Narrow-band spectral analysis and thin-bed tuning

Abstract Running window seismic spectral decomposition has proven to be a very powerful tool in analyzing difficult-to-delineate thin-bed tuning effects associated with variable-thickness sand channels, fans, and bars along an interpreted seismic horizon or time slice. Unfortunately, direct application of spectral decomposition to a large 3-D data set can result in a rather unwieldy 4-D cube of data. We develop a suite of new seismic attributes that reduces the input 20–60 running window spectral components down to a workable subset that allows us to quickly map thin-bed tuning effects in three dimensions. We demonstrate the effectiveness of these new attributes by applying them to a large spec survey from the Gulf of Mexico. These two thin-bed seismic attributes provide a fast, economic tool that, when coupled with other attributes such as seismic coherence and when interpreted within the framework of geomorphology and sequence stratigraphy, can help us quickly evaluate large 3-D seismic surveys. Ironically, in addition to being more quantitatively linked to bed thickness, the thin-bed attributes described here allow us to analyze thicker features than the conventional instantaneous and response frequencies, which cannot calculate the spectral interference between two well-separated reflectors.