Earthquake Early Warning Technology Progress in Taiwan

2009 ◽  
Vol 4 (4) ◽  
pp. 530-538 ◽  
Author(s):  
Kuo-Liang Wen ◽  
◽  
Tzay-Chyn Shin ◽  
Yih-Min Wu ◽  
Nai-Chi Hsiao ◽  
...  
Keyword(s):  
Real Time ◽  
Early Warning ◽  
Monitoring Network ◽  
Earthquake Early Warning ◽  
P Waves ◽  
Strong Base ◽  
Warning Messages ◽  
Online Experimentation ◽  
Near Future ◽  
Remarkable Progress

The dense real-time earthquake monitoring network established in Taiwan is a strong base for the development of the earthquake early warning (EEW) system. In remarkable progress over the last decades, real-time earthquake warning messages are sent within 20 sec after an event using the regional EEW system with a virtual subnetwork approach. An onsite EEW approach using the first 3 sec of P waves has been developed and under online experimentation. Integrating regional and onsite systems may enable EEW messages to be issued within 10 sec after an event occurred in the near future. This study mainly discusses the methodology for determining the magnitude and ground motion of an event.

scholarly journals Status of Earthquake Early Warning in Switzerland

2021 ◽  
Vol 9 ◽  
Author(s):  
Frédérick Massin ◽  
John Clinton ◽  
Maren Böse
Keyword(s):  
Real Time ◽  
Open Source Software ◽  
Early Warning ◽  
Seismic Network ◽  
Earthquake Early Warning ◽  
Fault Rupture ◽  
Time Data ◽  
Origin Time ◽  
P Waves ◽  
Finite Fault

The Swiss Seismological Service (SED) at ETH has been developing methods and open-source software for Earthquake Early Warning (EEW) for more than a decade and has been using SeisComP for earthquake monitoring since 2012. The SED has built a comprehensive set of SeisComP modules that can provide EEW solutions in a quick and transparent manner by any seismic service operating SeisComP. To date, implementations of the Virtual Seismologist (VS) and Finite-Fault Rupture Detector (FinDer) EEW algorithms are available. VS provides rapid EEW magnitudes building on existing SeisComP detection and location modules for point-source origins. FinDer matches growing patterns of observed high-frequency seismic acceleration amplitudes with modeled templates to identify rupture extent, and hence can infer on-going finite-fault rupture in real-time. Together these methods can provide EEW for all event dimensions from moderate to great, if a high quality, EEW-ready, seismic network is available. In this paper, we benchmark the performance of this SeisComP-based EEW system using recent seismicity in Switzerland. Both algorithms are observed to be similarly fast and can often produce first EEW alerts within 4–6 s of origin time. In real time performance, the median delay for the first VS alert is 8.7 s after origin time (56 earthquakes since 2014, from M2.7 to M4.6), and 7 s for FinDer (10 earthquakes since 2017, from M2.7 to M4.3). The median value for the travel time of the P waves from event origin to the fourth station accounts for 3.5 s of delay; with an additional 1.4 s for real-time data sample delays. We demonstrate that operating two independent algorithms provides redundancy and tolerance to failures of a single algorithm. This is documented with the case of a moderate M3.9 event that occured seconds after a quarry blast, where picks from both events produced a 4 s delay in the pick-based VS, while FinDer performed as expected. Operating on the Swiss Seismic Network, that is being continuously optimised for EEW, the SED-ETHZ SeisComP EEW system is achieving performance that is comparable to operational EEW systems around the world.

Source Process-Related Delays in Earthquake Early Warning for Example Cases in Greece

2021 ◽  
Author(s):  
Nikolaos Vavlas ◽  
Anastasia A. Kiratzi ◽  
Zafeiria Roumelioti
Keyword(s):  
Early Warning ◽  
Monitoring Network ◽  
Earthquake Early Warning ◽  
Source Process ◽  
Spatial Density ◽  
Intermediate Depth ◽  
National Monitoring ◽  
Gulf Of Corinth ◽  
Crete Island ◽  
Finite Fault

ABSTRACT We explore a hypothetical zero-latency earthquake early warning (EEW) system in Greece, aiming to provide alerts before warning thresholds of the intensity of ground motion are exceeded. Within the seismotectonic context of Greece, both shallow- and intermediate-depth earthquakes (along the Hellenic subduction zone) are plausible and, thus, examined. Using regionally applicable attenuation relations, we combine and adjust the methodologies of Minson et al. (2018) and Hoshiba (2020) to examine what are the minimum magnitudes required to invoke the warning thresholds at the user site. With simple modeling, we examine how fast an alert can be issued and what is the available warning time when taking into account delays due to finite-fault rupture propagation, alongside other delays. These computations are merged with delays introduced due to the present-day configuration of the Greek national monitoring network (varying spatial density of permanent monitoring stations). This approach serves as a tool to assess the feasibility of an EEW system at specific sites and to redesign the national permanent monitoring network to serve such a system more effectively (we provide results for four sites.). Warning times for on-land crustal earthquakes are found to be shorter, whereas for intermediate-depth earthquakes in Greece an EEW system is feasible (provides warning times of several tens of seconds at large cities, e.g., on Crete Island) even with the current configuration of the national monitoring network, which is quite sparse in the southern part of the country. The current network configuration also provides sufficient early warning (e.g., of the order of 10 s for a warning threshold of 0.05g) at the center of Athens from earthquakes of the eastern Gulf of Corinth—a zone posing elevated hazard in the broader area of the Greek capital. Several additional assumptions and factors affecting the operability of an EEW system in Greece (i.e., source process complexity and uncertainty in attenuation laws) are also discussed.

Too-Late Warnings by Estimating Mw: Earthquake Early Warning in the Near-Fault Region

2020 ◽  
Vol 110 (3) ◽  
pp. 1276-1288
Author(s):  
Mitsuyuki Hoshiba
Keyword(s):  
Real Time ◽  
Ground Motion ◽  
Early Warning ◽  
Strong Motion ◽  
Earthquake Early Warning ◽  
Motion Generation ◽  
Near Fault ◽  
Time Gap ◽  
Fault Region ◽  
Strong Ground

ABSTRACT Earthquake early warning (EEW) systems aim to provide advance warnings of impending strong ground shaking. Many EEW systems are based on a strategy in which precise and rapid estimates of source parameters, such as hypocentral location and moment magnitude (Mw), are used in a ground-motion prediction equation (GMPE) to predict the strength of ground motion. For large earthquakes with long rupture duration, the process is repeated, and the prediction is updated in accordance with the growth of Mw during the ongoing rupture. However, in some regions near the causative fault this approach leads to late warnings, because strong ground motions often occur during earthquake ruptures before Mw can be confirmed. Mw increases monotonically with elapsed time and reaches its maximum at the end of rupture, and ground motion predicted by a GMPE similarly reaches its maximum at the end of rupture, but actual generation of strong motion is earlier than the end of rupture. A time gap between maximum Mw and strong-motion generation is the first factor contributing to late warnings. Because this time gap exists at any point of time during the rupture, a late warning is inherently caused even when the growth of Mw can be monitored in real time. In the near-fault region, a weak subevent can be the main contributor to strong ground motion at a site if the distance from the subevent to the site is small. A contribution from a weaker but nearby subevent early in the rupture is the second factor contributing to late warnings. Thus, an EEW strategy based on rapid estimation of Mw is not suitable for near-fault regions where strong shaking is usually recorded. Real-time monitoring of ground motion provides direct information for real-time prediction for these near-fault locations.

End-to-end PGA estimation for earthquake early warning using transformer networks

2020 ◽  
Author(s):  
Jannes Münchmeyer ◽  
Dino Bindi ◽  
Ulf Leser ◽  
Frederik Tilmann
Keyword(s):  
Neural Network ◽  
Real Time ◽  
Ground Motion ◽  
Early Warning ◽  
Earthquake Early Warning ◽  
P Wave ◽  
Lead Times ◽  
Ground Shaking ◽  
Target Sites ◽  
End To End

<p>The key task of earthquake early warning is to provide timely and accurate estimates of the ground shaking at target sites. Current approaches use either source or propagation based methods. Source based methods calculate fast estimates of the earthquake source parameters and apply ground motion prediction equations to estimate shaking. They suffer from saturation effects for large events, simplified assumptions and the need for a well known hypocentral location, which usually requires arrivals at multiple stations. Propagation based methods estimate levels of shaking from the shaking at neighboring stations and therefore have short warning times and possibly large blind zones. Both methods only use specific features from the waveform. In contrast, we present a multi-station neural network method to estimate horizontal peak ground acceleration (PGA) anywhere in the target region directly from raw accelerometer waveforms in real time.</p><p>The three main components of our model are a convolutional neural network (CNN) for extracting features from the single-station three-component accelerograms, a transformer network for combining features from multiple stations and for transferring them to the target site features and a mixture density network to generate probabilistic PGA estimates. By using a transformer network, our model is able to handle a varying set and number of stations as well as target sites. We train our model end-to-end using recorded waveforms and PGAs. We use data augmentation to enable the model to provide estimations at targets without waveform recordings. Starting with the arrival of a P wave at any station of the network, our model issues real-time predictions at each new sample. The predictions are Gaussian mixtures, giving estimates of both expected value and uncertainties. The model can be used to predict PGA at specific target sites, as well as to generate ground motion maps.</p><p>We analyze the model on two strong motion data sets from Japan and Italy in terms of standard deviation and lead times. Through the probabilistic predictions we are able to give lead times for different levels of uncertainty and ground shaking. This allows to control the ratio of missed detections to false alerts. Preliminary analysis suggest that for levels between 1%g and 10%g our model achieves multi-second lead times even for the closest stations at a false-positive rate below 25%. For an example event at 50 km depth, lead times at the closest stations with epicentral distances below 20 km are 6 s and 7.5 s. This suggests that our model is able to effectively use the difference between P and S travel time and accurately assess the future level of ground shaking from the first parts of the P wave. It additionally makes effective use of the information contained in the absence of signal at other stations.</p>

Ground Motion Estimation Using Front Site Wave Form Data Based on RVM for Earthquake Early Warning

2015 ◽  
Vol 10 (4) ◽  
pp. 667-677
Author(s):  
Yincheng Yang ◽  
◽  
Masato Motosaka ◽  
Keyword(s):  
Ground Motion ◽  
Early Warning ◽  
Warning System ◽  
Earthquake Early Warning ◽  
Peak Ground Velocity ◽  
Wave Form ◽  
Source Information ◽  
Ground Acceleration ◽  
P Waves ◽  
Earthquake Early Warning System

The use of the earthquake early warning system (EEWS), one of the most useful emergency response tools, requires that the accuracy of real-time ground motion prediction (GMP) be enhanced. This requires that waveform information at observation points along earthquake wave propagation paths (hereafter, front-site waveform information) be used effectively. To enhance the combined reliability of different systems, such as on-site and local/regional warning, we present a GMP method using front-site waveform information by applying a relevant vector machine (RVM). We present methodology and application examples for a case study estimating peak ground acceleration (PGA) and peak ground velocity (PGV) for earthquakes in the Miyagi-Ken Oki subduction zone. With no knowledge of source information, front site waveforms have been used to predict ground motion at target sites. Five input variables – earthquake PGA, PGD, pulse rise time, average period and theVpmax/Amaxratio – have been used for the first 4 to 6 seconds of P-waves in training a regression model. We found that RVM is a useful tool for the prediction of peak ground motion.

scholarly journals Operational real-time GPS-enhanced earthquake early warning

2014 ◽  
Vol 119 (10) ◽  
pp. 7944-7965 ◽  
Author(s):  
R. Grapenthin ◽  
I. A. Johanson ◽  
R. M. Allen
Keyword(s):  
Real Time ◽  
Early Warning ◽  
Earthquake Early Warning
Sign in / Sign up

Export Citation Format

Share Document