INTRODUCTION

Almost three years have passed since the Kobe earthquake of January 17, 1995. This earthquake, which resulted in more than 6,000 deaths, was one of the most disastrous to hit Japan since the Great Kanto earthquake of 1923. A large number of buildings and houses collapsed, increasing the final death toll, while severe damage to highway and railway bridges seriously affected rescue and ambulance operations and hindered restoration activities after the earthquake. Lique-faction occurred extensively on reclaimed coastal land; the city's lifeline networks of water, gas, and electricity were cut off; and urban functions were paralyzed for a long time.

Why did such serious losses occur in Japan which is often thought of as being in the forefront of earthquake engineering? The author of this study conducted research in earthquake engineering for some 30 years after completing university, but did not even dream of the occurrence in Japan of an earthquake that would take such a heavy toll of lives. The earthquake has made it obvious that a groundless belief in structures sufficiently resistant to earthquakes exists in Japan and that overconfidence in disaster preventive measures in urban areas contributed to the considerable damage. It goes without saying that we civil engineers have an obligation to turn our full powers to the creation of a society safe from such earthquakes by reviewing what we have done so far and by fully assimilating the lessons we have learned from the Kobe earthquake. To this end, the full extent of the damage to various structures and facilities has been investigated and relevant groups have analyzed the causes of the damage. Based on the results of the investigations and analyses, earthquake-resistant design standards are now being updated.

Shortly after the earthquake, the Japan Society of Civil Engineers (JSCE) organized a Special Task Committee on Earthquake Resistance of Civil Engineering Structures. Taking into account the damage to civil engineering structures caused by the earthquake, the committee has continually discussed critical subjects, such as prompt review of measures against earthquakes and what future earthquake-resistant structural designs should be. The results of the discussions have already been released as phase 1 and phase 2 proposals.

This paper introduces the basic concept underlying the earthquake resistance of civil engineering structures and the future measures for the mitigation of earthquake hazard discussed in the proposals. Further, I would like to discuss what has been done and solved after the earthquake, and which subjects require further study.

DEVELOPMENT OF EARTH-QUAKE ENGINEERING AND INFRASTRUCTURES IN JAPAN

Before recalling the damage caused by and the lessons learned from the earthquake, a brief description is given below of the history of earthquake engineering and development of infrastructure in Japan. Table 1 lists the characteristics of damage caused by earthquakes that have struck Japan over the past one hundred years, and the history of revisions to earthquake-resistant design standards. With the magnitude 8 Nobi earthquake of 1891 as a turning point, the government organized the Earthquake Disaster Prevention Committee, and research into earthquake damage prevention engineering in Japan entered its current, modern period. It was after the terrible experience of the 1923 Great Kanto earthquake that the seismic coefficient method, a method by which a certain proportion of the total dead load of a structure is applied horizontally to the structure in earthquake-resistant design, was proposed and earthquake engineering started in Japan. By taking a general view of the history of both earthquakes and the subsequent revisions made to earthquake-resistant design methods, it can be understood that phenomena which caused severe damage were isolated and research then began to supplement and revise the design methods each time an earthquake caused heavy damage.

In the 1964 Niigata earthquake, liquefaction occurred in sandy sub-soil, causing damage to many buildings and lifeline facilities; this included the collapse of the Showa Bridge shortly after the completion of its construction. With this earthquake as a starting point, research into liquefaction began and the results were subsequently taken into consideration in earthquake-resistant design. In the 1968 Tokachi-Oki earthquake, many RC structures were shear-fractured. Based on this, a study was launched on the method of increasing shear resistance of concrete structural members, and a method of checking ultimate strength extending beyond elastic into plastic regions was adopted for the earthquake-resistant design of RC structures.

In the 1978 Miyagiken-Oki earthquake, the first earthquake to cause a severe damage to a modern city such as Sendai, lifelines such as gas, city water and sewer systems suffered considerably. With this as an impetus, lifeline earthquake engineering began and earthquake-resistant standards for buried piping were updated. In the 1983 Nihonkai-Chubu earthquake, a phenomenon called lateral ground flow, in which liquefied subsoil moved as much as a few meters horizontally, occurred and seriously damaged buried pipelines. Although research into lateral ground flow began with this earthquake, sadly, the Kobe earthquake struck and caused identical damage before the research results could be reflected in the earthquake-resistant design standards.

As described above, when looking back at the history of earthquakes and earthquake-resistant design standards in recent years, it can be said that research in earthquake engineering and the updating of design standards has always been stimulated by earthquake occurrences. An earthquake is a natural phenomenon; the extent of damage it will cause is not entirely foreseeable. However, as far as history is concerned, it is quite appropriate that we be criticized that "earthquake engineering and design standards are never updated until after earthquakes."

Certainly, there has been such damage that could have been avoided if we had more carefully observed damage caused by past earthquakes and more thoroughly learned the lessons derived from those earthquakes. Lateral ground flow due to liquefaction is one such overlooked area. As mentioned earlier, no quantitative investigation of lateral ground flow was made until after the 1983 Nihonkai-Chubu earthquake. Truthfully, recent studies have revealed that lateral ground flow also occurred in both the 1923 Great Kanto earthquake and the 1948 Fukui earthquake. Residents who experienced the Niigata earthquake were quoted as saying, "The ground moved a few meters horizontally," "a road was bent by the earthquake," and "the width of the Shinano River narrowed because the riverbank moved toward the center of the river." These are all evidence of the occurrence of lateral ground flow, but no investigation was made and earthquake damage was left unnoticed.

After the Kobe earthquake, the most talked about subject has been the liquefaction of so-called Masado that was used for reclaiming near-shore water areas. Masado, granite decomposed by weathering, is found extensively in the Rokko Mountains, and contains a high content of gravel. Since soil with a high gravel content was generally considered to be resistant to liquefaction, the liquefaction phenomenon captured much of the attention after the earthquake. One widely held view was that the higher the gravel content in soil, the higher the resistance to liquefaction because higher permeability, which was thought to increase the soil strength, means it is more difficult for pore water to increase. Nevertheless, due to strong earthquake motions, high-gravel-content soil liquefied and caused damage to many structures during the Kobe earthquake.

Looking back at past earthquake damage, there has been some evidence that soil containing gravel liquefied. In the 1948 Fukui earthquake, large scale liquefaction occurred in alluvial plains in the basin of the Kuzuryu River. According to reports, soil mixed with gravel spouted out of the riverbeds and surrounding ground. Since the Fukui earthquake was of the same type as the Kobe earthquake, i.e. an inland earthquake, both earthquakes must have had similar earthquake motion characteristics. A phenomenon similar to that observed in the Kobe earthquake had already occurred in the Fukui earthquake about 50 years ago! A later report on the 1994 Hokkaido Nansei-Oki earthquake also stated that volcanic sediments containing high amount of gravel liquefied.

Tracing the history of earthquake engineering and revisions to earthquake-resistant standards for structures, it cannot be over-emphasized that it is a history of earthquake damage oversights. In this sense, the responsibility of our profession is extremely significant. So as not to repeat history, we have to carefully learn lessons from the Kobe earthquake and ensure that there will be no serious errors or oversights in preventive measures against future earthquakes.

SUMMARY OF EARTHQUAKE DAMAGE TO CIVIL ENGINEERING STRUCTURES

The first point to note about damage to civil engineering structures caused by the Kobe earthquake is that elevated highway bridge piers were destroyed. Single column RC

(*):In-Land Earthquake

* Toll by Fire Defence Agency May 21, 1995

** Collapsed and Extensively Damaged

bridge piers collapsed in Fukae Ward, Kobe. Although there had been RC bridge piers damaged by earthquakes in the past, this was the first total collapse anywhere. Most of the seriously damaged piers were designed in accordance with pre-1980 earthquake-resistant design standards. The piers of concrete structures having low ductility, or low resistance, were shear-fractured, resulting in such major failures. Damage to RC piers designed in conformance with the current earthquake-resistant design standards (updated in 1980 and after) was not so severe as to result in bridge collapses. However, some of the bridge piers designed and constructed based on the current standards suffered appreciable damage. Photo 5 shows typical damage to a bridge pier of the Harbor Highway which was put into service in 1993, two years before the earthquake. Based on the current design standards, a study had been made on the ductility of the pier by checking its ultimate strength. Fortunately, the pier had not completely collapsed, but cracks penetrated the entire section of the pier, indicating a need to review the current design standards.

Another point to note is the damage to steel bridge piers. As typically shown in Photo 6, many steel bridge piers buckled. As a general rule, most steel structures are designed by the allowable stress design method, a method by which stresses in steel structural members fall within an elastic region. Little research has been done on the characteristics of deformation of steel structure in the plastic region. This is one of the technical subjects that the earthquake showed needs to be promptly investigated and studied.

Damage to underground structures, typified by the damage to subway structures, has also become a focus of attention. The most severe damage was at Daikai Station, which is of box-type RC construction, on the Kobe Express Railway, where intermediate columns were shear fractured and an upper floor deck plate collapsed along with the overburden soil. Damage to other underground station structures was also reported. Tunnels, such as subway tunnels, constructed by the cut-and-cover method as well as mountain tunnels were damaged.

Before the earthquake, it was generally thought that underground structures would not suffer heavy damage, even if they were subjected to strong earthquake motions. It had been considered that the ground would become unstable when excavated during construction, but that the structures would be brought into a highly stable condition after construction, even to withstand earthquakes. There has never previously been a collapse of an underground structure with a large section such as Daikai Station. The resistance of underground structures to earthquakes was revealed to be a problem.

The deformation of under-ground structures in earthquakes is governed by the displacement of surrounding ground Therefore, "the response displacement method," a method that focuses on the displacement of the ground, is used for designing the earthquake resistance of underground structures. In this method the deformation and stress of underground structures are calculated by subjecting structures to ground displacements through ground springs. Therefore, the response displacement method may be regarded as a method using ground displacement.

In contrast, the deformation of aboveground structures such as buildings and bridges during earthquakes is dominated by inertial forces acting on the structures, and acceleration in earthquakes becomes the key factor governing the deformation of the structures. The currently used seismic coefficient method, modified seismic coefficient method, and dynamic analytical method fall within the category of accelerated earthquake motion methods.

Demarcated by the ground surface, earthquake-resistant design above and below the ground surface is thus based on two completely different concepts. In the case of underground structures at a shallow depth, such as the collapsed Daikai Station, however, it is questionable whether the above-mentioned concepts should be applied. That is, there is a possibility that a shallow underground structure can vibrate independently of the surrounding ground. In such a case, the structure behaves similarly to an aboveground structure, and inertia forces, or accelerations, become a dominant factor. In addition, since shallow underground structure is subject to earthquake inertial forces, with large amounts of soil placed on the roof of the structure, extremely large external forces act on the structure.

Another typical characteristic of damage caused by the Kobe earthquake is large displacement of quay walls. Numerous collapses of revetments and quay walls have been reported in past earthquakes, but most of them had not been designed against earthquakes and had been decaying. This was the first time when recently constructed quay walls were displaced, by a few meters. All the damaged quay walls had been constructed using concrete caissons. The cause of large displacement of the caisson revetments was investigated by members, with the Port and Harbor Research Institute of the Ministry of Trans-port as a leader. Soft cohesive soil had been replaced with Masado to install caissons on the seabed. The primary cause of the damage is now considered to be that the replaced soil had partially liquefied during the earthquake.

It should be noted that all the so-called earthquake resistant quay walls survived, though many quay walls were badly damaged. The construction of earthquake-resistant quay walls has been promoted nationwide, mainly in major ports and harbors, based on lessons learned from the damage to quay walls in Akita Harbor during the 1983 Nihonkai-Chubu earthquake. The earthquake-resistant quay walls, whose seismic coefficients used in their earthquake-resistant design were larger than for conventional quay walls, were designed to withstand liquefaction.

Damage to RC elevated railway bridge piers of the Shinkansen (bullet trains) shocked not only civil engineers, but also the general public. RC bridge piers were shear-fractured and collapsed, and girders fell. Fortunately, because the earthquake struck before service hours, loss of life was minimized. A serious issue has surfaced of how to assure the safety of moving trains, including Shinkansen, against earthquakes directly below them.

What we have to focus attention on is the previously mentioned lateral flow of the ground due to liquefaction. Resulting from the large displacement of revetments, the ground behind the revetments moved as much as a few meters towards the sea. The lateral ground flow damaged the foundation piles of buildings and plant facilities. Strains in the ground due to large displacement ruptured buried pipes in many places. As previously stated, although lateral ground flow had been recognized at the time of the 1983 Nihonkai-Chubu earthquake, the Kobe earthquake struck before sufficient results could be obtained from the research on this subject to put it to practical use. In the current earthquake-resistant design standards, no effect of lateral ground flow is taken into account. The problem to be resolved is how to incorporate the effects of lateral flow into the earthquake-resistant design of underground structures such as buried pipes and foundations.

Thinking back upon the damage caused by the earthquake, conditional factors which mitigated the disaster should be remembered. Shinkansen, as mentioned earlier, is one example.

Had the earthquake struck one hour later, bullet trains traveling on elevated railway bridges would have run off the rails, fallen onto private housings, and caused disastrous train accidents. The same can be said of the collapse of subway stations. Concrete ceilings

along with their overburden collapsed onto subway tracks. If trains had been stopped there or had smashed into the collapsed sections, additional serious damage would have resulted.

Due to large-scale liquefaction of reclaimed land in coastal regions, many storage tanks for oil, petroleum products, and high pressure gas were greatly tilted. However no tank collapsed. This is considered due to the relatively short time period over which earthquake motion continued. Despite strong tremors specific to inland earthquakes, a fortunate characteristic of the Kobe earthquake was the short duration of earthquake motion. If the earthquake motion had continued one minute longer, many of the tanks would have collapsed, stored fluids and gas would have flowed out of the tanks into the sea, and these would have caused secondary disasters, such as fires.

There were other factors that contributed to lessening the secondary damage. One important factor was that the earthquake struck early in the morning. If the earthquake had struck a few hours later during the rush hour, the results would have been much more tragic. Another factor was that dawn broke over the disaster-stricken area after the earthquake. The daylight aided the evacuation of victims and the rescue of people trapped under collapsed houses. If the earthquake had struck at midnight, the death toll would have been much greater.

Although it is quite natural to investigate in depth the damage done by the earthquake from both the viewpoints of structural construction and design standards concepts, etc., and to apply the results of the investigation in future preventive measures against earthquakes, we should also keep in mind the above-mentioned factors which mitigated the disaster.

PROPOSITIONS BY THE JSCE'S SPECIAL TASK COMMITTEE

(1)Level II earthquake motions

The JSCE organized a Special Task Committee of Earthquake Resistance of Civil Engineering Structures in March 1995, about two months after the earthquake, to discuss various subjects, such as what an earthquake-resistant structure should be in the future. The committee members first discussed whether the strong earthquake motions that had occurred in Kobe should be taken into account in the future earthquake-resistant design of civil engineering structures. According to researchers on active faults, the fault that caused the earthquake would becomes active every 1,000 to 2,000 years. Assuming that the fault becomes active in a cycle of 1,000 years and the useful service life of civil engineering structures is about 50 years, a probability that the structures would undergo such strong earthquake motions as those observed in the earthquake during the serviceable life is only five percent. The subject of discussions by the committee was how to tackle great disasters with such low frequency of occurrence.

The committee came to the conclusion that the resistance of civil engineering structures to earthquakes should be checked for such strong earthquake motions as observed during the Kobe earthquake which was caused by an inland fault, in addition to the earthquake motions that have thus far been used for earthquake-resistant design. These two earthquake motions are respectively called Level II and Level I motions. There are various reasons for this conclusion.

One is that the occurrence of inland earthquakes of a magnitude of 7 and larger is not

uncommon in Japan. Asterisked in Table 1 are inland earthquakes of a magnitude of 7 and larger that have done damage over the last 100 years. As is clear from the table, earthquakes of a magnitude struck six times in about 100 years, indicating a mean return period of 15 to 20 years. It should be noted that inland earthquakes resulted in many deaths, indicating the severity of damage such earthquakes can cause.

One of the reasons the committee reached the above conclusion was that inland earthquakes of magnitude 7 and larger, although extremely infrequent in a specific area, have a probability of occurrence that cannot be neglected in terms of earthquake-resistant design when viewed nationwide and can cause considerable damage when they do occur.

The committee proposed that application of Level II earthquake motion in the earthquake-resistant design of structures should be determined based on the results of studies on active faults in the neighborhood of a construction site.

However, there seems to be few cases where the existence of a fault can be defined clearly and the fault parameters such as the size of a fault section and dip angle are necessary to decide earthquake motions. Since many big cities in Japan are located on thick alluvial and diluvial deposits, it is generally very difficult to ascertain whether a fault exists deep under the cities. In such a case, the committee proposed applying a nationwide standard Level II earthquake motion which is evaluated from data recorded during past earthquakes. However, considering the present situation in which only a few observation records on inland earthquakes are available, further investigation and research are required. A study is now underway on this subject by a joint workshop organized by members from the JSCE and the Architectural Institute of Japan.

1. Structural resistance to Level II earthquake motions

It was proposed that the large earthquake motions observed in Kobe during the earthquake should be taken into account in the earthquake-resistant design of structures. However, this proposition does not mean that all structures should be designed and constructed to sustain Level II earthquake motions. As shown in Fig. 1, the proposition states that the earthquake-resistant capability of a structure should be determined by comparing the importance of the structure with the probability of occurrence of earthquake motion acting on the structures. For instance, against earthquake motions having a probability of occurrence once or twice during the service life of structures, e.g. Level I earthquake motions, the earthquake-resistant design should stipulate that the deformation of structure falls within an elastic limit and that residual deformation does not remain after an earthquake. In contrast to this, against very rare earthquake motions, e.g. Level II earthquake motions, the earthquake-resistant design should change the target earthquake resistance capability according to the importance of the structure: for example, the structure may be fractured, or the structure will have an allowable residual deformation limit, or the structure will not be damaged.

The significance level of a structure is determined by (1) the effects of the structure on human life and survival if fractured, (2) the effects on rescue and ambulance operations and restoration activities immediately after earthquakes, (3) the effects on civic life after earthquakes, (4) the effects on economic activities, and (5) the difficulty in reconstruction work. For instance, buildings in which many people gather such as department stores, movie theaters, and city halls are permitted to be damaged to some degree but should not collapse from the viewpoint of Item (1). Local government office buildings in which emergency headquarters are set up immediately after earthquakes, hospitals, police stations and emergency transportation roads should maintain their functions after earthquakes from the view-point of Item (2). In contrast, for roads and waterfront facilities which are not associated directly with the transportation of personnel and materials urgently needed for relief operations and lifeline networks which may be substituted by others, earthquake-resistant capabilities may be lower from the viewpoints of Items (3) and (4).

The method by which two types of earthquake motion, i.e. Levels I and II, are taken into account in earthquake-resistant design is called the two-level earthquake-resistant design method. The method has been adopted for the design of nuclear power plants and RC bridges. One of the committee's propositions is to, in the future, extend the application of the method into the design of general civil engineering structures.

However, many technical problems remain. One is the method of checking the critical state of steel structures. As mentioned previously, the earthquake-resistant design of steel structures has been generally made by the allowable stress design method. That is, the design is made, not in a plastic region beyond an elastic region. Little research has been done on the characteristic behaviors of steel structures in plastic regions. The same can be said of the critical state of buried steel pipes. If large displacement of the ground due to lateral ground flow is incorporated into the earthquake-resistant design of buried pipes, stresses of the pipes will reach a plastic region. But only little data has been collected on the deformation characteristics and critical conditions of buried pipes in a plastic region.

Further, the critical conditions of the so-called earth structures, e.g. embankments, revetments, retaining walls, and quay walls, are difficult to check against Level II earthquake motions. These have usually been designed based on the balance of forces, and the resistance of the structures has been assured by checking that the resistance of the structures is greater than external forces such as earthquake and dead loads by a certain ratio (a safety factor). This design method has not considered that the external loads exceed the resistance. However, if strong earthquake motions, such as Level II earthquake motions, are taken into account in the design, the external loads in many cases will exceed the resistance, the forces will be out of balance, and residual deformation could remain in the structures. Therefore, it is important to correctly estimate the residual deformation of the structures after forces become out of balance. Numerical analyses by the finite element method and model experiments are available but insufficient in terms of ease to use in practical design.

1. Earthquake-resistant reinforce ment of existing structures

Although the future earthquake-resistant design of civil engineering structures will be based on the concept described above, an additional problem that needs to be addressed is the earthquake-resistant reinforcement of existing structures. In large Japanese cities, such as Tokyo and Osaka, there are countless civil engineering structures similar to those damaged in the Hansin District by the 1995 earthquake. Some of them, e.g. highway roads, Shinkansen lines, subways, and revetments, were constructed earlier or have decayed more than those damaged in the Hanshin district. The earthquake-resistant reinforcement of these structures becomes an inevitable problem if disaster preventive measures are taken by predicting that earthquakes of a similar scale of the earthquake will hit these cities. Considering the incredible number of structures requiring reinforcement, as well as the fact that it would be necessary to reinforce them while they are still in service, it can be quickly understood how difficult it would be to reinforce these structures. As is clear from the damage caused by the Kobe earthquake, most critical and urgent is the earthquake-resistant reinforcement of reclaimed lands, where no soil improvement has been taken against liquefaction, and a huge number of buildings, bridges, and lifeline facilities exist. Therefore, it is necessary to quickly develop technologies and methods of diagnosing the earthquake resistance of existing structures and reinforcing them.

In addition, because reinforcement would be undertaken in stages, it is also necessary to develop a basic idea to decide the priority of reinforcement. The previously mentioned significance level of structures may be referred to in deciding the priority of the reinforcement. That is, the effects of structures on human life and survival and on rescue and ambulance operations and restoration activities immediately after earthquake, as well as other effects, need to be taken into account in deciding the order in which the earthquake-resistant reinforcement is to be performed. Moreover, consensus on the costs necessary for the reinforcement and how to share these costs needs to be reached based on extensive discussion.

LIMITS OF EARTHQUAKE RESISTANCE OF STRUCTURES AND ACCEPTABLE RISK LEVELS

The current earthquake resistance standards for various structures and facilities are now being reviewed and revised. An extremely important point in updating the current earthquake-resistant

standards is that we engineers must understand that the earthquake resistance and safety of structures have thresholds, and that we have to give people a clear explanation of this. It is of importance to give a quantitative explanation of the extent to which structures and facilities are able to withstand earthquakes and to disclose the limit of assurance.

A wrong impression has been given so far that structures and facilities are safe if they are designed and constructed in conformance with design standards set forth by governmental offices having control over the structures and facilities, such as the ministries of Construction and International Trade and Industry. Needless to say, unlimited safety is impossible when dealing with natural phenomena such as earthquakes.

A revision to the current design standards will incorporate Level II earthquake motions, earthquake motions which are much stronger than before. However, there is still a possibility that a stronger earthquake motion than the design Level II motions will occur in the future. In setting Level II earthquake motions, it is necessary to show quantitatively the degree of earthquake resistance achieved by the newly set values and at the same time show what the probability is of exceeding the set value.

This applies without exception to the construction of nuclear power plants. Those concerned have often repeated explanations, giving an impression that the plants are completely safe; yet each time minor or moderate accidents have occurred, such explanations have only raised suspicions and lowered public confidence.

The earthquake-resistant design of structures has been explained in the same way. Every time earthquake damage was reported from overseas, a groundless explanation was given through the mass media that structures in Japan will be safer because of stricter earthquake-resistant requirements than those overseas. It is undeniable that a wrong impression about the earthquake resistance of structures has been given to the public. An explanation should be given that structures have a limitation of earthquake resistance and that they might collapse in future earthquakes.

It is important to make efforts to form a consensus on acceptable risk levels with the public. The acceptable risk levels are decided by comparing the expected values of benefits that people usually enjoy with the expected values of losses once a disaster occurs. Comparing human lives with property, i.e. values of a different nature, is an essentially difficult problem.

The Shinkansen is a good example in thinking about acceptable risk levels. As described earlier, if a bullet train is struck by inland earthquake like the Kobe earthquake, the train can run off the rails and fall from an elevated railway bridge. However, it should be remembered that the bullet trains play a big part in economic activities in Japan and offer great convenience to the nation. The argument does not hold that the bullet trains should be removed from service unless 100% safety is assured. The public accepts this risk by a tacit consent. The same can be said of car accidents resulting in more than 10,000 deaths annually. The probability of annual per capita death by car accident is 10^-4, far higher than an estimated accident probability at nuclear power plants of 10^-6, but in this age of motorcars no one thinks that automobiles should be banned. It can also be considered that some sort of consensus is formed tacitly among the public with regard to acceptable risks posed by motorcars.

As stated above, it is necessary to make efforts to obtain national consensus about acceptable risk levels by clearly and quantitatively showing the limitations in the earthquake resistance of structures before updating the current design standards.

CONCLUSIONS

here are, as of now, two subjects we must confront to avoid a repeat of the disastrous damage caused by the Kobe earthquake. The first subject is to thoroughly record the actual damage conditions and pass this data on to future generations. The occurrence frequency of damage as seen from the Kobe and Great Kanto earthquakes seems to be once every tens of years. Among the problems posed by the damage are some that cannot be solved by the present generation, as well as many problems where work on the solution must pass from generation to generation. From this point of view, what we should first do is to leave clear records of the damage.

The second subject is to openly discuss with researchers and engineers in many fields the causes of damage to structures and facilities and the knowledge about the earthquake derived from accurate damage data, and to chart correct courses of what the earthquake resistance of civil engineering structures and preventive measures against earthquake disasters should be in the future.

The Northridge earthquake occurred in the U.S. on January 17, 1994, exactly one year before the Kobe earthquake, causing highways to collapse and severe damage to houses and buildings. Besides the damage caused by the Northridge and Kobe earthquakes, a great deal of damage is done by quakes that take place roughly once a year somewhere in the world. There are at times earthquakes that take a heavier toll of lives than the Kobe earthquake. To lessen earthquake damage and create a society safe from earthquakes is a duty charged to all persons concerned in this field, not only in Japan but throughout the world. Engineers and researchers in the world must team up to eliminate earthquake damage by further advancing international cooperation, including activities through the United Nations International Decade for Natural Disaster Reduction which began in 1990.