Development of a New Early Earthquake Detection and Alarm System

Kimitoshi ASHIYA
Senior Researcher, Laboratory Head,
Earthquake Disaster Prevention,
Disaster Prevention Technology Div.

1. Introduction
    The Railway Technical Research Institute (RTRI) has developed an urgent earthquake detection and alarm system (UREDAS) mainly for Shinkansen and put it into practical use. Valuable knowledge is being accumulated to give an urgent alarm for earthquakes due to the development of the real-time seismology in recent years. The Japan@Meteorological AgencyiJMAj and other governmental organizations are improving the nationwide earthquake observation network and planning to distribute prompt earthquake information (Now-cast earthquake information). Under the circumstances, RTRI is studying new methods to predict seismic source parameters based on the P-wave and urgently estimate earthquake damage, and promoting the development of a new early earthquake detection and alarm system to utilize the prompt earthquake information provided by governmental organizations jointly with JMA.

2. New early earthquake detection and alarm system
2.1 New method to predict seismic source parameters

    The conventional system roughly estimates the magnitude and the epicenter based on the period and the maximum amplitude of the initial P-wave measured at a observation point, and issues an alarm before the arrival of the main shock. RTRI has discussed a method to predict seismic source parameters based on the P-wave by utilizing the latest knowledge in seismology and other scientific fields, and developed a new method to estimate the magnitude and the distance to the epicenter from the maximum amplitude and the amplitude increasing rate of the initial P-wave (Fig. 1).
    When compared with the conventional method, it estimates the magnitude at higher precision (Fig. 2) while less affected by the noise of ground motion.

Fig.1 New method to predict seismic
source parameters

Fig.2 Comparison of the precision of
the estimation of magnitude

Fig.3 Seismograph for the new system
2.2 Seismograph for the new system
    RTRI has developed a seismograph for the new system by applying the new method to detect seismic source parameters (Fig. 3), by assuming the application to the observation points of railway operators. Since it applies the latest information technology, it has several merits when compared with the conventional system, in addition to the aforementioned improved precision in the detection of source parameters. A built-in PC, for example, makes the seismograph compact and lightweight and enables parallel processing and remote operation with a real-time OS. Circuits are designed to incorporate countermeasures against electromagnetic noise. This makes it possible to use the seismograph at observation points in wayside substations.

2.3 Now-cast earthquake information
    JMA has a plan to measure the seismic waves near the epicenter at 180 observation points across the country and distribute the information on the epicenter, magnitude, main shock arrival time and predicted seismic intensity, which is called the Now-cast earthquake information, before the main shock arrives. The Now-cast information will be distributed when the P-wave has arrived at the observation point nearest the epicenter and repeatedly thereafter at certain time intervals. The information which is distributed first (the 0'th information) is on the seismic source parameters estimated from the data of the P-wave observed at an observation point for a few seconds. The new method mentioned above will be used to process this information. After that, more precise information (the 1st information, 2nd information and so on) will be distributed, as the seismic motion is observed at other observation points.
    The new system uses the Now-cast information (mainly the 0'th information) provided by JMA in addition to the information obtained by railways. It is thought, therefore, that the function of the UREDAS for Shinkansen will significantly be improved. It will also be possible to apply it to narrow-gauge lines at low costs.

3. Operation restart support system
    One of the most important roles of the railway earthquake disaster prevention system is to urgently detect earthquakes, control train operation and ensure safety. What is equally important from the viewpoint of stable transport is to immediately check the safety of facilities after an earthquake and promptly restart train operation when safety is ensured. In addition to the distribution of the above-mentioned Now-cast information, JMA plans to summarize the data obtained by the seismic intensity meters at about 3,000 points in the country, add the information on the ground thereto, calculate the seismic intensities in a wide area surrounding the epicenter divided into 1km square areas and provide the information on the two-dimensional distribution of seismic intensity thus established in about five minutes after an earthquake (Fig. 4). In the future, JMA is expected to distribute the information not only on the seismic intensity, but also on the maximum acceleration, maximum velocity and dominant period of seismic motion. RTRI is now developing an operation restart support system in the belief that it is possible to cut the down-time of train operation by selecting the sections where safety should be confirmed based on the two-dimensional earthquake information (Fig. 5).

Fig. 4 Example of the two-dimensional distribution of estimated seismic intensities (Geiyo earthquake)

Fig.5 Output image of the operation restart support system

4. Future plan of RTRI
    RTRI will construct a prototype compound system to combine the new early earthquake detection and alarm system and the post-earthquake operation restart support system on its premises by 2002 autumn for operation tests, based on which it will propose the specifications for a practical compound system at the end of fiscal 2002. In 2003 autumn, it will start a test of the normal operation of the compound system based on the now-cast information and two-dimensional information on the seismic intensity estimated by JMA.