Fast location of seismic sources using migration techniques
Abstract
In this work we present a new approach for precise location of seismic sources using a Gaussian-beam-type-migration of multicomponent data. Standard location procedures require the identification of seismic phases and the picking of P- and S-wave arrival times. They are characterized by a strong dependence on the picking accuracy of arrival times and a low degree of automation. In particular, manual picking in the case of large data sets makes the process of location very slow, expensive and not practicable. Recently, we have developed a fast and semi-automated location procedure that takes into account the full waveform of three component data in a preselected time interval. We estimate the polarization for each time sample from the three components of the signal and perform initial-value ray tracing using this polarization information as the starting direction of the ray. If the time sample is part of the signal then the ray is traced towards the hypocenter and if not the ray is traced in a random direction. A wave field back-propagation along the ray is performed by assigning image values to every grid point of the model. These image values are obtained by weighting the energy of the time sample with a Gaussian-beam-type factor. Significant energy-values are concentrated within the width of the Gaussian beam whereas outside these values decrease rapidly. Hence, the weighting restricts the back-propagation along the rays. The summation of all obtained image values of the selected time interval over all receivers yields regions of distinct stacked energy. Thus, the region in the final image with maximum stacked energy is assumed to represent the hypocenter of the seismic event. Our approach was tested on various synthetic data in order to analyze different receiver geometries with different signal-to-noise ratios and their effects on the location accuracy. We have successfully applied our location method to real data of a hydraulic fracturing experiment from the Cotton Valley tight gas reservoir (East Texas, USA) where microearthquakes with very small magnitudes (M~<~0) were recorded.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2005
- Bibcode:
- 2005AGUFM.S33A0306R
- Keywords:
-
- 0520 Data analysis: algorithms and implementation;
- 0910 Data processing;
- 3285 Wave propagation (0689;
- 2487;
- 4275;
- 4455;
- 6934);
- 7215 Earthquake source observations (1240);
- 7290 Computational seismology