Centro and Taft in the United States and the quakes in question were the Imperial Valley Earthquake and Kern County Earthquake. I believe, however, that recently designs have become somewhat stronger as architects take into account the fact that the body waves engendered are likely to be somewhat more powerful than in those cases.
Here I would like to emphasize that I am a scientist, not an engineer. That is to say, someone who is designing a house will take his calculations and then multiply them by a "safety coefficient." His rationale here is that the use of this safety coefficient will cover for the unknown dangers that cannot be accurately predicted. For example, we have many aircraft flying through the skies every day, even though they are all subject to very alarming degrees of metal fatigue. That, I believe, is another example of how, through the use of safety coefficients, the engineering side seeks to allow for the unknown, incalculable factors.

Q: I would like to ask a question from the standpoint of a complete layman. How far ahead can seismologists accurately predict earthquakes? Also, you said that the cycle for the occurrence of ocean-trench-type quakes runs between 80 and 150 years. Where are we along in this cycle in terms of the recurrence in the Tokyo area of a major quake on the scale of the Great Kanto Earthquake of 1923?
A: The occurrence of physical phenomena involving breakage, such as earthquakes, involves a great deal of fuzziness. For example, if you drop these two glasses on the floor, they will shatter in different ways. So when science looks at the breakage of physical objects, it faces great difficulties, and inevitably this area of science remains rife with ambiguities.
Next, let me explain about how a major ocean-trench-type earthquake could affect Tokyo. Concern about the likelihood of a Tokai earthquake began to be voiced loudly from the middle of the 1970s, and at that time the majority of Japanese seismologists thought that such a quake was likely to take place in about a decade, or at the latest within two decades. However, already close to 30 years have passed since then. So in that sense unfortunately we are still lost in ambiguities. Perhaps we should just accept the fact that by definition nature remains cloaked with that degree of ambiguity. As for your question about how a quake might hit Tokyo, in the event of a Tokai earthquake, I would expect that it would represent an M5 level quake for the Tokyo area. That being so, the damage would likely not be all that great. More threatening would be a repeat of the same kind of fearsome quake that hit Tokyo in the form of the Great Kanto Earthquake of 1923. However, the majority of scientists do not think this is likely to happen for more than another century.
Nevertheless, if a direct-hit-type quake struck Tokyo, that could be truly terrifying. In 1855, during the Edo period, the city was hit by just such a quake in the form of the Ansei Earthquake, which cost the lives of more than 10,000 people. Although such activity has been relatively rare in recent years for unknown reason, somewhat smaller tremors than what we have been discussing occurred on countless occasions during and before the Meiji period. That being so, it is quite possible that a direct-hit-type quake could strike Tokyo some day. But we cannot tell at all exactly when. As I mentioned earlier, there is a difference of 30:1 in terms of the energy released by an M8 quake versus an M7 quake. This means that a quake with a lower magnitude value takes less time to accumulate that much energy. For that reason, the odds are very high that energy that has been accumulated in this fashion could hit Tokyo with devastating force at any time.

* Given on September 24, 2003, at the Foreign Press Center/Japan. This paper is reserved for internal use; any reproduction or quotation is forbidden without prior permission from the FPC.