Previous geophysical surveys in the Palmer Deep area comprise the 1992 cruise of the Polar Duke (U.S. Antarctic Program) and the 1997 survey of the OGS-Explora (Programma Nazionale di Ricerche in Antartide) (see "Seismic Stratigraphy" in the "Explanatory Notes" chapter, and "Appendix" and Fig. AF1, both in the "Leg 178 Summary" chapter). During these surveys, very high-resolution (i.e., decimeter scale) Huntec deep-tow boomer (DTB) and lower resolution, single-channel air gun seismic profiles, respectively, were collected (see "Background and Scientific Objectives"). Seismic stratigraphic interpretations of these profiles were made by Kirby (1993) and Rebesco et al. (1998). Both data sets have been used at Sites 1098 and 1099 to establish the acoustic stratigraphy described below. Physical properties of cores are used to correlate the cored section with seismic units and assign true depths to reflectors.
Bulk densities for the formation have been derived from MST (GRAPE) measurements (2-cm spatial resolution) and index property measurements (3.1-m spatial resolution, Site 1098; 3.5-m spatial resolution, Site 1099 [see "Physical Properties"]). Both sections were cored using an APC, which provides undisturbed sediment that yields a high-quality MST (GRAPE) density record. Low GRAPE density values resulting from gas expansion gaps in the core from the lower part of Site 1099 have been removed using methods described in "Seismic Stratigraphy" in the "Explanatory Notes" chapter. Spliced density and velocity data (see "Composite Depths") have been used for acoustic models at both sites (Figs. F42, F43).
Two differently derived velocities are available. The MST logger provided continuous data (4-cm spatial resolution) for all cores of Site 1098 and the upper 70 mbsf of Site 1099. Additionally, discrete PWS3 Hamilton Frame measurements with a lower spatial resolution (1.4-1.5 m; see "Physical Properties") are available. The model for Site 1098 uses MST velocities, and PWS3 velocities are used for the Site 1099 model. High MST P-wave velocities below 35 mbsf at Site 1098 are possibly an artifact (they show no correlation with the density) and should be re-evaluated (Fig. F42 [note that not all of the scatter is displayed in Fig. F26]). The lack of P-wave data below 77.5 mbsf at Site 1099, resulting from core disturbance by gas exsolution, has been compensated for by extrapolation, using a third-order polynomial (Fig. F43).
Three different source signals are used for the seismic models at Sites 1098 and 1099: the far-field and seafloor signatures of the Generator Injector (GI) gun (as described in "Seismic Stratigraphy" in the "Site 1095" chapter) and a high-resolution digital recording of a Huntec far-field signal (P. Simptkan, pers. comm., 1997). The very short Huntec signal with a length of 0.25 ms and a continuous energy spectrum up to 10,000 Hz is displayed in Figure F44. With a vertical resolution of 0.2 m (under optimal circumstances), the DTB system is capable of producing ultrahigh-resolution records of non-indurated sediments.
Depth/traveltime curves have been calculated using the available velocity data (Fig. F45). At both sites, the traveltime/depth relationship is close to linear. The steepening of the Site 1098 curve in the lower part is possibly the result of incautious use of MST P-wave data (velocities increase in the lower part [see PWS3 data, in "Physical Properties"]).
The spliced and corrected velocity and density data were resampled at 0.2-m (Site 1098) and 0.15-m (Site 1099) resolution. The reflectivity series of Site 1098 was convolved only with the Huntec signal because no profile taken over Basin I with any other acoustic source (air gun and GI gun) showed comparable detail within the shallow basin fill. The velocity/density data, the impedance curve, reflectivity coefficients, and an unfiltered synthetic trace are displayed in Figure F42.
For Site 1099, the acoustic impedance model was convolved with the GI gun far-field and seafloor, and with Huntec signals (Fig. F43). A depth axis is displayed next to the time scale to allow convenient traveltime/depth conversion for both sites. Extreme frequency and resolution differences are obvious within the synthetic traces derived from the three different source signals at Site 1099. The time delay between an impedance contrast (e.g., at 0.043 s) and the corresponding reflection within the synthetic trace increases dramatically with the wavelength of the different signals. The response of the Huntec trace is "instant." The GI far-field response occurs at 0.048 s, and the GI seafloor trace of 40 ms shows a response at 0.055. To eliminate these effects, reflector depths must be measured with respect to the onset of the seafloor reflection.
The Huntec traces for Sites 1098 and 1099 were resampled at 20,000 Hz and subsequently filtered using a zero-phase equi-ripple band-pass filter (pass band = 800-4500 Hz; attenuation = 40 dB; and filter order = 40 [Fig. F44]) to match their frequency content to a commonly used frequency range for the DTB-record display (Fig. F46). Most of the prominent reflectors are represented within the synthetic traces for Site 1098 (see "Seismic Units"). At Site 1099, the upper 40 ms reveal all the important reflectors of the DTB profile (Fig. F47). For a detailed discussion and description of the seismic section, see "Seismic Units".
Additionally, the GI gun seafloor synthetic trace (based on the same impedance model) has been filtered using a zero-phase equi-ripple band-pass filter (pass band = 30-110 Hz; attenuation = 35 dB; and filter order = 70) and processed to meet the processing specifications of seismic line I97H-219G crossing Basin III (zero-phase Butterworth band-pass filter [pass band = 30-110 Hz], automatic-gain recovery window of 100 ms, static shift and cutting, and interpolation to 0.5-ms resolution). The only feature that can be identified with some confidence is a broad reflector at 50 ms TWT (Fig. F48).
At Site 1098, we sampled sediments within a narrow basin, Basin I of Kirby (1993) and Rebesco et al. (1998) in Palmer Deep, to a depth of 46.7 mbsf, reaching acoustic basement. In the single-channel air gun seismic profile I97H-218G, we identified one seismic unit above acoustic basement (Fig. F49).
Seismic Unit I (0-46.7 mbsf) is acoustically semitransparent with some low-amplitude reflectors toward the base of the unit that onlap the irregularities of acoustic basement (Fig. F49). High-resolution DTB seismic profiles penetrate to ~43 mbsf and allow division of seismic Unit I into three subunits (Fig. F46): (1) Subunit Ia (0-7 mbsf) is characterized by stratified continuous reflectors with a high-amplitude reflector at its base, (2) Subunit Ib (7-23 mbsf) consists of parallel reflectors at the top but becomes more transparent toward the bottom, and (3) Subunit Ic (23-43 mbsf) is mainly acoustically transparent but with high-amplitude reflectors at 30, 33, and 38 mbsf (T1, T2, and T3 in Fig. F46). The synthetic seismogram clearly reveals the high-amplitude reflections that occur at the base of Subunit Ia and at the T1, T2, and T3 reflectors in Subunit Ic. Higher and lower amplitude reflection patterns in the synthetic traces can be correlated with acoustically stratified and semitransparent acoustic facies, respectively, in the DTB seismic profile (Fig. F46). Seismic Unit I has a drape geometry, indicated by uniform thickness of the unit in the central part of the basin and a bottom morphology that follows the irregularities of acoustic basement (Fig. F46).
Parallel reflectors and sheet drape geometries in both air gun and DTB seismic profiles are compatible with sedimentation dominated by hemipelagic/pelagic settling and low-density gravity flows. High-resolution DTB profiles allow direct correlation between acoustic character, synthetic traces, and lithology (Fig. F46). The reflector marking the base of seismic Subunit Ia corresponds roughly to the base of a massive, bioturbated muddy diatom ooze (see "Lithostratigraphy"). This reflector is located ~2.5 m below the base of Subunit Ia of Rebesco et al. (1998). Semitransparent and stratified reflectors of seismic Subunit Ib correspond to a 15-m interval where laminated sediments predominate. The base of Subunit Ib coincides with the base of Unit Ib of Rebesco et al. (1998) (Fig. F46) (see "Lithostratigraphy"). High-amplitude reflections in the synthetic traces of seismic Subunit Ic correspond to an acoustically semitransparent subunit in the DTB profiles. This suggests that the acoustic character of this unit is the result of limitations in the penetration of the DTB system and is not a real consequence of the stratigraphic and lithologic character of Subunit Ic. Three turbidite layers, T1 through T3, are correlated with high-amplitude reflectors at 30, 33, and 38 mbsf (see "Lithostratigraphy").
Site 1099 was drilled to a depth of 107.5 mbsf in Palmer Deep Basin III (Kirby, 1993; Rebesco et al., 1998). Two main seismic units can be differentiated in the air gun profile (Fig. F50).
Seismic Unit I (0-76 mbsf) is acoustically semitransparent (Fig. F50). A high-amplitude reflector, previously referred to as the MBR (Kirby, 1993; Rebesco et al., 1998), divides Unit I in two. Thickness of Unit I is uniform in the center of the basin and reduces toward the edges of the basin. High-resolution DTB profiles penetrate the upper 34 m of seismic Unit I to the MBR (Fig. F47). Acoustic character above the MBR in this high-resolution profile allows us to differentiate seven acoustic packages characterized by low-amplitude reflectors at the top and parallel higher amplitude reflectors at the bottom. In Figure F47, we assign letters from a through g to the base of acoustic packages at 4.5, 8, 10, 12, 14, 19, and 21 mbsf, respectively. The synthetic seismogram at this site shows the correlation between high-amplitude reflections and the base of the acoustic packages (Fig. F47). The deepest strong reflector in the DTB profiles and a high-amplitude reflection in the synthetic seismogram are found at 34 mbsf and correspond to the MBR (also named h in Fig. F47). The MBR in the air gun seismic profile is a horizontal reflector that can be traced across the basin. In the middle of the basin, the MBR has lower amplitude and an irregular surface. Toward the sides of the basin, it grades to a higher amplitude reflector that onlaps a more acoustically transparent and chaotic package of reflectors. The MBR in the DTB profiles is a reflector that parallels the irregularities of the sea bottom.
Seismic Unit II (76-107.5 mbsf) in the air gun profiles is characterized by horizontally stratified high-amplitude and continuous reflectors (Fig. F50). Reflectors in the upper part of this unit are flat-lying in the center of the basin, "climb" up at its eastern edge, and terminate abruptly in the western part of the basin. Below, reflectors exhibit typical onlap fill geometry either against the acoustic basement or against a package of irregular and chaotic reflectors. Thickness of seismic Unit II varies across the basin because of irregularities in the acoustic basement.
Seismic Unit I above the MBR (the top 34 mbsf) is dominantly massive and laminated muddy diatom ooze. The higher amplitude reflectors (i.e., Reflectors a through g) at the base of the seven acoustic packages identified in the DTB seismic profiles correspond to seven thin interbedded turbidites in the upper 23 m of the core. Reflectors a, e, and f coincide with the base of Units Ia, Ib, and Ic, respectively, of Rebesco et al. (1998). The MBR, apparent in both air gun and DTB profiles at 34 mbsf, corresponds to a layer of coarse sand and pebbles interpreted as a clast-poor diamict layer with a diatom clayey silt matrix (see "Lithostratigraphy"). The acoustically semitransparent unit below the MBR corresponds roughly to a thick interval of massive, diatom clayey silt and muddy diatom ooze (Fig. F11). High-amplitude stratified reflectors of seismic Unit II correspond to alternations of massive, bioturbated, muddy diatom ooze, laminated mud-bearing diatom ooze, and very fine-grained graded turbidites (see "Lithostratigraphy").
The overall onlapping fill character of the seismic units drilled in the Palmer Deep basins suggests that the style of sedimentation was dominated by hemipelagic drape and by infilling of the basins with sediment gravity flows, probably from a local source. An intrabasinal source for sediments is suggested by the irregular reflector packages on the sides of the basin (i.e., the western side in Fig. F50). The "climbing" reflectors of Unit I and the uppermost reflectors of Unit II at the edge of the basin may be interpreted as suggesting neotectonic activity in the basin.