This thesis aims to advance methodology for acquiring and processing 3D multicomponent seismic data for shallow (< 300 m) exploration targets. The primary focus is to improve seismic resolution, and hence geological interpretation, for coal-scale targets.The coal industry is a significant contributor to the energy security and economy of Australia and the world. Conventional P-wave seismic methods are widely used in this industry, providing economic, safety and environmental benefits. There is potential for expanding these benefits by including multicomponent procedures. In this thesis, the primary focus is on converted-wave (PS) reflection, which is a logical extension to the standard approach. This has theoretical potential to provide extra geological information. It has also been proposed that there may be resolution advantages for shallower targets such as coal.A valuable starting point for understanding resolution in shallow P and PS reflection is via visco-elastic finite-difference simulation. This provides a useful indication of the reflection response of different targets, and can include the influence of different processing flows. For typical coal-scale environments, modelling with reasonable anelasticity assumptions suggests that PS resolution is unlikely to be superior to P resolution, even with idealised acquisition and processing. In real-world situations, achieving good PS resolution may be even more challenging. There are a number of factors across the acquisition and processing flow which incrementally influence resolution.Survey design is intrinsically more complex for PS surveys than for P, primarily because of ray-path asymmetry. In addition, phase and amplitude effects require careful analysis, and finite-difference modelling provides a useful tool. Such modelling suggests that for shallow surveys in particular, it may be possible to incorporate longer relative offsets, compared to the petroleum scale. An examination of bin fold, and offset/azimuth distribution, suggests that the natural bin size of a PS survey is mostly dependent on the receiver spacing, and is generally larger than for P waves. This favours the use of higher receiver densities for multicomponent surveys. These observations can be more important at the shallow scale, again because of greater ray-path asymmetry and potentially higher V P /V S value.One of the most critical steps in the shallow PS processing sequence is correcting for the S-wave receiver statics. Three relatively conventional approaches have been evaluated. A surface-consistent inversion approach has been shown to fail for shallow targets, or in the presence of strong noise. For our 3D data set, PPS refraction analysis provided the preferred solution. In other cases where PPS refractions are poorly defined, our robust statistical approach may be useful for determining short-wavelength statics, although additional long-wavelength control would then be needed.ii In exploration seismology, surface waves are commonly considered to be noise. However, the d...
