Acoustic seismic impedance inversion is investigated using a migrated, stacked, reflectivity image. Basic inversion formulae are derived and sensitivity to the bandwidth of the reflectivity is analyzed. It is shown that high frequencies are responsible for the detail in the impedance image and low frequencies are responsible for the trend, where the low frequencies must be acquired from an external data type.
External data must be matched to the reflectivity image. Methods for data amplitude balancing and phase rotation are discussed. A Hilbert envelope calibration algorithm is designed to calibrate the time-depth curves between the seismic reflectivity and the external data. Both the use of well logs and stacking impedances are explored, and while well logs produce the most accurate images, a combination of the two types is also acceptable. Inversions are sensitive to the amount of external data that is added and reliable impedance images can be produced by using just 1.5-2.5 Hz of the external data.
View full article as PDF (5.12 Mb)