Toward EEW engineering applications: leadToward EEW ... · With an area of 13.595 km², Campania is...
1
EGU General Assembly 2008, Vienna, Austria, 13 - 18 April 2008 Toward EEW engineering applications: lead Toward EEW engineering applications: lead Toward EEW engineering applications: lead Toward EEW engineering applications: lead- - -time maps for the ISNet infrastructure in the Campanian region time maps for the ISNet infrastructure in the Campanian region time maps for the ISNet infrastructure in the Campanian region time maps for the ISNet infrastructure in the Campanian region Department of Structural Engineering, University of Naples Federico II, Italy (* [email protected]) Motivation The feasibility of Earthquake Early Warning Systems (EEWS) depends I) I) on the possibility of real-time estimation of the size of the event (source parameters) and the intensity of subsequent ground shaking at various sites (although with large associated uncertainty) and, at the same time, II) II) on the ability to design earthquake engineering applications for real-time seismic risk mitigation. Among the information needed to evaluate the effectiveness of EEWS, there is the estimation of the available time to perform security actions before the arrival of the more energetic seismic phase at the site of interest. This interval is termed lead lead - - time time. A rational approach to the feasibility analysis of EEWS in a densely urbanized area requires the assessment of likely and minimum-maximum range lead-times (in a probabilistic manner) for the events the seismic infrastructure can detect. The Irpinia Seismic Network (ISNet) With an area of 13.595 km², Campania is only the twelfth largest region in Italy, but a population of around 5.8 million people makes it the second most populous, and the most densely populated region in Italy. In Campanian region the most hazardous seismogenic source is close to, if not overlapped, with densely urbanized areas and time consuming security actions may be unfeasible. Anyway, historical earthquakes damaged wide areas (even larger than 100 km in radius) due to a combination of the geology (low rate of seismic attenuation in the region) and of the low quality of the buildings in relation to their resistance to seismic action. For this reason, Campanian region is now provided with an advanced and dense sensor network mainly devoted to EW purposes. The Irpinia Seismic Network (ISNet) covers an area of about 100 x 70 km 2 , corresponding to a part of the entire seismic source area [1]. ISNet comprises 29, 6 components seismic station equipped with both accelerometers and velocimeters, with real-time telemetry. Following a distributed approach, the network is organized in 6 sub-nets: waveform data is collected and elaborated at local hubs (LCC, Local Control Centers) which, in turn send processed parameters to a Network Control Center (NCC) in Naples. Lead-time maps Our study has yielded the following results, presented in the form of maps. We evaluated the minimum, maximum and average lead times (T 4 , T 18 and T 29 ) for each node of a regular grid having a ~ 2 km spacing and covering the whole Campanian territory with over 2700 nodes. 13.5 14 14.5 15 15.5 16 40 40.5 41 41.5 Longitude [°] Latitude [°] Minimum lead-time T 4 Blind zone 0 - 5 s 5 - 10 s 10 - 15 s 15 - 20 s 20 - 25 s ISNet stations 13.5 14 14.5 15 15.5 16 40 40.5 41 41.5 Longitude [°] Latitude [°] Medium lead-time T 4 0 - 5 s 5 - 10 s 10 - 15 s 15 - 20 s 20 - 25 s 25 - 30 s 30 - 35 s 35 - 40 s ISNet stations 13.5 14 14.5 15 15.5 16 40 40.5 41 41.5 Longitude [°] Latitude [°] Maximum lead-time T 4 5 - 10 s 10 - 15 s 15 - 20 s 20 - 25 s 25 - 30 s 30 - 35 s 35 - 40 s > 40 s ISNet stations 13.5 14 14.5 15 15.5 16 40 40.5 41 41.5 Longitude [°] Latitude [°] Minimum lead-time T 18 Blind zone 0 - 5 s 5 - 10 s 10 - 15 s 15 - 20 s ISNet stations 13.5 14 14.5 15 15.5 16 40 40.5 41 41.5 Longitude [°] Latitude [°] Medium lead-time T 18 Blind zone 0 - 5 s 5 - 10 s 10 - 15 s 15 - 20 s 20 - 25 s 25 - 30 s 30 - 35 s ISNet stations 13.5 14 14.5 15 15.5 16 40 40.5 41 41.5 Longitude [°] Latitude [°] Maximum lead-time T 18 0 - 5 s 5 - 10 s 10 - 15 s 15 - 20 s 20 - 25 s 25 - 30 s 30 - 35 s 35 - 40 s > 40 s ISNet stations 13.5 14 14.5 15 15.5 16 40 40.5 41 41.5 Longitude [°] Latitude [°] Minimum lead-time T 29 Blind zone 0 - 5 s 5 - 10 s ISNet stations 13.5 14 14.5 15 15.5 16 40 40.5 41 41.5 Longitude [°] Latitude [°] Medium lead-time T 29 Blind zone 0 - 5 s 5 - 10 s 10 - 15 s 15 - 20 s 20 - 25 s 25 - 30 s ISNet stations 13.5 14 14.5 15 15.5 16 40 40.5 41 41.5 Longitude [°] Latitude [°] Maximum lead-time T 29 Blind zone 0 - 5 s 5 - 10 s 10 - 15 s 15 - 20 s 20 - 25 s 25 - 30 s 30 - 35 s 40 - 45 s ISNet stations Conclusions The study presented here examines the feasibility of EEW for the Campanian regional seismic network. We considered lead-time as the design parameter for EEW engineering applications. Result of the analyses, presented in the form of maps, confirms that evacuation of buildings is hardly possible in the region; however, the estimated warning times seem to be sufficient to activate several types of alternative security measures in selected critical structures/infrastructures. One of the limitations to the effectiveness of an EEWS is the existence of a “ blind zone blind zone” for each earthquake within which no warning is available. The size of blind zone is dependent on the time required to assess the hazard for an earthquake, which, in turn, is dependent on the proximity of seismic stations and the duration of P-wave data required at each station (typically ~ 4 sec). Lead-time and feasible risk reduction action Lead-time maps for the case-study region can be superimposed to real-time risk reduction actions for specific structural systems. These security measures can be classified according to the time required to be carried out. 13.5 14 14.5 15 15.5 16 40 40.5 41 41.5 Longitude [°] Latitude [°] Medium lead-time T 4 0 - 5 s 5 - 10 s 10 - 15 s 15 - 20 s 20 - 25 s 25 - 30 s 30 - 35 s 35 - 40 s ISNet stations • Semi-active structural control (~ 1 sec). • Turning all the traffic lights on the freeway to red to alert drivers and stop traffic (~ 5 sec). • At home, moving away from large appliances or pieces of furniture that could fall over and taking cover under a table or other object that provides similar protection (~ 5 – 10 sec). • Outdoors, looking out for collapsing objects (signs, broken glass, cornice, …) and taking shelter in a sturdy building if there is one close enough (~ 5 – 10 sec). • Stopping hazardous work (in building yards, …) to secure safety (~ 5 – 10 sec). • Moving away from hazardous chemical systems and machinery to secure safety (~ 5 – 10 sec). • In public buildings, do not panic and do not rush for the exit or stairs, following the attendant’s instructions (~ 10 sec). • Stopping elevators at the nearest floor and opening its doors immediately to prevent people from being trapped (~ 5 – 10 sec). • Activating backup and turning off important computers (~ 5 sec – 10 sec). • Slowing down trains (and stopping them if necessary/possible) to prevent railway accidents (a few tents of seconds dependent on train speed). • Preventing planes from landing (a few tents of seconds). • Conducted (and planned) evacuation (in public buildings) (~ > 40 sec). • Controlling production lines to mitigate damage (~ 15 sec). • Shutdown of critical systems (sensitive bio-medical equipments in hospitals, reactors and other hazardous equipments in energy and chemical plants) and lifelines/pipelines (~ 20 sec). • Suspending work in progress in operating rooms to avoid mistakes (~ 20 – 30 sec). P and S-waves arrival time as a function of epicentral distance from an earthquake 1D v P model for Campanian region (v P /v S = 1.68) Stations Stations Local Control Local Control Center Center Network Control Network Control Center Center city of Napoli 4 small towns hospitals railways highways industries gas/electric pipelines fire stations Potential targets for EEWS in Campanian region ISNet architecture INDIVIDUAL ACTIONS (to enable personal protection) AUTOMATIC ACTIONS (thanks to Automatic Control and Systems Engineering, Structural Engineering, …) Definitions of lead-time In order to compute the minimum, maximum and average lead times associated to each point (P j ) in Campania, we used a 1D v P model for the region with a constant v P /v S of 1.68. Furthermore, we considered events occurring in the area covered by the ISNet infrastructure, assuming a three-variate uniform distribution for the hypocenter coordinates (the earthquakes hypocentral depths in Campania are small ranging from ~ 4 to ~ 12 km). For P j, the resulting lead-time, T 1 j may be approximated as: where v P (z) and v S (z) are the P- and S-wave velocities, x st and x j are the distance of the closest (to the source) seismic station and the target being warned (P j ) from an earthquake’s epicenter, h is the depth of the event and ∆t p is the required processing time (time needed by the system to trigger and record a sufficient length of waveforms + time to process the data) [2]. (T S j |h) and (T P st |h) can be computed using appropriate travel time curves for the given source depth h. However, a real-time estimate of the size and hazard of an earthquake using a little of data (i.e. data from 1 seismic station) is not reliable and shows wide scattering. The accuracy of magnitude, location and ground motion estimates is a function of the number of stations providing P-wave data. We observed that these estimates tend to be reliable when averaged over at least 4 stations and stable, in a probabilistic sense, when averaged over 18 stations [3]. Then we calculated lead time as: Finally, it is interesting to estimate lead-time when all the seismic stations have triggered: Distribution for the hypocenter coordinates ( 29 ( 29 ( 29 ( 29 p st j p st j j z z t h | T h | T t v h x v h x T P S P 2 2 S 2 2 1 ∆ - - 2245 ∆ - + - + 2245 ( 29 ( 29 ( 29 ( 29 p i i st j j p i i st j j t h | T h | T T , t h | T h | T T 18 1 , P S 18 4 1 , P S 4 ∆ - - 2245 ∆ - - 2245 ∑ ∑ = = ( 29 ( 29 p i i st j j t h | T h | T T 29 1 , P S 29 ∆ - - 2245 ∑ = References [1] E. Weber, V. Convertito, G. Iannaccone, A. Zollo, A. Bobbio, L. Cantore, M. Corciulo, M. Di Crosta, L. Elia, C. Martino, A. Romeo, and C. Satriano - An Advanced Seismic Network in the Southern Apennines (Italy) for Seismicity Investigations and Experimentation with Earthquake Early Warning - Seismological Research Letters, 2007 [2] R. M. Allen - Probabilistic Warning Times for Earthquake Ground Shaking in the San Francisco Bay Area - Seismological Research Letters, 2006 [3] I. Iervolino, V. Convertito, M. Giorgio, G. Manfredi, A. Zollo - Real-time risk analysis for hybrid earthquake early warning systems - Journal of Earthquake Engineering, 2006 Iunio Iervolino, Carmine Galasso *, Gaetano Manfredi
Transcript of Toward EEW engineering applications: leadToward EEW ... · With an area of 13.595 km², Campania is...
EGU General Assembly 2008, Vienna, Austria, 13 - 18 April 2008
Toward EEW engineering applications: leadToward EEW engineering applications: leadToward EEW engineering applications: leadToward EEW engineering applications: lead----time maps for the ISNet infrastructure in the Campanian regiontime maps for the ISNet infrastructure in the Campanian regiontime maps for the ISNet infrastructure in the Campanian regiontime maps for the ISNet infrastructure in the Campanian region
Department of Structural Engineering, University of Naples Federico II, Italy (* [email protected])
���� Motivation The feasibility of Earthquake Early Warning Systems (EEWS) depends I)I) on the possibility of real-time estimation of the size of the event (source
parameters) and the intensity of subsequent ground shaking at various sites (although with large associated uncertainty) and, at the same time, II) II) on the ability to design earthquake engineering applications for real-time seismic risk mitigation. Among the information needed to evaluate the effectiveness of EEWS, there is the estimation of the available time to perform security actions before the arrival of the more energetic seismic phase at the site of interest. This interval is termed leadlead --timetime . A rational approach to the feasibility analysis of EEWS in a densely urbanized area requires the assessment of likely and minimum-maximum range lead-times (in a probabilistic manner) for the events the seismic infrastructure can detect.
���� The Irpinia Seismic Network (ISNet) With an area of 13.595 km², Campania is only the twelfth largest region in Italy, but a population of around 5.8 million people makes it the second most populous, and the most densely populated region in Italy. In Campanian region the most hazardous seismogenic source is close to, if not overlapped, with densely urbanized areas and time consuming security actions may be unfeasible. Anyway, historical earthquakes damaged wide areas (even larger than 100 km in radius) due to a combination of the geology (low rate of seismic attenuation in the region) and of the low quality of the buildings in relation to their resistance to seismic action. For this reason, Campanian region is now provided with an advanced and dense sensor network mainly devoted to EW purposes. The Irpinia Seismic Network (ISNet) covers an area of about 100 x 70 km2, corresponding to a part of the entire seismic source area [1] . ISNet comprises 29, 6 components seismic station equipped with both accelerometers and velocimeters, with real-time telemetry. Following a distributed approach, the network is organized in 6 sub-nets: waveform data is collected and elaborated at local hubs (LCC, Local Control Centers) which, in turn send processed parameters to a Network Control Center (NCC) in Naples.
���� Lead-time maps Our study has yielded the following results, presented in the form of maps. We evaluated the minimum, maximum and average lead times (T4, T18 and T29) for each node of a regular grid having a ~ 2 km spacing and covering the whole Campanian territory with over 2700 nodes.
13.5 14 14.5 15 15.5 1640
40.5
41
41.5
Longitude [°]
Latit
ude
[°]
Minimum lead-time T 4
Blind zone0 - 5 s5 - 10 s10 - 15 s15 - 20 s20 - 25 sISNet stations
13.5 14 14.5 15 15.5 1640
40.5
41
41.5
Longitude [°]
Latit
ude
[°]
Medium lead-time T 4
0 - 5 s5 - 10 s10 - 15 s15 - 20 s20 - 25 s25 - 30 s30 - 35 s35 - 40 sISNet stations
13.5 14 14.5 15 15.5 1640
40.5
41
41.5
Longitude [°]
Latit
ude
[°]
Maximum lead-time T 4
5 - 10 s10 - 15 s15 - 20 s20 - 25 s25 - 30 s30 - 35 s35 - 40 s> 40 sISNet stations
13.5 14 14.5 15 15.5 1640
40.5
41
41.5
Longitude [°]
Latit
ude
[°]
Minimum lead-time T 18
Blind zone0 - 5 s5 - 10 s10 - 15 s15 - 20 sISNet stations
13.5 14 14.5 15 15.5 1640
40.5
41
41.5
Longitude [°]
Latit
ude
[°]
Medium lead-time T 18
Blind zone0 - 5 s5 - 10 s10 - 15 s15 - 20 s20 - 25 s25 - 30 s30 - 35 sISNet stations
13.5 14 14.5 15 15.5 1640
40.5
41
41.5
Longitude [°]
Latit
ude
[°]
Maximum lead-time T 18
0 - 5 s5 - 10 s10 - 15 s15 - 20 s20 - 25 s25 - 30 s30 - 35 s35 - 40 s> 40 sISNet stations
13.5 14 14.5 15 15.5 1640
40.5
41
41.5
Longitude [°]
Latit
ude
[°]
Minimum lead-time T 29
Blind zone0 - 5 s5 - 10 sISNet stations
13.5 14 14.5 15 15.5 1640
40.5
41
41.5
Longitude [°]
Latit
ude
[°]
Medium lead-time T 29
Blind zone0 - 5 s5 - 10 s10 - 15 s15 - 20 s20 - 25 s25 - 30 sISNet stations
13.5 14 14.5 15 15.5 1640
40.5
41
41.5
Longitude [°]
Latit
ude
[°]
Maximum lead-time T 29
Blind zone0 - 5 s5 - 10 s10 - 15 s15 - 20 s20 - 25 s25 - 30 s30 - 35 s40 - 45 sISNet stations
���� Conclusions
The study presented here examines the feasibility of EEW for the Campanian regional seismic network. We considered lead-time as the design parameter for EEW engineering applications. Result of the analyses, presented in the form of maps, confirms that evacuation of buildings is hardly possible in the region; however, the estimated warning times seem to be sufficient to activate several types of alternative security measures in selected critical structures/infrastructures. One of the limitations to the effectiveness of an EEWS is the existence of a “blind zoneblind zone ” for each earthquake within which no warning is available. The size of blind zone is dependent on the time required to assess the hazard for an earthquake, which, in turn, is dependent on the proximity of seismic stations and the duration of P-wave data required at each station (typically ~ 4 sec).
���� Lead-time and feasible risk reduction action Lead-time maps for the case-study region can be superimposed to real-time risk reduction actions for specific structural systems. These security measures can be classified according to the time required to be carried out.
13.5 14 14.5 15 15.5 1640
40.5
41
41.5
Longitude [°]
Latit
ude
[°]
Medium lead-time T 4
0 - 5 s5 - 10 s10 - 15 s15 - 20 s20 - 25 s25 - 30 s30 - 35 s35 - 40 sISNet stations
• Semi-active structural control (~ 1 sec) .
• Turning all the traffic lights on the freeway to red to alert drivers and stop traffic (~ 5 sec) .
• At home, moving away from large appliances or pieces of furniture that could fall over and taking cover under a table or other object that provides similar protection (~ 5 – 10 sec) .
• Outdoors, looking out for collapsing objects (signs, broken glass, cornice, …) and taking shelter in a sturdy building if there is one close enough (~ 5 – 10 sec) .
• Stopping hazardous work (in building yards, …) to secure safety (~ 5 – 10 sec) .
• Moving away from hazardous chemical systems and machinery to secure safety (~ 5 – 10 sec) .
• In public buildings, do not panic and do not rush for the exit or stairs, following the attendant’s instructions (~ 10 sec) .
• Stopping elevators at the nearest floor and opening its doors immediately to prevent people from being trapped (~ 5 – 10 sec) .
• Activating backup and turning off important computers (~ 5 sec – 10 sec) .
• Slowing down trains (and stopping them if necessary/possible) to prevent railway accidents (a few tents of seconds dependent on train speed) .
• Preventing planes from landing (a few tents of seconds) .
• Conducted (and planned) evacuation (in public buildings) (~ > 40 sec) .
• Controlling production lines to mitigate damage (~ 15 sec) .
• Shutdown of critical systems (sensitive bio-medical equipments in hospitals, reactors and other hazardous equipments in energy and chemical plants) and lifelines/pipelines (~ 20 sec) .
• Suspending work in progress in operating rooms to avoid mistakes (~ 20 – 30 sec) .
���� P and S-waves arrival time as a function of
epicentral distance from an earthquake ���� 1D vP model for Campanian region (vP/vS = 1.68)
StationsStations
Local Control Local Control CenterCenter
Network Control Network Control CenterCenter
city of Napoli 4 small towns hospitals railways highways industries gas/electric pipelines fire stations
Potential targets for EEWS in Campanian region
ISNet architecture
� INDIVIDUAL ACTIONS (to enable personal protection)
� AUTOMATIC ACTIONS (thanks to Automatic Control and Systems Engineering, Structural Engineering, …)
���� Definitions of lead-time In order to compute the minimum, maximum and average lead times associated to each point (Pj) in Campania, we used a 1D vP model for the region with a constant vP/vS of 1.68. Furthermore, we considered events occurring in the area covered by the ISNet infrastructure, assuming a three-variate uniform distribution for the hypocenter coordinates (the earthquakes hypocentral depths in Campania are small ranging from ~ 4 to ~ 12 km). For Pj, the resulting lead-time, T1
j may be approximated as: where vP(z) and vS(z) are the P- and S-wave velocities, xst and xj are the distance of the closest (to the source) seismic station and the target being warned (Pj) from an earthquake’s epicenter, h is the depth of the event and ∆tp is the required processing time (time needed by the system to trigger and record a sufficient length of waveforms + time to process the data) [2] . (TS
j|h) and (TP
st|h) can be computed using appropriate travel time curves for the given source depth h. However, a real-time estimate of the size and hazard of an earthquake using a little of data (i.e. data from 1 seismic station) is not reliable and shows wide scattering. The accuracy of magnitude, location and ground motion estimates is a function of the number of stations providing P-wave data. We observed that these estimates tend to be reliable when averaged over at least 4 stations and stable, in a probabilistic sense, when averaged over 18 stations [3] . Then we calculated lead time as:
Finally, it is interesting to estimate lead-time when all the seismic stations have triggered:
���� Distribution for the hypocenter coordinates
( ) ( ) ( ) ( ) pstj
pstjj
zzth|Th|Tt
v
hx
v
hxT PS
P
22
S
22
1 ∆−−≅∆−+
−+
≅
( ) ( ) ( ) ( ) pi
istjjp
i
istjj th|Th|TT , th|Th|TT18
1
,PS18
4
1
,PS4 ∆−−≅∆−−≅ ∑∑
==
( ) ( ) pi
istjj th|Th|TT29
1
,PS29 ∆−−≅ ∑
=
References [1] E. Weber, V. Convertito, G. Iannaccone, A. Zollo, A. Bobbio, L. Cantore, M. Corciulo, M. Di Crosta, L. Elia, C. Martino, A. Romeo, and C. Satriano - An Advanced Seismic Network in the Southern
Apennines (Italy) for Seismicity Investigations and Experimentation with Earthquake Early Warning - Seismological Research Letters, 2007
[2] R. M. Allen - Probabilistic Warning Times for Earthquake Ground Shaking in the San Francisco Bay Area - Seismological Research Letters, 2006
[3] I. Iervolino, V. Convertito, M. Giorgio, G. Manfredi, A. Zollo - Real-time risk analysis for hybrid earthquake early warning systems - Journal of Earthquake Engineering, 2006
Iunio Iervolino, Carmine Galasso *, Gaetano Manfred i
