Two-Component Recording with a P-Wave Source to Improve Seismic Resolution

Abstract

In two-component seismic observations with vertical and in-line horizontal geophones, the compressional (P-) wave amplitudes, as well as the vertically polarized shear (SV-) wave amplitudes, are observed on both vertical and horizontal geophones. In our case, we use a P-wave source, while the SV waves are the result of mode conversion. The mode-conversion mechanism considered here is related to the near-surface layers, i.e. we have a P-leg from the source and mode conversion at/in the weathered layer. The resulting SV waves therefore will show lateral variations because the elastic parameters of the near-surface layers vary along the seismic line, but these variations will be consistent at the surface. This effect is demonstrated by a synthetic example based on elastic parameters representative of the actual seismic line being considered. To separate the individual P and SV arrivals, we apply a two-dimensional convolution filter designed to meet the wavenumber-frequency (k-f) domain transfer function for P-SV separation which can be derived from the k-f domain geophone-receiving characteristic and the near surface P- and S-wave velocities. The reason for P-SV separation filtering in the offset-traveltime (X-T) domain instead of directly filtering in the k-f domain, is a great saving in computer time, as X-T filters, with few coefficients, can be used. In this paper, after a short summary of the k-f domain P-SV separation filters and their transformation to the X-T domain, we apply the X-T filters to synthetic data in order to demonstrate that our design is correct. We also work on actual data and discuss the problems being faced, which mainly originate from the different geophone groups and, as a consequence, the different scalings of vertical and horizontal geophones. The main advantage of two-component seismic observations is two-fold: firstly, a clean P-wave section is obtained (SV-energy arriving at the receivers is cancelled by applying the foresaid separation filter) and, secondly, we obtain an additional SV-wave section at almost no cost to data acquisition. These two sections contribute towards distinguishing between true and false bright spots, so they are used as direct hydrocarbon indicator tools.