2009/04/22 Earthquake Early Warning and Realtime Disaster Prevention Yutaka NAKAMURA, Dr. Eng. System and Data Research, Co., Ltd. Visiting Professor, Tokyo Institute of Technology System and Data Research Co., Ltd. 1 Tokyo Institute of Technology
2009/04/22 Videos Recorded at the Moment of Earthquake Attack System and Data Research Co., Ltd. 2 Tokyo Institute of Technology
2009/04/22 EEW and Earthquake Disaster Mitigation � The Basic Countermeasure is Strengthening the Facilities � EEW is only a Trigger for Quick Response against Quake � It is important for EEW to avoid Overestimation � Late EEW is Unnecessary � Accurate Information is Extremely Important for Quick Response after Quake Damage of the 2009 L’Aquila Earthquake (Mw 6.3) System and Data Research Co., Ltd. 3 Tokyo Institute of Technology
2009/04/22 Concept of Earthquake Early Warning There are two kinds of the On- On -Site Site earthquake alarm. One is Detection/Alarm Detection/Alarm “On-site Alarm” which is the Front Front alarm based on the Detection/Alarm Detection/Alarm observation at the side of the objects to be warned. The other is “Front Alarm” P which is the alarm based on P-wave Alarm S-wave Alarm S the observation near the epicentral area to warn for Fault Rupture the possible damaged area. For each, there are more two kinds of alarm. One is so-called “S-wave Alarm” or “Triggered Alarm”. And the other is “P-wave Alarm”. We have developed a prototype system for EEW as UrEDAS in early 1980’s. System and Data Research Co., Ltd. 4 Tokyo Institute of Technology
2009/04/22 Introduction of UrEDAS UrEDAS, Urgent Earthquake Detection and Alarm System, is the first real time P-wave alarm system over the world in practical use in 1992 for Tokaido Shinkansen. It is characterized to be able to process digitized waveform step by step without storing waveform. Amount of procedure is not differ from each other either earthquake occurs or not, so it expected not to be occurred the system down due to the over load. UrEDAS is able to use not only for the On-site alarm but also for the Front alarm. System and Data Research Co., Ltd. 5 Tokyo Institute of Technology
2009/04/22 Functions of the UrEDAS There are two types of UrEDAS; “ UrEDAS ” and “ Compact UrEDAS ”. Function of the UrEDAS (1985) is to estimate the magnitude and the location of detected earthquake in three seconds after initial P-wave detection and issuing the alarm for expected damage area. On the other hand, Compact UrEDAS (1998) can evaluate whether the earthquake will be destructive or not using Destructive Intensity DI and issues alarm one second after P-wave detection if needed. System and Data Research Co., Ltd. 6 Tokyo Institute of Technology
2009/04/22 The UrEDAS Technique for Estimation Methods of Location, Depth and Magnitude using a Single Station Data in Realtime The development of UrEDAS had been almost completed in the middle of 1980’s. UrEDAS realized realtime independent process. Although the JMA system intermittingly processes with several seconds, UrEDAS processes in every sampling time. Warning time of UrEDAS can set arbitrary. At first the warning time set to three seconds, but we have found the time is able to set one second without problems at least for M7 class earthquake. For New generation of UrEDAS, FREQL, the warning time is set to one second. I would like to show the potentiality of them instead of explanation the UrEDAS techniques in detail. System and Data Research Co., Ltd. 7 Tokyo Institute of Technology
2009/04/22 UrEDAS is only true realtime system UrEDAS can estimate Epicentral Azimuth, P-wave Incident Angle and etc. in realtime I would like to show the potentiality of UrEDAS; Rupture Trace for Real Event in realtime using one UrEDAS station System and Data Research Co., Ltd. 8 Tokyo Institute of Technology
2009/04/22 UrEDAS Application to Rupture Process Estimation in Realtime Pasadena Pasadena Pasadena Pasadena In case of the 1994 Northridge Eq. (after Nakamura 2001) This is an example of M<3 M<4 M<5 M<6 M<3 M<4 M<5 M<6 M<3 M<4 M<5 M<6 M<3 M<4 M<5 M<6 rupture process tracing in realtime using the data of Pasadena UrEDAS station. The plane between back U U r gent Ear t hquake D r gent Ear t hquake D et ect i on & A et ect i on & A l ar m S l ar m S yst em yst em azimuth and incident angle is correspond to left upper cross- section. Rupture can be seen to go to Pasadena surface from hypocenter. System and Data Research Co., Ltd. 9 Tokyo Institute of Technology
2009/04/22 Estimation Destructivity and Warning Methods of Compact UrEDAS System and Data Research Co., Ltd. 10 Tokyo Institute of Technology
2009/04/22 Captured Photos at the time of the 1995 Kobe Earthquake Captured Photos at the time of the 1995 Kobe Earthquake P wave arrival Time Principal motion arrival Motivation of Compact UrEDAS development is the Kobe Earthquake. On the VTR, they noticed the initial P wave motion as something happen, and then the severe motion attacked them after a few seconds. Although there was only a few seconds between something happen and recognition of earthquake, it was anxiousness and fearful because they could not understand what happened and felt relieved after recognition of the earthquake occurrence. As the counter of this kind of feeling, the earlier earthquake alarm is required and I developed the Compact UrEDAS to make the alarm within one second after P wave arrival. System and Data Research Co., Ltd. 11 Tokyo Institute of Technology
The P wave alarm of Compact UrEDAS demonstrates 2009/04/22 the effectiveness as making the derailment not catastrophe 0 1 2 3 4 5 6 7 8 9 10 11 12 13 0 1 2 3 4 5 6 7 8 9 10 11 12 13 P -wave Arrival P -wave Arrival P-wave Alarm P-wave Alarm S -wave Arrival S -wave Arrival Principal Motion Principal Motion 5 seconds Main Derailments Main Derailments Vehicle ② : Successive You can see derailment situation Time Time Example of and contact situation between body Derailment Derailment Location Location and railroad, Compact UrEDAS Final Derailment Situation Final Situation normal vehicle normal vehicle Warning derailment vehicle by derailment vehicle by flange climbing flange climbing large relative displacement large relative displacement railway deformation railway deformation Front ⑩ Rear ① (a)先頭車 (b)最後尾車 Tunnel Exit System and Data Research Co., Ltd. 12 Tokyo Institute of Technology
2009/04/22 Ultimate Earthquake Early Warning System FREQL series and AcCo - PS System and Data Research Co., Ltd. 13 Tokyo Institute of Technology
2009/04/22 FREQL (F ast Re spo nse E quipme nt against Quake L o ad) FREQL is developed for the earthquake warning system based on the experiences of development and operation of the world first P wave alarm system UrEDAS. FREQL function is combined the functions of UrEDAS, Compact UrEDAS and AcCo. P wave alarm is available 0.2 seconds in minimum after P wave detection ( the fastest time will be 0.1 seconds in 2009) S wave alarm is also available. (based on acceleration and real-time seismic intensity RI.) System and Data Research Co., Ltd. 14 Tokyo Institute of Technology
2009/04/22 Change of processing time for EEW This figure shows the 10 10 Processing Time for P-wave Alarm in sec. Processing Time for P-wave Alarm in sec. change of the cf. Equipment developed by JMA cf. Equipment developed by JMA SDR products SDR products processing time for 5 5 Average 5.4 sec. Average 5.4 sec. UrEDAS 3 sec. UrEDAS 3 sec. EEW. 3 3 UrEDAS UrEDAS JR Tokaido Shinkansen JR Tokaido Shinkansen 警報処理時間 （秒） 警報処理時間 （秒） Minimum 2.0 sec. Minimum 2.0 sec. Test observation Test observation Tsunami warning system Tsunami warning system 2 2 While JMA system for Wakayama Prefecture for Wakayama Prefecture performs every one 1 1 Compact UrEDAS 1 sec. Compact UrEDAS 1 sec. second for the alarm Tohoku Shinkansen Tohoku Shinkansen Tokyo Metro Tokyo Metro 0.5 0.5 processing Min. Min. intermittently with FREQL Minimum 0.2 sec. FREQL Minimum 0.2 sec. 0.1 0.1 0.2 0.2 stored data, UrEDAS sec. sec. Tokyo Metro, Odakyu Tokyo Metro, Odakyu Fire Depart ments of Tokyo, Osaka and so on Fire Depart ments of Tokyo, Osaka and so on and FREQL perform 0.1 0.1 the procedure 1980 1980 1985 1985 1990 1990 1995 1995 2000 2000 2005 2005 2010 2010 continuously in every Year Year Development of Processing Time Development of Processing Time sampling time. System and Data Research Co., Ltd. 15 Tokyo Institute of Technology
Recommend
More recommend
