(2)Shallow inland earthquakes

Sometimes, large earthquakes occur on land in the Japanese archipelago, such as the 1995 Southern Hyogo Prefecture Earthquake (M 7.2) that caused the Great Hanshin-Awaji Earthquake Disaster. In the underground area of the Japanese archipelago, there are generally powerful forces of compression exerted in an east-west or northwest-southeast direction which cause earthquakes of this type to occur. Past research has shown that with earthquakes occurring on land, rock mass with hardness sufficient to cause earthquakes extends to about 15-20 km underground. Temperatures are high at depths lower than that, so it is thought that the rock mass will not suddenly break up when pressure is exerted on it. Rather, fluid distortion will occur. Therefore, the hypocenter of large earthquakes that occur on land will be located less than 20 km below the surface. Most of these earthquakes will be at the M 7 level, and earthquakes at the M 8 level seldom occur, with an example being the 1891 Nobi Earthquake (M 8.0).

Because these earthquakes have a shallow hypocenter, in many cases, fault movement reaches the surface during large earthquakes of M 7 or greater. We have learned from topographical and geological surveys that slip caused by fault movement accumulate for at least several hundred thousand years on faults that have slipped the earth's surface. This signifies repeated slipping on an existing, roughly uniform weak plane (fault plane); in other words, several large earthquakes must have occurred. It also signifies that a roughly equivalent force has been exerted in the subterranean area of the Japanese archipelago for at least last several hundred thousand years. Therefore, it is believed that large earthquakes will continue to occur at faults such as these in the future as long as this force is applied. Faults at which this activity has repeatedly occurred in the past and at where it is likely to occur in the future are known as "active faults".

Active faults form characteristic topographical conditions through the accumulation of slip that accompanied earthquakes in the past (Fig.2-21, Fig.2-22). One such condition is the formation of topographical discontinuities at locations that had been uniform. These conditions can be used as clues to discover active faults and estimate their extent of activity. Surveys to locate active faults often make use of aerial photography to reveal the topographical conditions such as those seen in Fig.2-21. A look at the distribution of the primary active faults in and around the Japanese archipelago in Fig.2-23 shows there is a close relation between topography and the distribution of active faults. For example, active faults are often found at the boundary between mountains and lowlands.

A detailed survey of active faults reveals sharp differences in the proportion of accumulated slip among active faults. That is, the extent of activity for active faults differs greatly with each fault. There are three categories for this activity, termed A, B, and C class, ranging from the greatest to the least. The categories are based on the average volume of slip per 1,000 years. This is called the "mean slip rate". For the A class, which has the greatest activity, the average amount of slip over 1,000 years will be from 1-10 m. The activity is an important indicator when considering the interval between earthquakes on each active fault and the characteristics of regional seismic activity. For example, regions with many other active faults in the A class of activity have a higher incidence of large earthquakes than other regions.

Scientists also study the history of this activity by excavating these active faults (trenching) (Fig.2-24). These studies enable an evaluation to a certain degree of the next period and the magnitude of activity if the history of previous activity is clarified. This history includes such factors as the intervals of activity, the most recent period of activity, and the size of the fault where the activity occurred. It also may include reactions to earthquakes that are known of from historical accounts. The intervals between activity on active faults stretch to about 1,000 years at the shortest. Further, there is a certain degree of variations, so the accuracy of evaluating the next period of activity can only be in the range of several hundred years. Also, slip does not occur on the earth's surface during M 6 earthquakes, which sometimes can cause significant local destruction. Therefore it is difficult to understand earthquakes of this size through ordinary active fault surveys.

Repeated active fault movement in the Japanese archipelago creates plains and basins, where many cities have developed. Hence, large earthquakes sometimes occur directly below or very close to these areas. In addition, in some parts in these plains and basins, deposit carried by rivers can accumulate rapidly. However, visual examination alone of the topography in these areas is insufficient to determine the existence of active faults.

Strong ground motion immediately followed by tsunami occurs with earthquakes on active faults at the sea floor near coastal areas.

Earthquakes, which occur at deeper locations (50-200 km) in land areas, are classified as earthquakes that occur within the subducting Pacific Plate and the Philippine Sea Plate. [Refer to 2-4(1) 3)].