Multicomponent seismic surveys in a complex setting: Catatumbo, Colombia

Saúl Ernesto Guevara, Gary F. Margrave, Robert R. Stewart

Compared with single component seismic data, multicomponent data have potentially higher information content even in structurally complex areas. Converted wave data are an obvious potential added value, but also important are the sensitivity of the 3C sensors to the vector properties of the data, which may improve the P-wave information. Data processing which takes into account the topography illustrates these points. Multicomponent seismic surveys acquired in a Northern Andes foothills setting are presented here. The survey is in a location that includes rough topography and folded geology. The seismic data were acquired with the 3C sensors oriented vertically, and it can be assumed that waves arrived normal to the surface. Consequently, the rotation of the radial and vertical data into a direction normal to the topography would better estimate the P-waves. Radial filtering was used to attenuate coherent noise, Gabor deconvolution was applied, and finally, a rotation was carried out to obtain the P-wave normal to topography. Differences can be observed in the stacked sections with and without correction; however, the result can not be considered definitive. Information about the near surface layer suggests the possibility of non-normal wavefront incidence, which implies the need to consider the free surface effect on this wave-mode separation. Converted waves must also be given similar consideration. Besides the usual difficulties related to converted wave processing, such as static corrections, velocity analysis and trace binning, the geologic structure presents a particular challenge for converted waves, which was illustrated by another multicomponent data set previously acquired in the area. Prestack depth migration appears to be an appropriate tool to overcome these shortcomings of converted wave imaging.