Online detection and location estimation of earthquake events using continuous wavelet transform

Online detection of onset of an earthquake and estimation of its epicenter is of paramount importance. Manual analysis is still the final resort to characterize a seismic event and various seismic phases (waves) associated with it. In this paper, an attempt has been made at multi-scale modeling of seismic array signal to automate detection of event onset, estimation of event location and automatic/ unsupervised report generation. The technique works in wavelet domain, exploiting the non-stationary property of seismic waves. After an event is detected, a section of the data around the P onset is used to calculate the azimuth and apparent velocity of the signal. These two parameters are thereafter used to estimate the epicenter latitude-longitude. All these parameters are estimated in near real time as soon as the P-phase of seismic signal reaches the detector(s). Data from Gauribidanur seismic array is used for demonstrating the efficacy of the proposed methodology.