The focus of this thesis is to determine the most accurate processes and operators to properly image PS data.
Converted-wave seismic data have intrinsic characteristics that make their processing different than that of conventional PP seismic data. For a simple horizontal layer embedded in a constant-velocity medium the concept of Common Midpoint Gathers (CMP), does not hold for PS data; therefore, each trace represents a different point in the subsurface. For PS data, the most common representation is known as Common Reflection/Conversion-point Gathers (CRP/CCP). Figure shows an schematic representation for both the CMP and the CCP/CRP definition. The top panel represents the raypath geometry for a converted-mode ray in a CMP distribution; because of the difference between the P-velocity and the S-velocity, each source-receiver ray illuminates a different point in the subsurface. The bottom panels presents the raypath for the converted-mode ray distribution in CRP/CCP geometry. The appropriate knowledge of the P-wave velocity, the S-wave velocity, and the physics of wave propagation, makes it possible to arrange each source-receiver ray to illuminate the same point in the subsurface, this point is also known as the image point.
CCP stacking is a processing technique for converted-wave data that approximates the correction from the CMP to the CCP domain; however, the final result is in the time domain and the knowledge of the P and S velocity models is required.
Another characteristic of converted-wave data is that the polarity flips along the reflection hyperbola. This polarity change is an intrinsic property of the shear-wave displacement Danbom and Domenico (1988). In a constant-velocity medium, the vector displacement field produces opposite movements in the receivers at either side of the intersection of the normal ray with the surface (Figure ). This leads to the polarity flip along the same reflection. Figure presents a characteristic Common Shot Gather (CSG) for the PS section of the 2-D OBS Mahogany data set in the Gulf of Mexico. This is an example of the polarity flip characteristic of converted-wave data.
Figure 3 Polarity inversion in converted waves seismic data. +g and -g correspond to positive and negative polarity in a common shot gather. The vector displacement field produces opposite movements at either size of the intersection of the normal ray with the surface.
Figure 4 PS Common Shot Gather (CSG) from the 2-D Mahogany data set in the Gulf of Mexico. Note the polarity flip along the same hyperbolas.
In a complex velocity medium, the normal ray determines the location of the polarity flip. For flat reflectors in v(z) media, and in areas with constant P-to-S velocity ratio (), the normal incidence ray emerges at the surface at zero-offset. However, in general, the P and S raypaths corresponding to the normal-incidence (zero-amplitude) ray will not necessarily emerge at the surface at the zero-offset point. Figure illustrates this for the case of a dipping layer and a non-constant .
Figure 5 Polarity flip problem for a dipping layer and a non-constant .
For 3-D OBS data, the irregularity in the acquisition geometry is an additional problem for the imaging of converted-wave data. Irregular geometries are a serious impediment for accurate subsurface imaging Beasley (1994); Chemingui (1996); Gardner and Canning (1994). Irregularly sampled data affect the image with amplitude artifacts and phase distortions if the missing data are assumed to be zero traces. Irregular geometry problems are more acute in cases in which the amplitude information is one of the main goals of study. Typical OBS seismic data acquisition presents processing problems similar to those of land data. Gardner and Canning (1994) demonstrate some of the effects of irregular sampling on 3-D prestack migration, through synthetic examples using real 3-D land-acquisition geometry. Figure illustrates this problem. This figure presents the source and receiver distribution for 3-D OBS data set in the Alba oil field in the North Sea. The gap in the source distribution, panel (a), is due to an oil production platform. These gaps in the data result in artifacts in the final 3-D image.