Predicting geometry and stacking pattern of thin beds by interpreting geomorphology and waveforms using sequential stratal-slices in the Wheeler domain

Abstract

I used a seismic sedimentology-based approach for interpreting the spatial geometry and stacking pattern of superimposed, seismically thin beds. The method was applicable to conventional low-frequency seismic data and required knowledge of the basic shape of the seismic wavelet and lithology-impedance relationship (impedance model). It was primarily aimed at those beds that were spatially extensive enough to be resolved horizontally, but so thin as to only be detected vertically with a given bandwidth, and it assumed that the data had been converted to stratal slices in the relative geologic time (or Wheeler) domain. A simple one-bed model illustrated that a thin-bed depositional system can be characterized by a seismic-geomorphologic pattern of the same spatial shape on sequential relative geologic-time (stratal) slices, but the amplitude, phase, and polarity would vary depending upon the known (estimated) seismic wavelet. This phenomenon could be captured and evaluated in the thin bed’s response window (RW) in the Wheeler domain. If multiple thin-bed units were present in the RW, the seismic responses from vertically adjacent units would interfere with the “true” seismic-geomorphologic pattern of any single thin-bed unit. The composite waveform for each of the units could be restored in variable quality, depending on its geomorphologic character, thickness, and stratigraphic position. Relative traveltime differences of thin-bed waveforms revealed the depositional history (stacking pattern) of the thin-bed sequence. A field-data test confirmed that within a stratigraphic interval of a composite seismic event ([Formula: see text] or 15 m) at least two, and possibly three, thin fluvial channel sandstones (2–6 m each) could be identified and their spatial localities and distribution could be unraveled.