Reflections are typically interference patterns caused by the complex interaction between source wavelet and impedance contrasts. These interference patterns are difficult to directly correlate with observed downhole changes in velocity and density (lithology), except in cases such as the sediment/water and sediment/basement interfaces and cases where a homogenous, relatively thick layer lies within a uniform sediment matrix (e.g., thick chert or sill). Downhole logs obtain in situ measurements every 0.1524 m, regardless of core recovery, of velocity and density data used for producing synthetic seismograms. A log-generated synthetic seismogram can accurately tie the cored stratigraphy (depth) to seismic reflection images in seconds TWT if the synthetic seismogram matches the observed seismic character.
Velocity and density data from downhole measurements were used to calculate reflection coefficients that were convolved with a source wavelet to produce synthetic seismograms using software provided by Dr. Sun of the Borehole Research Group, Lamont-Doherty Earth Observatory (LDEO). A 10-trace average of the seafloor reflection, including the bubble pulse at ~0.1 s TWT, which represents the true source wavelet, is used to construct the synthetic seismogram (Fig. 5). Except for the addition of the bubble pulse, this wavelet is nearly identical to a minimum phase Ricker wavelet of 21 Hz. The seismic data and synthetic seismogram are both displayed with a band-pass filter (5-50 Hz) and an AGC with a 500-ms window. The synthetic seismogram, displayed as five adjacent traces, is placed adjacent to observed seismic data at the closest approach to Site 999 and is shifted until the best visual correlation with reflections is achieved. A reasonable match occurs with both amplitudes and traveltimes of reflections between observed and synthetic seismograms within the CB4 and CB5 seismic units.