A synthetic seismogram can constrain depth-traveltime correlations if the synthetic seismogram closely matches the observed seismic character. Reflections, however, are most often interference patterns caused by the source wavelet character and combined impedance contrasts associated with closely spaced downhole changes in porosity and lithology. Exceptions include reflections created by large and abrupt impedance contrasts between extended intervals of relatively uniform physical properties where the reflection appears as a recognizable reproduction of the source wavelet scaled by the reflection coefficient (e.g., the seafloor). Reflections created at such interfaces are especially useful for constraining depth-traveltime correlations. Reflection amplitude is directly related to the size of the impedance contrast and can also be an important constraint when matching synthetic and observed seismograms. At Site 1149, a large contrast in physical properties at the interface between the thick (180 m) section of relatively homogenous pelagic clay and a sequence of interbedded chert/porcellanite/clay is expected to produce the shallowest subseafloor high-amplitude reflection (e.g., Wilkens et al., 1993).
Figure F2 summarizes the core-log-seismic correlation at Site 1149 based on synthetic seismograms. The observed seismic data and synthetic seismogram are both displayed using identical parameters, including true relative amplitude, in order to clearly identify the highest amplitude events. A preliminary synthetic seismogram (not shown) matched all the major features of the observed seismic profile including the high-amplitude reflections expected from the largest impedance contrasts at the seafloor, the pelagic clay/chert interface (Unit II/III boundary), and the top of the oceanic crust. However, the TWT of the major subseafloor reflections on the preliminary synthetic seismogram were slightly larger (10 ms deeper) than those observed on seismic profiles. This indicates velocities used in this synthetic were slightly too low. The simplest solution to achieve a perfect match between the onset of the highest amplitude reflections of the synthetic and observed seismograms was to raise the average velocity (1525 m/s) of the laboratory-derived measurements in the upper 180 mbsf by 69 m/s, a 4.5% increase. Is this a valid solution? It is well known that laboratory measurements of velocity are lower than in situ measurements in unlithified marine sediments because of core expansion and disturbance created from the coring process. This indicates the sign of the correction is correct (i.e., laboratory velocities must increase), but is the amount of increase reasonable? The ~20-m interval (160-180 mbsf) of in situ velocity measurements overlapping with laboratory measurements in the pelagic clay (Subunit IIB) average of ~1568 m/s, a 3.5% increase over the average laboratory velocity in this interval. This result is very close to the amount of velocity increase required to exactly match synthetic and observed seismograms; however, downhole velocity measurements in this interval are extremely noisy because of widely varying borehole diameters and may be too low (Fig. F3).
Various site specific and general velocity vs. depth functions have been developed that correct laboratory velocities to reflect in situ conditions (e.g., Urmos et al., 1993; Carlson et al., 1986). Although these corrections are not specific to the high-porosity, high-clay content of Units I and II, they can serve as a general guide to in situ velocities in this short and shallow depth interval. Application of the corrections of either Urmos et al. (1993) or Carlson et al. (1986) to our data raises the average velocity in the 0-180 mbsf interval by ~7% (~1630 m/s) and places the first high-amplitude subseafloor reflection only 6 ms shallower than on the observed seismogram (i.e., decreases the TWT). Because the character of the observed seismic (i.e., both low and high amplitudes) is so well matched by the synthetic, the use of any of these corrections to laboratory velocities results in similar conclusions; the Unit II/III boundary at 180.12 mbsf corresponds to the shallowest subseafloor high-amplitude reflection at 226 milliseconds below seafloor (msbsf), and the top of oceanic crust at 410 mbsf corresponds to the high-amplitude reflection at 438 msbsf. The synthetic seismogram utilized in this paper applies the simplest solution of raising the laboratory velocities by a constant 69 m/s, resulting in an average velocity of 1594 m/s for the upper 180 m of sediment.
These results indicate that the correlation between lithologic unit boundaries (core depth) and TWT given by the Shipboard Scientific Party (2000c), which were established without the aid of synthetic seismograms, must be adjusted to slightly greater traveltimes. The Unit II/III boundary is placed at 226 msbsf rather than 200 msbsf. The Unit III/IV boundary correlates to 336 msbsf rather than 280 msbsf, and the top of oceanic crust is at 438 msbsf instead of 420 msbsf.
Ewing et al. (1968) originally defined the acoustic stratigraphy of large portions of the western Pacific as consisting of two or more of the following four seismic units: (1) an upper transparent layer (weakly reflective), (2) an upper opaque layer (highly reflective or well stratified), (3) a lower transparent layer, and (4) acoustic basement. Acoustic basement has been referred to as "Horizon B," the "deep opaque layer," and as the "reverberant layer" where it is characterized by an interval of flat-lying, smooth, high-amplitude, closely spaced reflections (Ewing et al., 1968; Heezen, MacGregor, et al., 1973; Houtz et al., 1973; Houtz and Ludwig, 1979).
These regional seismic facies are also apparent at Site 1149. The upper transparent layer is highlighted in brown, the upper opaque and lower transparent facies are shown in light blue, and acoustic basement is displayed in gray (Figs. F2, F4). The lowermost seismic facies, acoustic basement, is characterized by a single or two to three closely spaced high-amplitude continuous reflector(s) (Figs. F2, F4). These continuous reflections range in appearance from relatively smooth and flatlying to diffractive and undulating depending on seismic source and azimuth of profile direction. In Holes 1149A and 1149B, acoustic basement begins at 8138 milliseconds two-way traveltime (mstwt), equivalent to 438 msbsf, and is created from the impedance contrast between nanno-chalk/marl (Unit IV) and fractured basalt at 410 mbsf.
The lower transparent layer is poorly represented at Site 1149, and often, it appears that the upper opaque layer directly overlies oceanic crust, depending on processing parameters such as automatic gain control. The lower transparent facies appears as relatively low-amplitude discontinuous chaotic to hummocky reflections extending down to acoustic basement and exhibits pelagic sheet drape character because it is generally of uniform thickness and concordant with the underlying basement topography. This interval correlates to the interbedded chert/chalk/marl of lithologic Unit IV. The lithologic Unit III/IV boundary (282.9 mbsf = 336 msbsf), however, is not associated with a distinct continuous reflection because there is no large and abrupt impedance contrast at this boundary. Instead, it is characterized by a reflection interference pattern with a subtle transition from the upper opaque to the poorly developed lower transparent seismic facies.
The upper opaque layer as originally defined by analog air gun records appears stratified rather than "opaque" on these SCS water gun records (Figs. F2, F4). The upper opaque layer in Holes 1149A and 1149B begins at 226 msbsf (7926 mstwt) and consists of high-amplitude continuous reflections that mimic the underlying basement relief and appear as a stratified pelagic drape deposit of generally uniform thickness. The continuous high-amplitude reflection that appears at 226 msbsf is the result of the large and abrupt impedance contrast, which is present between pelagic clay (Unit II) and the shallowest abundant chert at ~180 mbsf (Unit III) (Figs. F2, F4). The stratified appearance is an interference pattern from closely spaced impedance contrasts associated with interbedded chert, porcellanite, and clay similar to that observed by Wilkens et al. (1993).
The upper transparent layer is relatively thick and extends from the seafloor to 226 msbsf (7926 mstwt). This unit has a pelagic sheet-drape form and a relatively reflection-free seismic character in the upper portion with semicontinuous reflections of low amplitude apparent in the lower portion of this interval. The "transparent" character is indicative of a relatively homogenous interval containing no significant impedance contrasts and is correlated to the unlithified siliceous ash-bearing clay of lithologic Subunit IA. The weak semicontinuous reflections beginning at 148 msbsf correlate to lithologic Unit II (118.2-180.12 mbsf). The Unit I/II boundary is marked by abrupt but relatively small changes in porosity, grain density, and downhole resistivity, and extreme changes in borehole diameter are present within Unit II (Fig. F3) (Shipboard Scientific Party, 2000c).