EARTHQUAKE EARLY WARNING Perspectives for developing countries Paolo Gasparini AMRA Scarl and Dept. Physical Sciences, University of Naples Federico II Gaetano Manfredi AMRA Scarl and Dept. Structural Engineering, University of Naples Federico II
The Disaster Reduction Chain Early warning is an important component! Process Understanding Hazard Assessment Risk Assessment Safe Buildings Infrastruct. Scientists Scientists Scientists, Engineers Engineers Knowledge Transfer Early Warning Rapid Resp. and Loss Assessment Task force Decision Makers, Stake Holders Scientists, Engineers
Geohazards: A rapid disaster information system consists of the following components: The Early Warning refers to actions it can be taken after the threating event has started but before it strikes at the target (the available time window is the «lead time»). The Prediction is the capacity to deterministically evaluate time, location and dimension of a coming event. Not feasible. The Forecasting is the capacity to probabilistically evaluate time, location and dimension of a coming event. The Rapid Response Support corresponds to the information given within the minutes after the event has hit the target. It will support the response actions. The informations generated by an EW System can be used for this purpose.
The Urban Explosion From It takes 8 Megacities place in the in 1950 developing to 60 world! 2015! million Number of people in cities 1950 2015 418 400 Risk mitigation in mega-cities347require 300 preventive long term actions. risk is a highly dynamic quantity They 200 are not enough and are too costly. Mega-cities are becoming Hot Spots of Risk Real 100time risk mitigation by early 71warning systems is a budget friendly need. 1950 1970 1990 2015 Industrialized countries Developing countries All Source: National Geographic, Nov. P. Gasparini, 2002 G. Manfredi - EARTHQUAKE EARLY WARNING, Washington, D.C. - January 2009
TYPICAL LEAD TIMES MOST SIGNIFICANT ACTIONS EARTHQUAKES: seconds to tens of seconds TSUNAMIS: minutes to hours METEROLOGICAL EVENTS: hours to days FLOODS AND LANDSLIDES: hours to days VOLCANIC ERUPTIONS: hours to weeks AUTOMATIC ALERT + INFORMATION ALERT + INFORMATION COPING CAPACITY COPING CAPACITY
Principles of Earthquake Early Warning (EEW) Energy Containing Waves (S) are slower than Information Containing Waves (P) Time (s) 20 S P 10 50 100 Distance (Km)
Principles of Earthquake Early Warning A seismic sensor detects an earthquake P-wave close to its source. A Warning Signal can be transmitted. It reaches the target with a lead time of a few seconds to a few minutes before the arrival of destructive waves. Earthquake Seismic sensor Destructive S waves (3.5 km/s) Transmitted information of seismic wave arrival (no delay) Target site
Possible actions with a few seconds of lead time 5 s 10 s 15 s 20-60 s Alert schools and workmen in dangerous situations Switch off lab, industrial equipments Stop traffic (fast railways, bridges) Disactivate hazardous industries
Components of an EEW System 1. A seismic sensor network with real-time capabilities 2. 3. A not-satellite based system to transmit information to the target site A local or central data processing system 4. An interface/device with the target to be protected
Configurations of Early Warning Systems REGIONAL: sensors around the earthquake source transmit signals to the target sites One network for all targets; Longer lead times; high cost BARRIER: seismic sensors are located between the seismic source and the target site One network for many targets; Intermediate lead times; high cost ON SITE: sensors are located very near or on the target site One system for each target, very short lead times; low cost, easy maintenance
A typical REGIONAL EEW network
Typical BARRIER system: Earthquake sources are located offshore
ON SITE Early Warning Systems The Uredas system in Japan is positioned along the Shinkansen railway line. It can predict the magnitude of and distance from a potentially destructive event starting from an analysis of the first seconds of the P wave.
ON SITE Early Warning Systems The Ignalina Nuclear Power Plant In the Ignalina NPP the EEWS is made up of a number of stations circling around the plant (Wieland et al. 2000). Each station is made up of three sensors triggering at 0.025g. A 2-out-of- 3 logic is used to determine if a real seismic event has occurred. 30km Then an alarm is generated in the reactor control room. Two seconds are needed to reduce the reactor capacity and prevent a meltdown during a severe accident. 30km The false alarm issue is reduced by the redundancy of measurement at the same geographic location.
Exisiting Regional Earthquake Early Warning Systems Italy Romania Turkey Taiwan Japan USA Mexico
FP6-Global-4 SAFER Seismic early warning For EuRope STREP Project One of Safer Project objectives is the reduction of seismic risk in: Athens Bucharest Cairo Istanbul Naples Lead partners: GFZ Postdam (Germany) AMRA Scarl (Italy) NOA Athens (Greece) 23 Partners from 15 countries Supported by: Eureopean Commission
ERGO the Campanian EEW
The most tested EEWS is implemented in Japan. It started to protect Shinkaze trains I. Alarm seismometers installed along all rail tracks shut off power when horizontal acceleration exceeds a threshold Front detection: deployed along coast gives ~15 sec warning II. UrEDAS event parameters determined from the P-arrival
An EEW Amendment of Weather Service Law came into force on the 1st of Dec. 2007 Points 1) JMA MUST issue EEW in case it is necessary. 2) Designated organization MUST transmit EEW to the relevant organizations and public. 3) Those who start provision of EEW to individual house and building, need to satisfy Technical Standard determined by JMA.
Conceptual Image of Seismic Wave Propagation and Earthquake Early Warning (EEW) 0 th issuance of EEW rapidity 1 st issuance of EEW 2 nd issuance of EEW N th issuance of EEW accuracy
FALSE ALARMS In 3 years 1713 EEWs were issued based on information from one single station 30 EEWs (1.75%) were false alarms issued by various reasons Only 7 were for M>4.5 No false alarm was issued when two or more stations data were processed. No missed alarmas are reported Measures against false alarm can minimize possibility of erroneous alarm. (improvement of hardware, software and operation manual) Cause of errors Initial defective device & Mis-operation Defective device and noise signal triggered EEW Estimated maximum seismic intensity 5-5+ 6-6+ 7 Total 3 0 1 0 0 4 3 0 0 0 0 3 Total 7
Early Warning in Seismic Risk Reduction Pre-Event Tens of seconds engineering: Limiting exposure; Specific actions and Protection Systems; Increasing resilience; Reducing vulnerability. Post-Event: Emergency management (real time shake maps, expected damage maps); Structural intervention directions.
EWS Time-Scale minutes Shake maps, expected damage Evacuation of Buildings Time Shut-down of critical systems Activation of Structural Control Systems Stop of High-Speed trains seconds Actions Impact
Design Targets Lead Time High Perception Impact (e.g. Lifelines Interruption ) Medium Perception Impact (e.g.trasportation( Interruption Low Perception Impact (e.g. Elevetor) False Alarm Probability Performances/ Consequences
LEAD TIME MAP AND POSSIBLE ACTIONS
Structural Prospectives for EWS EWS can be an indirect protection system in the case of structures which cannot be retrofitted; EWS can integrate structural monitoring systems; EWS can be complementary to other protection systems for critical structures.
EEW-Structure Interaction Source Processing Decision/Action Segnale P Predizione dell intensità al sito e probabilità di superamento EEW Time Tp Ta Te Tc Detection Forecasting Waring arrival Earthquake Time Tuning To Origin dell Evento Td StrikeSismico
Near-real time shake and damage maps
Engineering Requirements of EEWS Quantitative real-time assessment of seismic risk (losses for specific application) Time dependent decision making (quantification of trade-off between lead-time and costs of missed/false alarms) Automated decision/control structural system Consequence-based approach
Real-Time loss assessment Extending the hazard approach it is possible to determine the expected losses conditioned to the measurements of the seismic network in the case of alarming or not [ ], ( τ) LDEDPIMMR ( ) ( ) ( ) ( ) ( τ) E L l f l d g f d edp f edp im f im g dlddd EDPd IM = Expected Economic Loss 1.Loss probability depending on the alarming decision 2. Structural damage probability depending on building s seismic response 3. Bulding Seismic response probability depending on ground acceleration 4. Real-time Ground acceleration estimate
When to activate security measures? A possible decisional rule is to alarm when the expected value of the PGA is larger than a given threshold From real-time hazard analysis Alarm if E [ PGA]> PGAC Another possible decisional rule is to alarm when the probability, that the PGA will be larger than a given threshold, exceeds a specific value (Pc). PGA C 0 ( ) Alarm if P [ PGA > PGA ] = 1 f PGA dpga > P C From real-time hazard analysis C
What is a false/missed alarm? { True C} { } MA: noalarm PGA > PGA FA: Alarm PGA True PGA C These probabilities may be computed as a function of time and depend also on the decisional rule adopted.
Expected Economic loss vs. Ground Shaking Intensity Expected Loss [ ] No alarm Alarm Ground Shaking Intensity Iervolino et al., 2006, Expected loss-based alarm threshold set for Earthquake Early Warning Systems,EESD, (modified) Optimal Alarm threshold
Low cost EEW systems Development of low cost sensors (some hundreds $ each) and interfaces; In situ systems (total cost of the order of 1,000 $ each) for simple actions; Barrier configurations using shared transmission systems; Much cheaper than retrofitting large building stocks!
Benefits EEWS are a powerful tool to prevent triggered events from human activity (city fires, industrial accidents, toxic substance escape ) EEWS can be used for warning areas of potential impact of cascade events (Landslides, tsunami, after-shocks ) EEWS can help for evacuating severely damaged buidings before they collapse EEWS can be used to keep strategic buildings in operation
Rapid Impact Assessment Ground Shaking Intensity Can be Overlayed on Infrastructure! Shake-Maps within a few minutes! (The Northridge Scenario) Ground Shaking Intensity The shake-map-technology is now implemented in many places in Europe! and real-time damage assessment is feasible, but needs more testing (GEM?)
The effective implementation of EEWS requires: A detailed study of information diffusion at every level; An extensive program of education of people, administrators and civil defense officials.
People Centered Early Warning System Self Organizing Mesh Networks for Early Warning, Rapid Response and Added Value for People Mesh node Decentralized versus centralized systems A New Technology People centered versus one voice System components: Standard equipment and own development Basic Features dense! self-organizing! decentralized! low-cost! multi-purpose! Client station
A case of missed alarm: The Madrid snow storm of January 9, 2009 45.000 people was trapped at Airports and 400 km of vehicles were trapped on Highways because authorities waited to give warning untill the false alarm probability was very low. Too Late to reach people timely!
A CASE WHERE NO EERW WAS GIVEN Smoke from the KASHIWAZAKI (NW Japan) nuclear plant struck by the M6.8 July 17,2007 earthquake. GA vas higher that design earthquake. No proper EEW was active. More than 50 problems were produced by the earhquake.
THANK YOU FOR YOUR ATTENTION CONTACTS: Paolo Gasparini paolo.gasparini@na.infn.it Gaetano Manfredi gamanfre@unina.it Amra Scarl 11, Nuova Agnano Napoli 80125 Ph. +39 081 7685124/25 fax. +39 0817685144 www.amracenter.com info@amracenter.com