Abstract
This study is motivated by the necessity to quantitatively characterize subtuned reservoirs. The conventional autocorrelation-based spectral-decomposition technique uses frequency notches to calculate vertical traveltime thickness of a layer of dipole reflectivity. Those notches tend to move outside the usable frequency band of the seismic data as the layer exceeds the tuning threshold of the wavelet. Assuming wavelet stationarity and nondipole reflectivity, a similar analysis performed on a crosscorrelation between an intercept and gradient trace extends the resolution limits to one-half the tuning threshold. That is a major improvement; however, many economic reservoirs still do not meet the half-tuning requirement. Such thin reservoirs led to the development of an optimization scheme. This approach, which does not require any wavelet stationarity or reflectivity assumptions, theoretically is not limited by the thickness of the target interval. The optimization scheme was applied successfully to a marine seismic survey in an attempt to estimate the traveltime thickness of a chalk reservoir.