Rapid Estimation Of Magnitude And Epicentral Distance Using Envelope Waveform Of The Initial Phrase Of P-wave For Earthquake Early Warning

Earthquake early warning (EEW) system is one of the most effective approachesfor mitigating the seismic hazard of high speed railways, which can rapidly provideearthquake information for emergency management so as to reduce the direct andsecondary seismic disasters.For the urgent need of high speed railway early warning system, this paperfocuses on the rapid estimation on epicenter distance and magnitude based on therecords of single station from K-net network in Japan. A probabilistic relationshipbetween the amplitude increasing rate and epicenter distance is established based onthe envelop information on initial phase of P wave from strong motion records ofsingle station. And then the study on the magnitude estimation by multi-parameterregression analysis is explored based on the P-wave envelope area and amplitudeincreasing rate. An overview of the research is as follows:1. The development progress and current status of B-Δ method is introduced, inwhich the P wave amplitude increasing rate is used to estimate the epicenter distance.Then a statistical regression relationship of epicenter distance and amplitude increasingrate is established according to the K-net strong motion observation data of Japan, andthe error analysis of epicenter distance estimation by the B-Δ method is done.2. Two characteristic parameters of P wave for magnitude rapid estimation areproposed: the acceleration envelope areaS aand velocity envelope areaS v, and therelevance of the two parameters with magnitude is analyzed. Then the linear regressionrelationships ofS aand magnitude,S vand magnitude are given respectively based onthe strong motion records of K-net network in Japan. Finally, the method is comparedwith the magnitude estimation method by displacement amplitudePd, and the resultsshow that the magnitude estimation by velocity envelope areaS vis in accordance withthat by displacement amplitudePd.3. The regression relationships of multi-parameter and magnitude are establishedbased on different regression parameters such as velocity amplitudeVm ax, displacementamplitudePd, acceleration envelope areaS a, velocity envelope areaS v., envelopeparameter of P wave A and amplitude increasing rate B. Then the variance analysis ofdifferent multi-parameter regression relationships are done. The results show that themagnitude estimation methods by velocity parametersS aandVm ax, displacementparametersS vandPd, A and B can effectively reduce the estimation error and mayimprove the problem of magnitude underestimation during large earthquakes.