Large earthquakes produce crustal deformation that can be quantified by geodetic measurements, al... more Large earthquakes produce crustal deformation that can be quantified by geodetic measurements, allowing for the determination of the slip distribution on the fault. We used data from Global Positioning System (GPS) networks in Central Chile to infer the static deformation and the kinematics of the 2010 moment magnitude (Mw) 8.8 Maule megathrust earthquake. From elastic modeling, we found a total rupture length of ~500 kilometers where slip (up to 15 meters) concentrated on two main asperities situated on both sides of the epicenter. We found that rupture reached shallow depths, probably extending up to the trench. Resolvable afterslip occurred in regions of low coseismic slip. The low-frequency hypocenter is relocated 40 kilometers southwest of initial estimates. Rupture propagated bilaterally at about 3.1 kilometers per second, with possible but not fully resolved velocity variations.
Most of worldwide developed Earthquake Early Warning Systems (EEWS) are conceived as either “regi... more Most of worldwide developed Earthquake Early Warning Systems (EEWS) are conceived as either “regional” (network-based) or “on-site” (stand-alone) systems (Kanamori,2005). A regional EEWS is based on a dense sensor network covering a portion or the entirety of an area that is threatened by earthquakes. The relevant source parameters (event location and magnitude) are estimated from the early portion of recorded signals and are used to predict, with a quantified confidence, a ground motion intensity measure at a distant site where a target structure of interest is located. An on-site EEWS consists of a single sensor or an array of sensors deployed in the proximity of the target structure that is to be alerted, and whose measurements of peak amplitude and/or predominant period on the initial P-wave motion are used to predict the ensuing peak ground motion (mainly related to the arrival of S and surface waves) at the same site. Wu and Kanamori (2005,2008) have proposed a methodology for on-site early warning applications aimed at the rapid assessment of earthquake damage potential, which is based on the near source (epicentral distance smaller than 30 km), real-time measurement of the peak displacement (Pd) and a period parameter(τc) from early P-wave signals. In this note we explore the feasibility of applying a similar approach in Italy, based on the analysis of the strong motion records from the European Strong Motion Data Base and from the main events (M≥3.4) of the recent 2009, Mw 6.3 Central Italy (L’Aquila) earthquake sequence. The empirical relationships for early warning parameters (PGV,peak ground velocity, vs Pd and τc vs Magnitude) are compared with measurements on Japan and Taiwan earthquakes in a wide distance/magnitude range. By extending the observation distance to 100 km, and by comparing data from different worldwide seismic regions, we generalize the Wu and Kanamori’ method by proposing an integrated regional/on-site early warning method which can be used in the very first seconds after a moderate to large earthquake to map the most probable damaged zones. The settings of specific thresholds for Pd and τc allows to rapidly issue an earthquake alert based on a four entries, decision-table scheme. The proposed approach is suitable for Italy, where, during the last two decades, a dense network of wide dynamic-range accelerometer arrays has been deployed by the Department of Civil Protection, INGV and other regional research agencies. References Kanamori, H., 2005, Real-time seismology and earthquake damage mitigation, Annu. Rev. Earth Planet. Sci. , 33, 195-214. Wu, Y-M. and Kanamori, H., 2005. Rapid Assessment of Damage Potential of Earthquakes in Taiwan from the Beginning of P Waves. Bulletin of the Seismological Society of America,, 95(3), 1181-1185.
1] It has been shown that the initial portion of P and S wave signals can provide information abo... more 1] It has been shown that the initial portion of P and S wave signals can provide information about the final earthquake magnitude in a wide magnitude range. This observation opens the perspective for the real-time determination of source parameters. In this paper we describe a probabilistic evolutionary approach for the real-time magnitude estimation which can have a potential use in earthquake early warning. The technique is based on empirical prediction laws correlating the low-frequency peak ground displacement measured in a few seconds after the P and/or S phase arrival and the final event magnitude. The evidence for such a correlation has been found through the analysis of 256 shallow crustal events in the magnitude range M jma 4-7.1 located over the entire Japanese archipelago. The peak displacement measured in a 2-s window from the first P phase arrival correlates with magnitude in the range M = [4-6.5]. While a possible saturation effect above M ' 6.5 is observed, it is less evident in an enlarged window of 4 s. The scaling of S peaks with magnitude is instead also observed at smaller time lapses (i.e., 1 s) after the first S arrival. The different scaling of P and S peaks with magnitude when measured in a 2-s window is explained in terms of different imaged rupture surface by the early portion of the body wave signals. We developed a technique to estimate the probability density function (PDF) of magnitude, at each time step after the event origin. The predicted magnitude value corresponds to the maximum of PDF, while its uncertainty is given by the 95% confidence bound. The method has been applied to the 2007 (M jma = 6.9) Noto Hanto and 1995 (M jma = 7.3) Kobe earthquakes. The results of this study can be summarized as follows: (1) The probabilistic algorithm founded on the predictive model of peak displacement versus final magnitude is able to provide a fast and robust estimation of the final magnitude. (2) The information available after a few seconds from the first detection of the P phase at the network can be used to predict the peak ground motion at a given regional target with uncertainties which are comparable to those derived from the attenuation law. (3) The near-source S phase data can be used jointly with P data for regional early warning purposes, thus increasing the accuracy and reliability of magnitude estimation. Citation: Lancieri, M., and A. Zollo (2008), A Bayesian approach to the real-time estimation of magnitude from the early P and S wave displacement peaks,
1] We show that the low-pass filtered, peak amplitudes of initial P-and S-wave seismic signals re... more 1] We show that the low-pass filtered, peak amplitudes of initial P-and S-wave seismic signals recorded in the vicinity of an occurring earthquake source correlates with the earthquake magnitude and may be used for real-time estimation of the event size in seismic early warning applications. The earthquake size can be therefore estimated using only a couple of seconds of signal from the P-or S-wave onsets, i.e. while the rupture itself is still propagating and rupture dimension is far from complete. We argue that dynamic stress release and/or slip duration on the fault in the very early stage of seismic fracture, scales both with the observed peak amplitude and with the elastic energy available for fracture propagation. The probability that a fracture grows to a larger size should scale with the energy initially available. Citation: Zollo, A., M. Lancieri, and S. Nielsen (2006), Earthquake magnitude estimation from peak amplitudes of very early seismic signals on strong motion records, Geophys. Res. Lett., 33, L23312,
The Northern Chile gap is generally considered to the site of the next megathurst event in Chile.... more The Northern Chile gap is generally considered to the site of the next megathurst event in Chile. The Tocopilla earthquake of 14 November 2007 (Mw 7.8) and aftershock series broke the southern end of this gap. The Tocopilla event ruptured a narrow strip of 120 km of length and a width that (Peyrat et al.; Delouis et al. 2009) estimated as 30 km. The aftershock sequence comprises five large thrust events with magnitude greater than 6. The main aftershock of Mw 6.7 occurred on November 15, at 15:06 (UTM) seawards of the Mejillones Peninsula. One month later, on December 16 2007, a strong (Mw 6.8) intraplate event with slab-push mechanism occurred near the bottom of the rupture zone. These events represent a unique opportunity for the study of earthquakes in Northern Chile because of the quantity and quality of available data. In the epicentral area, the IPOC network was deployed by GFZ, CNRS/INSU and DGF before the main event. This is a digital, continuously recording network, equipped with both strong-motion and broad-band instrument. On 29 November 2007 a second network named “Task Force” (TF) was deployed by GFZ to study the aftershocks. This is a dense network, installed near the Mejillones peninsula. It is composed by 20 short-period instruments. The slab-push event of 16 december 2007 occurred in the middle of the area covered by the TF network. Aftershocks were detected using an automatic procedure and manually revised in order to pick P and S arrivals. In the 14-28 November period, we detected 635 events recorded at the IPOC network; and a further 552 events were detected between 29 November and 16 December before the slab-push event using the TF network. The events were located using a vertically layered velocity model (Husen et al. 1999), using the NLLoc software of Lomax et al. From the broadband data we estimated the moment magnitude from the displacement spectra of the events. From the short-period instruments we evaluated local magnitudes using the technique proposed by Kanamori and Jennings that consists in digitally converting the short period records of S waves to those that would have been recorded by a classical Wood-Anderson seismometer. The Tocopilla earthquake and the 16 December intermediate depth earthquake had no precursors. The results of location show no events located in the southern area of the Mejillones peninsula, confirming that this peninsula acted like a barrier to rupture. With the TF network we find the aftershocks principally in the southern area of the main rupture. This locations shows a kink in the subduction plane and we determine that it occurs near to 45km of depth . We are currently studying the focal mechanism of the main aftershocks and we search to compare the slip distribution of the earthquakes with the aftershock distribution.
The ISNet (Irpinia Seismic Network) consists of 29 six-components seismic stations deployed in a ... more The ISNet (Irpinia Seismic Network) consists of 29 six-components seismic stations deployed in a 100×70 km2 area in Southern Italy, containing the fault system that generated the 1980, M=6.9, Irpinia earthquake. The seismic stations, equipped with both accelerometers and seismometers, are able to follow both the strong and the weak motion. The computation is distributed along the network, with local control centers collecting data-streams from the single stations and performing a redundant computation of the early warning parameters (picking, location, magnitude and hazard integral). The decision is finally sent to the network control center, which is located far-away from the network itself. The performance of the infrastructure therefore depends on the latency of the information in the acquisition and in the transmission, on the computation time and on the failures of the system. We build up a virtual simulator to analyze off-line both the lead-time of the whole infrastructure and the influence of the failures of the system on the final decision. We use the simulator as a platform to test and optimize the algorithms for the computation of the early warning parameters, to investigate the optimum topology for the existing network, to asses weaknesses of the present infrastructure and define interventions plans to make the alarm decision robust. The simulator is a flow chart, in which the single blocks (acquisition, transmission, computation and decision) are either mathematical operators based on empirically estimated physical quantities or code lines running on the same platform. We evaluate the performances of the network on a synthetic database of records simulating the 1980, Irpinia earthquake and a set of real data from an earthquake of magnitude 3.0.
A prototype system for earthquake early warning and rapid shake map evaluation is being developed... more A prototype system for earthquake early warning and rapid shake map evaluation is being developed and tested in southern Italy based on a dense, wide dynamic-range seismic network (accelerometers + seismometers) under installation in the Apenninic belt region (Irpinia Seismic Network). This system forms a regional Earthquake Early Warning System consisting of a seismic sensor network covering a portion of the expected epicentral area for large earrthquakes. Considering a warning window ranging from tens of seconds before to hundred of seconds after an earthquake, several public infrastructures and buildings of strategic relevance (hospitals, gas pipelines, railways, railroads, ...) of the Regione Campania are potential test-sites for testing innovative technologies for data acquisition, processing and transmission. A potential application of an early warning system in the Campania region based on the Irpinia network, should consider an expected time delay to the first energetic S wave train varying between 14-20 sec at 40-60 km distance to 26-30 sec at about 80-100 km, from a crustal earthquake occurring in the source region. The latter is the typical time window available for mitigating earthquake effects through early warning in the city of Naples (about 2 million of inhabitants including suburbs). We have developed a method for real time earthquake location following a probabilistic approach. The earthquake location is expressed as a probability density function for the hypocenter location in 3D space based on the concept of equal differential-time (EDT). It provides a location as the maximum of a stack over quasi-hyperbolic surfaces. On each surface the difference of calculated travel-times at a pair of stations is equal to the difference of observed arrival times at the same pair of stations. For an increasing number of P-wave readings, progressively acquired in the short time after the occurrence of an earthquake, the EDT method can be generalized by introducing an evolutionary strategy which is aimed at obtaining a more and more refined estimate of the maximum probability volume as the time goes on. The real time magnitude estimate will take advantage from the high spatial density of the network in the source region and the wide dynamic range of installed instruments. Based on the offline analysis of high quality strong-motion data bases recorded in Italy and worldwide, several methods will be checked and validated , using different observed quantities (peak amplitude, dominant frequency, square velocity integral, .) to be measured on seismograms, as a function of time. Following the ElarmS methodology (Allen,2004), peak ground attenuation relations can be used to predict the distribution of maximum ground shaking, as updated estimates of earthquake location and magnitude are progressively available from the Early Warning system starting from the time of first P-wave detection. As measurements of peak ground quantities for the current earthquake become available from the network, these values are progressively used to adjust an "ad hoc" determined attenuation relation for the Campania region using the stochastic approach proposed by Boore (1993).
The Irpinia Seismic Network (ISNet) is deployed in Southern Apennines along the active fault syst... more The Irpinia Seismic Network (ISNet) is deployed in Southern Apennines along the active fault system responsible for the 1980, November 23, M s 6.9 Campania–Lucania earthquake. It is set up by 28 stations and covers an area of about 100 × 70 km2. Each site is equipped with a 1-g full-scale accelerometer and a short-period velocimeter. Due to its design characteristics, i.e., the wide dynamic range and the high density of stations, the ISNet network is mainly devoted to estimating in real-time the earthquake location and magnitude from low- to high- magnitude events, and to providing ground-motion parameters values so to get some insights about the ground shaking expected. Moreover, the availability of high-quality of data allows studying the source processes related to the seismogenetic structures in the area. The network layout, the data communication system and protocols and the main instrumental features are described in the paper. The data analysis is managed by Earthworm software package that also provides the earthquake location while custom software has been developed for real-time computation of the source parameters and shaking maps. Technical details about these procedures are given in the article. The data collected at the ISNet stations are available upon request.
1] It has been shown that the initial portion of P and S wave signals can provide information abo... more 1] It has been shown that the initial portion of P and S wave signals can provide information about the final earthquake magnitude in a wide magnitude range. This observation opens the perspective for the real-time determination of source parameters. In this paper we describe a probabilistic evolutionary approach for the real-time magnitude estimation which can have a potential use in earthquake early warning. The technique is based on empirical prediction laws correlating the low-frequency peak ground displacement measured in a few seconds after the P and/or S phase arrival and the final event magnitude. The evidence for such a correlation has been found through the analysis of 256 shallow crustal events in the magnitude range M jma 4-7.1 located over the entire Japanese archipelago. The peak displacement measured in a 2-s window from the first P phase arrival correlates with magnitude in the range M = [4-6.5]. While a possible saturation effect above M ' 6.5 is observed, it is less evident in an enlarged window of 4 s. The scaling of S peaks with magnitude is instead also observed at smaller time lapses (i.e., 1 s) after the first S arrival. The different scaling of P and S peaks with magnitude when measured in a 2-s window is explained in terms of different imaged rupture surface by the early portion of the body wave signals. We developed a technique to estimate the probability density function (PDF) of magnitude, at each time step after the event origin. The predicted magnitude value corresponds to the maximum of PDF, while its uncertainty is given by the 95% confidence bound. The method has been applied to the 2007 (M jma = 6.9) Noto Hanto and 1995 (M jma = 7.3) Kobe earthquakes. The results of this study can be summarized as follows: (1) The probabilistic algorithm founded on the predictive model of peak displacement versus final magnitude is able to provide a fast and robust estimation of the final magnitude. (2) The information available after a few seconds from the first detection of the P phase at the network can be used to predict the peak ground motion at a given regional target with uncertainties which are comparable to those derived from the attenuation law. (3) The near-source S phase data can be used jointly with P data for regional early warning purposes, thus increasing the accuracy and reliability of magnitude estimation. Citation: Lancieri, M., and A. Zollo (2008), A Bayesian approach to the real-time estimation of magnitude from the early P and S wave displacement peaks,
We study the dynamic rupture propagation of a M 6.7 intraplate earthquake that occurred 16 Decemb... more We study the dynamic rupture propagation of a M 6.7 intraplate earthquake that occurred 16 December 2007, a month after a large thrust event of Tocopilla, Chile (M 7.7). The occurrence of a slab push event after a large subduction earthquake is well explained by Coulomb stress transfer models and crack dynamics. A dense seismic network, equipped with short period and accelerometers was deployed after the event of 14 November 2007 by the Task Force of GFZ Potsdam and the University of Chile in Santiago. This network was in place on December 16 providing an excellent data set for this earthquake. We used these data to make a detailed study of rupture processes. We localized the main event of December 16 and the aftershocks that occurred within 24 h of the main event. The main event was located at 43 km depth, while the aftershocks distribution covered a circular zone of 5 to 8 km of radius centred on the main shock epicentre. The aftershocks are distributed on an almost vertical plane that agrees with one of the fault planes of the mechanism (86° dip) and all the aftershock have the same mechanism as the main event. We used nearest accelerometric records in order to do dynamic inversion, two of these accelerometers were situated right above the hypocentre. We performed a non-linear dynamic inversion based on the neighbourhood algorithm (NA) and MonteCarlo methods with an L2 norm. The data was initially filtered in the 0.05-1 Hz. The velocity model was derived from previous work by GFZ. The earthquake was modelled using finite differences on a grid of variable size. Friction was modelled by the standard Ida slip weakening friction law. At each step of the inversion more than 32 full numerical simulations are carried in parallel. These simulations have been optimized in order to reduce the computer time to a minimum. The best models that result from dynamic inversion reduced the variance by more than 30 %, these models ruptured a relatively small zone of the fault plane with a diameter of about 6 km. The rupture speed was quite slow and the energy release rate that gives the best results was of the order of 10 MJ/m2. We explored a large area of the parameter space in search for solutions with faster rupture speeds but we have failed to find them. It appears then that this earthquake, located right under the transition zone from steady to stick slip, propagated very slowly. We have confronted the results of dynamic inversion with displacement records filtered at several frequency bands up to 4 Hz. The inverted models depend on the frequency band used for the inversion because the observed and synthetic records are controlled by a large stopping phase that contains most of the radiated energy. We also study the initiation of the event on seismograms, the event started by a clear cascade process involving several initial ruptures.
1] We investigate the effect of extended faulting processes and heterogeneous wave propagation on... more 1] We investigate the effect of extended faulting processes and heterogeneous wave propagation on the early warning system capability to predict the peak ground velocity (PGV) from moderate to large earthquakes occurring in the southern Apennines (Italy). Simulated time histories at the early warning network have been used to retrieve early estimates of source parameters and to predict the PGV, following an evolutionary, probabilistic approach. The system performance is measured through the Effective Lead-Time (ELT), i.e., the time interval between the arrival of the first S-wave and the time at which the probability to observe the true PGV value within one standard deviation becomes stationary, and the Probability of Prediction Error (PPE), which provides a measure of PGV prediction error. The regional maps of ELT and PPE show a significant variability around the fault up to large distances, thus indicating that the system's capability to accurately predict the observed peak ground motion strongly depends on distance and azimuth from the fault.
1] We show that the low-pass filtered, peak amplitudes of initial P-and S-wave seismic signals re... more 1] We show that the low-pass filtered, peak amplitudes of initial P-and S-wave seismic signals recorded in the vicinity of an occurring earthquake source correlates with the earthquake magnitude and may be used for real-time estimation of the event size in seismic early warning applications. The earthquake size can be therefore estimated using only a couple of seconds of signal from the P-or S-wave onsets, i.e. while the rupture itself is still propagating and rupture dimension is far from complete. We argue that dynamic stress release and/or slip duration on the fault in the very early stage of seismic fracture, scales both with the observed peak amplitude and with the elastic energy available for fracture propagation. The probability that a fracture grows to a larger size should scale with the energy initially available. Citation: Zollo, A., M. Lancieri, and S. Nielsen (2006), Earthquake magnitude estimation from peak amplitudes of very early seismic signals on strong motion records, Geophys. Res. Lett., 33, L23312,
1] From inspection of a large set of Japanese events, we investigate the scaling of the early rad... more 1] From inspection of a large set of Japanese events, we investigate the scaling of the early radiated energy, inferred from the squared velocity integral (IV2) with the final magnitude of the event. We found that the energy can only discriminate whether the event has a magnitude larger or smaller than 5.8, and in the latter case it can allow for realtime magnitude estimation. However, by normalizing IV2 for the rupture area, the initial slip scales with the magnitude between 4 < M < 7 following the expected scaling laws. We show that the ratio between the squared peak displacement and IV2 is a proxy for the slip following the same scaling but it can be directly derived from the data, without any assumption on the rupture area. The scaling relationship between initial slip and magnitude can be used for early warning applications, when integrated in a probabilistic, evolutionary approach. Citation: Festa, G., A. Zollo, and M.
In the framework of earthquake early warning (EEW) experimentation in Southern Italy, we have dev... more In the framework of earthquake early warning (EEW) experimentation in Southern Italy, we have developed an integrated software platform called PRESTo (PRobabilistic and Evolutionary early warning SysTem). PRESTo is a methodology for regional EEW applications, based on a fully probabilistic and evolutionary approach. This means that probability density functions of source parameters and ground shaking levels are computed , which make possible to evaluate the uncertainties, and to set an appropriate level of reliability of the issued warning. Moreover, the estimates of earthquake location, size and expected ground motion are continuously updated, as new data is available in real-time from the network. The earthquake location algorithm is based on an equal differential time formulation, and a probabilistic framework for describing the hypocenter solution. The algorithm, at each time step, relies on both the information from triggered arrivals and not-yet-triggered stations. The magnitude estimation is based on a Bayesian approach, and exploits an empirical relationship that correlates the final event magnitude with the logarithm of the distance-corrected, peak ground displacement, measured over the first 2÷4 seconds of P- and S-signals. Finally, a ground motion prediction equation valid at regional scale is used to predict the peak velocities and accelerations at the target sites. The core infrastructure of the EEW system in Southern Italy is the high density, wide dynamic-range seismic network ISNet (Irpinia Seismic Network), deployed along the Apennine chain, which is frequently stricken by moderate to large magnitude damaging earthquakes. For earthquakes originating inside the ISNet seismic network, a first alarm can be issued within 4-6 seconds from the origin time, and a stable estimate with relatively small error on location (less than 5 km) and magnitude (about 0.5 magnitude units) is achieved within 10 seconds after the origin time. PRESTo can easily be configured, tailored and exported to other digital seismic networks worldwide. A simple integration with the network is granted by the widely available SeedLink communication protocol used for data acquisition. In this note we illustrate and discuss the method flexibility and exportability to different acquisition layouts through off-line simulations, using real data recorded by the Japanese K-Net and KiK-Net, and RAN, the Italian strong motion network operated by the Department of Civil Protection.
1] From inspection of a large set of Japanese events, we investigate the scaling of the early rad... more 1] From inspection of a large set of Japanese events, we investigate the scaling of the early radiated energy, inferred from the squared velocity integral (IV2) with the final magnitude of the event. We found that the energy can only discriminate whether the event has a magnitude larger or smaller than 5.8, and in the latter case it can allow for realtime magnitude estimation. However, by normalizing IV2 for the rupture area, the initial slip scales with the magnitude between 4 < M < 7 following the expected scaling laws. We show that the ratio between the squared peak displacement and IV2 is a proxy for the slip following the same scaling but it can be directly derived from the data, without any assumption on the rupture area. The scaling relationship between initial slip and magnitude can be used for early warning applications, when integrated in a probabilistic, evolutionary approach. Citation: Festa, G., A. Zollo, and M.
Most earthquake early warning systems (EEWS) developed so far are conceived as either `regional' ... more Most earthquake early warning systems (EEWS) developed so far are conceived as either `regional' (network-based) or `on-site' (stand-alone) systems. The recent implementation of nationwide, high dynamic range, dense accelerometer arrays makes now available, potentially in real time, unsaturated waveforms of moderate-to-large magnitude earthquakes recorded at very short epicentral distances (<10-20 km). This would allow for a drastic increase of the early warning lead-time, for example, the time between the alert notification and the arrival time of potentially destructive waves at a given target site. By analysing strong motion data from modern accelerograph networks in Japan, Taiwan and Italy, we propose an integrated regional/on-site early warning method, which can be used in the very first seconds after a moderate-to-large earthquake to map the most probable damaged zones. The method is based on the real-time measurement of the period (τc) and peak displacement (Pd) parameters at stations located at increasing distances from the earthquake epicentre. The recorded values of early warning parameters are compared to threshold values, which are set for a minimum magnitude 6 and instrumental intensity VII, according to the empirical regression analyses of strong motion data. At each recording site the alert level is assigned based on a decisional table with four alert levels defined upon critical values of the parameters Pd and τc, which are set according to the error bounds estimated on the derived prediction equations. Given a real time, evolutionary estimation of earthquake location from first P arrivals, the method furnishes an estimation of the extent of potential damage zone as inferred from continuously updated averages of the period parameter and from mapping of the alert levels determined at the near-source accelerometer stations. The off-line application of the method to strong motion records of the Mw 6.3, 2009 Central Italy earthquake shows a very consistent match between the rapidly predicted (within a few seconds from the first recorded P wave) and observed damage zone, the latter being mapped from detailed macroseismic surveys a few days after the event. The proposed approach is suitable for Italy, where, during the last two decades, a dense network of wide dynamic-range accelerometer arrays has been deployed by the Department of Civil Protection (DPC), the Istituto Nazionale di Geofisica e Vulcanologia (INGV) and other regional research agencies.
PRESTo (PRobabilistic and Evolutionary early warning SysTem) is a software platform for regional ... more PRESTo (PRobabilistic and Evolutionary early warning SysTem) is a software platform for regional earthquake early warning that integrates recently developed algorithms for real-time earthquake location and magnitude estimation into a highly configurable and easily portable package. The system is under active experimentation in Southern Italy on the Irpinia Seismic Network (ISNet), which is deployed in a seismogenic area that is expected to produce a large earthquake within the next 20 years. In this paper we describe the architecture of the system and test its performances using both small earthquakes (M o3.5) recorded at the ISNet and a large event recorded in Japan, through a simulation mode. The results show that, when a dense seismic network is deployed in the fault area, PRESTo can produce reliable estimates of earthquake location and size within 5-6 s from the event origin. Each estimate is provided as a probability density function, with an uncertainty that typically decreases with time: a stable solution is generally reached within 10 s from the origin.
Large earthquakes produce crustal deformation that can be quantified by geodetic measurements, al... more Large earthquakes produce crustal deformation that can be quantified by geodetic measurements, allowing for the determination of the slip distribution on the fault. We used data from Global Positioning System (GPS) networks in Central Chile to infer the static deformation and the kinematics of the 2010 moment magnitude (Mw) 8.8 Maule megathrust earthquake. From elastic modeling, we found a total rupture length of ~500 kilometers where slip (up to 15 meters) concentrated on two main asperities situated on both sides of the epicenter. We found that rupture reached shallow depths, probably extending up to the trench. Resolvable afterslip occurred in regions of low coseismic slip. The low-frequency hypocenter is relocated 40 kilometers southwest of initial estimates. Rupture propagated bilaterally at about 3.1 kilometers per second, with possible but not fully resolved velocity variations.
Most of worldwide developed Earthquake Early Warning Systems (EEWS) are conceived as either “regi... more Most of worldwide developed Earthquake Early Warning Systems (EEWS) are conceived as either “regional” (network-based) or “on-site” (stand-alone) systems (Kanamori,2005). A regional EEWS is based on a dense sensor network covering a portion or the entirety of an area that is threatened by earthquakes. The relevant source parameters (event location and magnitude) are estimated from the early portion of recorded signals and are used to predict, with a quantified confidence, a ground motion intensity measure at a distant site where a target structure of interest is located. An on-site EEWS consists of a single sensor or an array of sensors deployed in the proximity of the target structure that is to be alerted, and whose measurements of peak amplitude and/or predominant period on the initial P-wave motion are used to predict the ensuing peak ground motion (mainly related to the arrival of S and surface waves) at the same site. Wu and Kanamori (2005,2008) have proposed a methodology for on-site early warning applications aimed at the rapid assessment of earthquake damage potential, which is based on the near source (epicentral distance smaller than 30 km), real-time measurement of the peak displacement (Pd) and a period parameter(τc) from early P-wave signals. In this note we explore the feasibility of applying a similar approach in Italy, based on the analysis of the strong motion records from the European Strong Motion Data Base and from the main events (M≥3.4) of the recent 2009, Mw 6.3 Central Italy (L’Aquila) earthquake sequence. The empirical relationships for early warning parameters (PGV,peak ground velocity, vs Pd and τc vs Magnitude) are compared with measurements on Japan and Taiwan earthquakes in a wide distance/magnitude range. By extending the observation distance to 100 km, and by comparing data from different worldwide seismic regions, we generalize the Wu and Kanamori’ method by proposing an integrated regional/on-site early warning method which can be used in the very first seconds after a moderate to large earthquake to map the most probable damaged zones. The settings of specific thresholds for Pd and τc allows to rapidly issue an earthquake alert based on a four entries, decision-table scheme. The proposed approach is suitable for Italy, where, during the last two decades, a dense network of wide dynamic-range accelerometer arrays has been deployed by the Department of Civil Protection, INGV and other regional research agencies. References Kanamori, H., 2005, Real-time seismology and earthquake damage mitigation, Annu. Rev. Earth Planet. Sci. , 33, 195-214. Wu, Y-M. and Kanamori, H., 2005. Rapid Assessment of Damage Potential of Earthquakes in Taiwan from the Beginning of P Waves. Bulletin of the Seismological Society of America,, 95(3), 1181-1185.
1] It has been shown that the initial portion of P and S wave signals can provide information abo... more 1] It has been shown that the initial portion of P and S wave signals can provide information about the final earthquake magnitude in a wide magnitude range. This observation opens the perspective for the real-time determination of source parameters. In this paper we describe a probabilistic evolutionary approach for the real-time magnitude estimation which can have a potential use in earthquake early warning. The technique is based on empirical prediction laws correlating the low-frequency peak ground displacement measured in a few seconds after the P and/or S phase arrival and the final event magnitude. The evidence for such a correlation has been found through the analysis of 256 shallow crustal events in the magnitude range M jma 4-7.1 located over the entire Japanese archipelago. The peak displacement measured in a 2-s window from the first P phase arrival correlates with magnitude in the range M = [4-6.5]. While a possible saturation effect above M ' 6.5 is observed, it is less evident in an enlarged window of 4 s. The scaling of S peaks with magnitude is instead also observed at smaller time lapses (i.e., 1 s) after the first S arrival. The different scaling of P and S peaks with magnitude when measured in a 2-s window is explained in terms of different imaged rupture surface by the early portion of the body wave signals. We developed a technique to estimate the probability density function (PDF) of magnitude, at each time step after the event origin. The predicted magnitude value corresponds to the maximum of PDF, while its uncertainty is given by the 95% confidence bound. The method has been applied to the 2007 (M jma = 6.9) Noto Hanto and 1995 (M jma = 7.3) Kobe earthquakes. The results of this study can be summarized as follows: (1) The probabilistic algorithm founded on the predictive model of peak displacement versus final magnitude is able to provide a fast and robust estimation of the final magnitude. (2) The information available after a few seconds from the first detection of the P phase at the network can be used to predict the peak ground motion at a given regional target with uncertainties which are comparable to those derived from the attenuation law. (3) The near-source S phase data can be used jointly with P data for regional early warning purposes, thus increasing the accuracy and reliability of magnitude estimation. Citation: Lancieri, M., and A. Zollo (2008), A Bayesian approach to the real-time estimation of magnitude from the early P and S wave displacement peaks,
1] We show that the low-pass filtered, peak amplitudes of initial P-and S-wave seismic signals re... more 1] We show that the low-pass filtered, peak amplitudes of initial P-and S-wave seismic signals recorded in the vicinity of an occurring earthquake source correlates with the earthquake magnitude and may be used for real-time estimation of the event size in seismic early warning applications. The earthquake size can be therefore estimated using only a couple of seconds of signal from the P-or S-wave onsets, i.e. while the rupture itself is still propagating and rupture dimension is far from complete. We argue that dynamic stress release and/or slip duration on the fault in the very early stage of seismic fracture, scales both with the observed peak amplitude and with the elastic energy available for fracture propagation. The probability that a fracture grows to a larger size should scale with the energy initially available. Citation: Zollo, A., M. Lancieri, and S. Nielsen (2006), Earthquake magnitude estimation from peak amplitudes of very early seismic signals on strong motion records, Geophys. Res. Lett., 33, L23312,
The Northern Chile gap is generally considered to the site of the next megathurst event in Chile.... more The Northern Chile gap is generally considered to the site of the next megathurst event in Chile. The Tocopilla earthquake of 14 November 2007 (Mw 7.8) and aftershock series broke the southern end of this gap. The Tocopilla event ruptured a narrow strip of 120 km of length and a width that (Peyrat et al.; Delouis et al. 2009) estimated as 30 km. The aftershock sequence comprises five large thrust events with magnitude greater than 6. The main aftershock of Mw 6.7 occurred on November 15, at 15:06 (UTM) seawards of the Mejillones Peninsula. One month later, on December 16 2007, a strong (Mw 6.8) intraplate event with slab-push mechanism occurred near the bottom of the rupture zone. These events represent a unique opportunity for the study of earthquakes in Northern Chile because of the quantity and quality of available data. In the epicentral area, the IPOC network was deployed by GFZ, CNRS/INSU and DGF before the main event. This is a digital, continuously recording network, equipped with both strong-motion and broad-band instrument. On 29 November 2007 a second network named “Task Force” (TF) was deployed by GFZ to study the aftershocks. This is a dense network, installed near the Mejillones peninsula. It is composed by 20 short-period instruments. The slab-push event of 16 december 2007 occurred in the middle of the area covered by the TF network. Aftershocks were detected using an automatic procedure and manually revised in order to pick P and S arrivals. In the 14-28 November period, we detected 635 events recorded at the IPOC network; and a further 552 events were detected between 29 November and 16 December before the slab-push event using the TF network. The events were located using a vertically layered velocity model (Husen et al. 1999), using the NLLoc software of Lomax et al. From the broadband data we estimated the moment magnitude from the displacement spectra of the events. From the short-period instruments we evaluated local magnitudes using the technique proposed by Kanamori and Jennings that consists in digitally converting the short period records of S waves to those that would have been recorded by a classical Wood-Anderson seismometer. The Tocopilla earthquake and the 16 December intermediate depth earthquake had no precursors. The results of location show no events located in the southern area of the Mejillones peninsula, confirming that this peninsula acted like a barrier to rupture. With the TF network we find the aftershocks principally in the southern area of the main rupture. This locations shows a kink in the subduction plane and we determine that it occurs near to 45km of depth . We are currently studying the focal mechanism of the main aftershocks and we search to compare the slip distribution of the earthquakes with the aftershock distribution.
The ISNet (Irpinia Seismic Network) consists of 29 six-components seismic stations deployed in a ... more The ISNet (Irpinia Seismic Network) consists of 29 six-components seismic stations deployed in a 100×70 km2 area in Southern Italy, containing the fault system that generated the 1980, M=6.9, Irpinia earthquake. The seismic stations, equipped with both accelerometers and seismometers, are able to follow both the strong and the weak motion. The computation is distributed along the network, with local control centers collecting data-streams from the single stations and performing a redundant computation of the early warning parameters (picking, location, magnitude and hazard integral). The decision is finally sent to the network control center, which is located far-away from the network itself. The performance of the infrastructure therefore depends on the latency of the information in the acquisition and in the transmission, on the computation time and on the failures of the system. We build up a virtual simulator to analyze off-line both the lead-time of the whole infrastructure and the influence of the failures of the system on the final decision. We use the simulator as a platform to test and optimize the algorithms for the computation of the early warning parameters, to investigate the optimum topology for the existing network, to asses weaknesses of the present infrastructure and define interventions plans to make the alarm decision robust. The simulator is a flow chart, in which the single blocks (acquisition, transmission, computation and decision) are either mathematical operators based on empirically estimated physical quantities or code lines running on the same platform. We evaluate the performances of the network on a synthetic database of records simulating the 1980, Irpinia earthquake and a set of real data from an earthquake of magnitude 3.0.
A prototype system for earthquake early warning and rapid shake map evaluation is being developed... more A prototype system for earthquake early warning and rapid shake map evaluation is being developed and tested in southern Italy based on a dense, wide dynamic-range seismic network (accelerometers + seismometers) under installation in the Apenninic belt region (Irpinia Seismic Network). This system forms a regional Earthquake Early Warning System consisting of a seismic sensor network covering a portion of the expected epicentral area for large earrthquakes. Considering a warning window ranging from tens of seconds before to hundred of seconds after an earthquake, several public infrastructures and buildings of strategic relevance (hospitals, gas pipelines, railways, railroads, ...) of the Regione Campania are potential test-sites for testing innovative technologies for data acquisition, processing and transmission. A potential application of an early warning system in the Campania region based on the Irpinia network, should consider an expected time delay to the first energetic S wave train varying between 14-20 sec at 40-60 km distance to 26-30 sec at about 80-100 km, from a crustal earthquake occurring in the source region. The latter is the typical time window available for mitigating earthquake effects through early warning in the city of Naples (about 2 million of inhabitants including suburbs). We have developed a method for real time earthquake location following a probabilistic approach. The earthquake location is expressed as a probability density function for the hypocenter location in 3D space based on the concept of equal differential-time (EDT). It provides a location as the maximum of a stack over quasi-hyperbolic surfaces. On each surface the difference of calculated travel-times at a pair of stations is equal to the difference of observed arrival times at the same pair of stations. For an increasing number of P-wave readings, progressively acquired in the short time after the occurrence of an earthquake, the EDT method can be generalized by introducing an evolutionary strategy which is aimed at obtaining a more and more refined estimate of the maximum probability volume as the time goes on. The real time magnitude estimate will take advantage from the high spatial density of the network in the source region and the wide dynamic range of installed instruments. Based on the offline analysis of high quality strong-motion data bases recorded in Italy and worldwide, several methods will be checked and validated , using different observed quantities (peak amplitude, dominant frequency, square velocity integral, .) to be measured on seismograms, as a function of time. Following the ElarmS methodology (Allen,2004), peak ground attenuation relations can be used to predict the distribution of maximum ground shaking, as updated estimates of earthquake location and magnitude are progressively available from the Early Warning system starting from the time of first P-wave detection. As measurements of peak ground quantities for the current earthquake become available from the network, these values are progressively used to adjust an "ad hoc" determined attenuation relation for the Campania region using the stochastic approach proposed by Boore (1993).
The Irpinia Seismic Network (ISNet) is deployed in Southern Apennines along the active fault syst... more The Irpinia Seismic Network (ISNet) is deployed in Southern Apennines along the active fault system responsible for the 1980, November 23, M s 6.9 Campania–Lucania earthquake. It is set up by 28 stations and covers an area of about 100 × 70 km2. Each site is equipped with a 1-g full-scale accelerometer and a short-period velocimeter. Due to its design characteristics, i.e., the wide dynamic range and the high density of stations, the ISNet network is mainly devoted to estimating in real-time the earthquake location and magnitude from low- to high- magnitude events, and to providing ground-motion parameters values so to get some insights about the ground shaking expected. Moreover, the availability of high-quality of data allows studying the source processes related to the seismogenetic structures in the area. The network layout, the data communication system and protocols and the main instrumental features are described in the paper. The data analysis is managed by Earthworm software package that also provides the earthquake location while custom software has been developed for real-time computation of the source parameters and shaking maps. Technical details about these procedures are given in the article. The data collected at the ISNet stations are available upon request.
1] It has been shown that the initial portion of P and S wave signals can provide information abo... more 1] It has been shown that the initial portion of P and S wave signals can provide information about the final earthquake magnitude in a wide magnitude range. This observation opens the perspective for the real-time determination of source parameters. In this paper we describe a probabilistic evolutionary approach for the real-time magnitude estimation which can have a potential use in earthquake early warning. The technique is based on empirical prediction laws correlating the low-frequency peak ground displacement measured in a few seconds after the P and/or S phase arrival and the final event magnitude. The evidence for such a correlation has been found through the analysis of 256 shallow crustal events in the magnitude range M jma 4-7.1 located over the entire Japanese archipelago. The peak displacement measured in a 2-s window from the first P phase arrival correlates with magnitude in the range M = [4-6.5]. While a possible saturation effect above M ' 6.5 is observed, it is less evident in an enlarged window of 4 s. The scaling of S peaks with magnitude is instead also observed at smaller time lapses (i.e., 1 s) after the first S arrival. The different scaling of P and S peaks with magnitude when measured in a 2-s window is explained in terms of different imaged rupture surface by the early portion of the body wave signals. We developed a technique to estimate the probability density function (PDF) of magnitude, at each time step after the event origin. The predicted magnitude value corresponds to the maximum of PDF, while its uncertainty is given by the 95% confidence bound. The method has been applied to the 2007 (M jma = 6.9) Noto Hanto and 1995 (M jma = 7.3) Kobe earthquakes. The results of this study can be summarized as follows: (1) The probabilistic algorithm founded on the predictive model of peak displacement versus final magnitude is able to provide a fast and robust estimation of the final magnitude. (2) The information available after a few seconds from the first detection of the P phase at the network can be used to predict the peak ground motion at a given regional target with uncertainties which are comparable to those derived from the attenuation law. (3) The near-source S phase data can be used jointly with P data for regional early warning purposes, thus increasing the accuracy and reliability of magnitude estimation. Citation: Lancieri, M., and A. Zollo (2008), A Bayesian approach to the real-time estimation of magnitude from the early P and S wave displacement peaks,
We study the dynamic rupture propagation of a M 6.7 intraplate earthquake that occurred 16 Decemb... more We study the dynamic rupture propagation of a M 6.7 intraplate earthquake that occurred 16 December 2007, a month after a large thrust event of Tocopilla, Chile (M 7.7). The occurrence of a slab push event after a large subduction earthquake is well explained by Coulomb stress transfer models and crack dynamics. A dense seismic network, equipped with short period and accelerometers was deployed after the event of 14 November 2007 by the Task Force of GFZ Potsdam and the University of Chile in Santiago. This network was in place on December 16 providing an excellent data set for this earthquake. We used these data to make a detailed study of rupture processes. We localized the main event of December 16 and the aftershocks that occurred within 24 h of the main event. The main event was located at 43 km depth, while the aftershocks distribution covered a circular zone of 5 to 8 km of radius centred on the main shock epicentre. The aftershocks are distributed on an almost vertical plane that agrees with one of the fault planes of the mechanism (86° dip) and all the aftershock have the same mechanism as the main event. We used nearest accelerometric records in order to do dynamic inversion, two of these accelerometers were situated right above the hypocentre. We performed a non-linear dynamic inversion based on the neighbourhood algorithm (NA) and MonteCarlo methods with an L2 norm. The data was initially filtered in the 0.05-1 Hz. The velocity model was derived from previous work by GFZ. The earthquake was modelled using finite differences on a grid of variable size. Friction was modelled by the standard Ida slip weakening friction law. At each step of the inversion more than 32 full numerical simulations are carried in parallel. These simulations have been optimized in order to reduce the computer time to a minimum. The best models that result from dynamic inversion reduced the variance by more than 30 %, these models ruptured a relatively small zone of the fault plane with a diameter of about 6 km. The rupture speed was quite slow and the energy release rate that gives the best results was of the order of 10 MJ/m2. We explored a large area of the parameter space in search for solutions with faster rupture speeds but we have failed to find them. It appears then that this earthquake, located right under the transition zone from steady to stick slip, propagated very slowly. We have confronted the results of dynamic inversion with displacement records filtered at several frequency bands up to 4 Hz. The inverted models depend on the frequency band used for the inversion because the observed and synthetic records are controlled by a large stopping phase that contains most of the radiated energy. We also study the initiation of the event on seismograms, the event started by a clear cascade process involving several initial ruptures.
1] We investigate the effect of extended faulting processes and heterogeneous wave propagation on... more 1] We investigate the effect of extended faulting processes and heterogeneous wave propagation on the early warning system capability to predict the peak ground velocity (PGV) from moderate to large earthquakes occurring in the southern Apennines (Italy). Simulated time histories at the early warning network have been used to retrieve early estimates of source parameters and to predict the PGV, following an evolutionary, probabilistic approach. The system performance is measured through the Effective Lead-Time (ELT), i.e., the time interval between the arrival of the first S-wave and the time at which the probability to observe the true PGV value within one standard deviation becomes stationary, and the Probability of Prediction Error (PPE), which provides a measure of PGV prediction error. The regional maps of ELT and PPE show a significant variability around the fault up to large distances, thus indicating that the system's capability to accurately predict the observed peak ground motion strongly depends on distance and azimuth from the fault.
1] We show that the low-pass filtered, peak amplitudes of initial P-and S-wave seismic signals re... more 1] We show that the low-pass filtered, peak amplitudes of initial P-and S-wave seismic signals recorded in the vicinity of an occurring earthquake source correlates with the earthquake magnitude and may be used for real-time estimation of the event size in seismic early warning applications. The earthquake size can be therefore estimated using only a couple of seconds of signal from the P-or S-wave onsets, i.e. while the rupture itself is still propagating and rupture dimension is far from complete. We argue that dynamic stress release and/or slip duration on the fault in the very early stage of seismic fracture, scales both with the observed peak amplitude and with the elastic energy available for fracture propagation. The probability that a fracture grows to a larger size should scale with the energy initially available. Citation: Zollo, A., M. Lancieri, and S. Nielsen (2006), Earthquake magnitude estimation from peak amplitudes of very early seismic signals on strong motion records, Geophys. Res. Lett., 33, L23312,
1] From inspection of a large set of Japanese events, we investigate the scaling of the early rad... more 1] From inspection of a large set of Japanese events, we investigate the scaling of the early radiated energy, inferred from the squared velocity integral (IV2) with the final magnitude of the event. We found that the energy can only discriminate whether the event has a magnitude larger or smaller than 5.8, and in the latter case it can allow for realtime magnitude estimation. However, by normalizing IV2 for the rupture area, the initial slip scales with the magnitude between 4 < M < 7 following the expected scaling laws. We show that the ratio between the squared peak displacement and IV2 is a proxy for the slip following the same scaling but it can be directly derived from the data, without any assumption on the rupture area. The scaling relationship between initial slip and magnitude can be used for early warning applications, when integrated in a probabilistic, evolutionary approach. Citation: Festa, G., A. Zollo, and M.
In the framework of earthquake early warning (EEW) experimentation in Southern Italy, we have dev... more In the framework of earthquake early warning (EEW) experimentation in Southern Italy, we have developed an integrated software platform called PRESTo (PRobabilistic and Evolutionary early warning SysTem). PRESTo is a methodology for regional EEW applications, based on a fully probabilistic and evolutionary approach. This means that probability density functions of source parameters and ground shaking levels are computed , which make possible to evaluate the uncertainties, and to set an appropriate level of reliability of the issued warning. Moreover, the estimates of earthquake location, size and expected ground motion are continuously updated, as new data is available in real-time from the network. The earthquake location algorithm is based on an equal differential time formulation, and a probabilistic framework for describing the hypocenter solution. The algorithm, at each time step, relies on both the information from triggered arrivals and not-yet-triggered stations. The magnitude estimation is based on a Bayesian approach, and exploits an empirical relationship that correlates the final event magnitude with the logarithm of the distance-corrected, peak ground displacement, measured over the first 2÷4 seconds of P- and S-signals. Finally, a ground motion prediction equation valid at regional scale is used to predict the peak velocities and accelerations at the target sites. The core infrastructure of the EEW system in Southern Italy is the high density, wide dynamic-range seismic network ISNet (Irpinia Seismic Network), deployed along the Apennine chain, which is frequently stricken by moderate to large magnitude damaging earthquakes. For earthquakes originating inside the ISNet seismic network, a first alarm can be issued within 4-6 seconds from the origin time, and a stable estimate with relatively small error on location (less than 5 km) and magnitude (about 0.5 magnitude units) is achieved within 10 seconds after the origin time. PRESTo can easily be configured, tailored and exported to other digital seismic networks worldwide. A simple integration with the network is granted by the widely available SeedLink communication protocol used for data acquisition. In this note we illustrate and discuss the method flexibility and exportability to different acquisition layouts through off-line simulations, using real data recorded by the Japanese K-Net and KiK-Net, and RAN, the Italian strong motion network operated by the Department of Civil Protection.
1] From inspection of a large set of Japanese events, we investigate the scaling of the early rad... more 1] From inspection of a large set of Japanese events, we investigate the scaling of the early radiated energy, inferred from the squared velocity integral (IV2) with the final magnitude of the event. We found that the energy can only discriminate whether the event has a magnitude larger or smaller than 5.8, and in the latter case it can allow for realtime magnitude estimation. However, by normalizing IV2 for the rupture area, the initial slip scales with the magnitude between 4 < M < 7 following the expected scaling laws. We show that the ratio between the squared peak displacement and IV2 is a proxy for the slip following the same scaling but it can be directly derived from the data, without any assumption on the rupture area. The scaling relationship between initial slip and magnitude can be used for early warning applications, when integrated in a probabilistic, evolutionary approach. Citation: Festa, G., A. Zollo, and M.
Most earthquake early warning systems (EEWS) developed so far are conceived as either `regional' ... more Most earthquake early warning systems (EEWS) developed so far are conceived as either `regional' (network-based) or `on-site' (stand-alone) systems. The recent implementation of nationwide, high dynamic range, dense accelerometer arrays makes now available, potentially in real time, unsaturated waveforms of moderate-to-large magnitude earthquakes recorded at very short epicentral distances (<10-20 km). This would allow for a drastic increase of the early warning lead-time, for example, the time between the alert notification and the arrival time of potentially destructive waves at a given target site. By analysing strong motion data from modern accelerograph networks in Japan, Taiwan and Italy, we propose an integrated regional/on-site early warning method, which can be used in the very first seconds after a moderate-to-large earthquake to map the most probable damaged zones. The method is based on the real-time measurement of the period (τc) and peak displacement (Pd) parameters at stations located at increasing distances from the earthquake epicentre. The recorded values of early warning parameters are compared to threshold values, which are set for a minimum magnitude 6 and instrumental intensity VII, according to the empirical regression analyses of strong motion data. At each recording site the alert level is assigned based on a decisional table with four alert levels defined upon critical values of the parameters Pd and τc, which are set according to the error bounds estimated on the derived prediction equations. Given a real time, evolutionary estimation of earthquake location from first P arrivals, the method furnishes an estimation of the extent of potential damage zone as inferred from continuously updated averages of the period parameter and from mapping of the alert levels determined at the near-source accelerometer stations. The off-line application of the method to strong motion records of the Mw 6.3, 2009 Central Italy earthquake shows a very consistent match between the rapidly predicted (within a few seconds from the first recorded P wave) and observed damage zone, the latter being mapped from detailed macroseismic surveys a few days after the event. The proposed approach is suitable for Italy, where, during the last two decades, a dense network of wide dynamic-range accelerometer arrays has been deployed by the Department of Civil Protection (DPC), the Istituto Nazionale di Geofisica e Vulcanologia (INGV) and other regional research agencies.
PRESTo (PRobabilistic and Evolutionary early warning SysTem) is a software platform for regional ... more PRESTo (PRobabilistic and Evolutionary early warning SysTem) is a software platform for regional earthquake early warning that integrates recently developed algorithms for real-time earthquake location and magnitude estimation into a highly configurable and easily portable package. The system is under active experimentation in Southern Italy on the Irpinia Seismic Network (ISNet), which is deployed in a seismogenic area that is expected to produce a large earthquake within the next 20 years. In this paper we describe the architecture of the system and test its performances using both small earthquakes (M o3.5) recorded at the ISNet and a large event recorded in Japan, through a simulation mode. The results show that, when a dense seismic network is deployed in the fault area, PRESTo can produce reliable estimates of earthquake location and size within 5-6 s from the event origin. Each estimate is provided as a probability density function, with an uncertainty that typically decreases with time: a stable solution is generally reached within 10 s from the origin.
Uploads
Papers by Maria Lancieri