Locating earthquakes is an important task not only in global seismology but also for industrial applications like the monitoring of a reservoir. The industrial interest in location procedures that provide real-time hypocenter estimates mainly inspired the development of a fast and robust location procedure in this thesis.
The investigation of state of the art technologies showed that the requirement of accurate P-and S-wave arrival picking is the most time-consuming part of standard location procedures. Furthermore it was found that other modern approaches that do not use arrival times of P- and S-waves are mainly based on the principles of reverse-time wavefield extrapolation. It was found that the main disadvantage of these methods is that they require a rather dense recording network. In order to obtain the event location with a reverse-time migration based method it is necessary to check the obtained images for the 'best-focused' source image at every time step. This check can be time-consuming as well as error-prone by itself. In this thesis a location procedure was developed that does not depend on accurate arrival time picks and which also does not require dense recording networks or a focusing-selection in time as modern reverse time wavefield extrapolation do.
The kernel of the location procedure is inspired by Gaussian-beam migration and requires only a selection of time intervals around the P-wave of a detected event. The polarization information of the three-component data in the selected time interval around the P-wave is estimated and used to perform initial-value ray tracing. By weighting the energy of the signal using Gaussian beams around these rays the energy back-propagation is restricted to physically relevant regions only. A summation of the Gaussian beams over all receivers yields regions of distinct energy and the event location corresponds to the region of maximum energy in the resulting image.
The developed location procedure is applied to synthetic data as well as two case studies. The first data set was from a hydraulic fracture experiment performed in the Carthage Cotton Valley gas field (East Texas, USA). The hypocenters obtained with the migration-based procedure developed in this thesis was in a very good agreement with the hypocenters obtained with arrival-time-based standard location procedures.
Furthermore, the location method was applied to data from the San Andreas Fault Observatory at Depth (SAFOD) near Parkfield (California, USA). The horizontal projection of the obtained hypocenters clearly shows that the epicenters are distributed along the San-Andreas Fault surface trace.
The successful application to synthetic data as well as to real data obtained in two different environments shows the high potential of the developed location procedure. Moreover, without the dependence on accurately picked arrival times of P- and S-Waves the presented method is characterized by a high degree of automation which allows for much faster location than standard location procedures.
Falls Ihr Browser eine Datei nicht öffnen kann, die Datei zuerst herunterladen und dann öffnen.