Stacking three-component seismograms

Standard stacking procedures are based on the combination of many scalar traces, and can exploit the coherency of arrivals between traces using non-linear methods. Such stacks can be used with some success for three-component data but do not exploit the vector character of the seismic wave¢eld. A set of stacking methods are introduced to work directly with three-component data and include intercomponent coherency through 3|3 matrix measures. These techniques include extensions of both semblance and phase stacks. Such techniques allow the exploitation of cross-component information, which can be particularly valuable in the analysis of the S wave¢eld. Potential applications arise for the analysis of three-component records from controlled-source experiments, arrays of portable broad-band instruments and beam-forming at permanent monitoring installations.