From Location to Stress Tensor Estimation: Automatically vs Manually Analyzed Data

An integrated methodology for the solution of the event detection problem and the accurate estimation of P- and S- phase arrival times has been developed and successfully applied in a passive seismic tomography survey. In such applications, the recorded seismic events must be analyzed, in order to get accurate locations and travel-times for tomography, but the analysis must also take into consideration those characteristics of the seismic signal (polarities) that will provide information on the events’ focal mechanisms and, consequently, the stress regime dominating a study area. All this information, which is very useful in the context of any Exploration and/or Production project, is more sensitive to the accuracy of the P-phase arrival estimations than simple event location and this fact has been seriously taken into consideration both in designing and in developing the proposed methodology. The above mentioned technique consists of a detection algorithm which constitutes an improved version of the classical STA/ LTA algorithm and an automatic picking algorithm based on the statistical characteristics of the seismic signals. Its effectiveness is confirmed by its application to a massive microseismic dataset, derived by the aftershock sequence following two major seismic events that occurred in Cephalonia Island in Greece.