Study And Application On Earthquake Early Warning

Earthquake early warning (EEW) system provides a few seconds to tens of seconds of warning time for impending ground motions due to a large earthquake and allows for mitigation measures in the short term. It is a new technology and new measure for seismic hazard mitigation in last twenty years and has the potential to reduce fatalities, casualties, costs. This paper focus on some key technologies for EEW include automatic picking phase arrival precisely, earthquake location, magnitude determination, the strong motion maps estimation of earthquake early warning target area such as seismic principal parameters and engineering seismology parameters. Real-time online operation is considered for all algorithms. An EEW frame is constructed rely on the results of this paper and the Fujian Digital Seismograph Network. The contents are summarized as follows:(1) A multi-step method has been developed for the automatic P-wave and S-wave accurate detection base on some common methods include STA/LTA, polarization analysis, and Akaike information criteria (AIC). The applicability and stability were improved. Also a new interference signal elimination method based the Delaunay triangulation has been developed and the detecting reliability was improved. We test our method on regional earthquake data from Fujian region. All results prove the expected low false rate, high timing accuracy and good detection probability for both P-wave and S-wave.(2) A real-time continuous earthquake location method has been developed base on Voronoi diagram, hyperbolic curve of P-wave arrival time difference of two stations, circle of S-wave and P-wave arrival time difference and not yet arriving S-wave, and epicenter azimuth. The influence of focal depth for different seismic phase arrive time is also considered. From the time that the first P-wave arrived to the nearest station to the time that three stations had detected P-wave and the first station had detected the S-wave, six cases are classified and the location process is continuous. We test all case on regional earthquake data from Fujian region using automatic P-wave and S-wave detection results. An off-line test proves that the location results are satisfying for EEW.(3) A new method to determinate the magnitude has been developed based on the effective cumulative rising width of first radial P-wave displacement pulse. As comparing with the most commonly used average'predominant period'Ï„c method it has the same precision. Farther for improving the precision and the stability for magnitude, another new real-time continuous method has been developed. Various period parameters and amplitude parameters for different frequency band include peak acceleration, peak velocity, peak displacement, predominant period, the ratio of peak velocity and peak acceleration Vmax/Amax, and effective cumulative rising width for every second after P-wave arrived were considered. All the parameters were synthesized using Artificial Neural Networks (ANN). So, various characters of every initial P-wave segment were considered synthetically and continuously. We test our method on the bedrock records of digital strong-motion seismograph of Japan (KiK-net). We find that the magnitude of even the single station for every second after the P-wave arrived is satisfying for EEW.(4) As similar principle of magnitude real-time continuous determination, we have developed a method to estimate the imminent ground shaking (PGA, PGV, PGD) for every second after P-wave arrived based on ANN. It can be applied to on-site warning directly. Also we present a method to continuous evaluated ground motion maps using the result of estimated value exist station.(5) An EEW frame is constructed rely on the study results of this paper and the Fujian digital Seismograph Network and the application of EEW in China is prospected.