, the deliverables include algorithms and codes for processing and inversion of multicomponent seismic data in anisotropic media and for modeling of seismic signatures of fracture systems. Most of this software is available from CWP as part of Seismic Unix (SU) -a package of programs for Unix-based machines.
SUMMARY OF RESEARCH RESULTS
New method for processing and inversion of multicomponent seismic dataDue to the high cost of shear-wave excitation, the difficulty of placing sources on the sea floor, and often poor quality of SS-wave data, it has become a practice in reflection seismology to replace pure SS reflections with mode-converted PS-waves. Converted waves are of particular importance in offshore seismic, where they represent the only available type of shear energy. For a number of exploration scenarios, PS-waves provide valuable information about the subsurface structure and anisotropic medium parameters that cannot be inferred from conventional PP-wave data. Kinematics and waveforms of reflected PS-waves, however, possess such undesirable features as moveout asymmetry, reflection point dispersal and polarity reversal, which preclude application of conventional velocity-analysis methods to mode conversions.Rather than using PS-waves directly, we developed a method for constructing SSwave reflections (if they are not physically excited in the survey) from PP and PS data. The original version of this technique ) is based on picking of PP and PS traveltimes on prestack data and identification (correlation) of the PP and PS events from the same interfaces. The key idea of the method is to match the reflection slopes (horizontal slownesses) on common-receiver PP and PS gathers. This matching allows us to find the coordinates of receivers that record PPand PS-waves reflected at exactly the same (albeit unknown) subsurface points and to determine the traveltime of the SS primary reflection as a simple combination of the PP and PS times. The constructed SS-wave moveout can then be processed by velocity-analysis methods designed for pure reflection modes (see below).To avoid traveltime picking on prestack data, we generalized the above algorithm (often called the "PP+PS=SS" method) to operate directly with recorded traces of PP-and PS-waves (Grechka and Dewangan, 2003). By designing a special convolution of PP-and PS-wave seismograms, it is possible to compute full-waveform SS data well-suited for moveout inversion and, potentially, for AVO (amplitude variation with offset) analysis.The PP+PS=SS method has the following attractive features: 1. Although it is necessary to establish correlation between the PP and PS sections, no information about the velocity field or anisotropic parameters is required to obtain SS-waves with the correct kinematics from the PP and PS data. 2. The moveout of the output SS-waves is symmetric with respect to zero offset and can be processed by algorithms designed for pure-mode reflections. 3. The reflection-point dispersal of PS-waves has no influence on the generated SS data. 3....