The drill sites in the Lower Congo Basin at ~5°-6°S latitude were the northernmost sites of Ocean Drilling Program (ODP) Leg 175, during which six working areas along the southwest African continental margin were drilled to study the evolution of the Benguela Current system in the late Neogene. To explain the strategy of the site survey (Bleil et al., 1995) a short summary of the evidence existing to date is given regarding the expected sedimentary structures at this part of the continental margin.
The survey area (Fig. 4) is located off the mouth of the Congo River, which is the second largest river of the world with respect to water discharge (Peters, 1978; Eisma and van Bennekom, 1978). Deposition in the Congo Fan area is characterized by significant sediment influx from the continent. A major proportion of the sediment is transported through the Congo Canyon directly into the deep sea (Heezen et al., 1964; Shepard and Emery, 1973; van Weering and van Iperen, 1984; Droz et al., 1996). The canyon extends far into the lower stretches of the river, with water depths of several hundred meters already inshore (Peters, 1978; Eisma and Kalf, 1984). The sediment is guided through the shelf and accumulates in small basins. From time to time, these materials are released and move down the canyon. The narrow, and in some sections more than 1000 m deep, channel opens into the Congo Cone at a water depth below 3000 m, where typical fan deposits and channel/levee systems are commonly associated with chaotic sedimentary structures.
The selection of drill sites in the original proposal was oriented along the seismic Line 62 of Emery et al. (1975), but the data quality was not sufficient to identify fine-scale sedimentary structures and establish the absence of disturbances. Also, we wished to survey the area in greater detail and with higher resolution using digital multichannel seismic lines. Six crossings were derived from analog on-line seismic records of Lines GeoB (Geosciences Bremen)/AWI 93-001 and 93-002 and were later all proposed as potential drill sites (Fig. 4).
Bathymetric survey data show that the seafloor is rather smooth in the working area. The lower Congo Cone is disrupted, however, by numerous small and a few larger distributary channels. Upslope, the general character of the seafloor changes between 3500 and 2500 m water depth from a rough, diffracting surface to a continuous layering with numerous parallel internal reflectors. No indications of channels, downslope transport, or slumping were found in the surface sediments shallower than 2500 m water depth.
Altogether, nine seismic lines were recorded in the northern Congo Fan area, with six crossings at potential drill sites in water depths between 1400 and 3000 m (Fig. 4). Connecting profiles will later allow a regional seismic correlation. The first line (GeoB/AWI 93-001; Fig. 5) started at 8°30'E/5°30'S to cover the depositional environment of the fan, particularly its change toward a predominantly (hemi-) pelagic sedimentation at the continental slope.
The area of the seismic survey generally shows similar overall acoustic characteristics. The southern Line GeoB/AWI 93-001 has basically recorded the same features as Line 62 of Emery et al. (1975). The basement near the continent is deeply buried, and no salt diapirs are observed close to the surface. Deep reflectors are undulating, however, indicating minor deformation of deeper layers by salt movement or an early tectonism. These undulations are smoothed out toward the surface and do not control the ocean-floor morphology. On Line GeoB/AWI 93-002 (Fig. 6), which is about 25 nmi farther to the north, intense salt diapirism is apparent, and numerous disturbed sequences are observed near the surface. For this reason, the crossing lines were concentrated on those sections of Lines GeoB/AWI 93-001 and 93-002 where penetration was high and a distinct layering was observed. The general seismostratigraphic characteristics are given in Uenzelmann-Neben et al. (1997) and will be summarized here for the part of the sedimentary column that was cored at Sites 1075 to 1077.
In the survey area, the characteristic pattern of fan deposits has completely disappeared in the upper sedimentary column. Two seismostratigraphic units can be identified in the range of the drill holes. They are distinguished mainly by their reflection amplitudes and reflector coherence, but lack distinct reflectors. These seismostratigraphic units are described below.
Seismostratigraphic Unit 1 is characterized by low reflection amplitudes and a few, partially coherent, weak reflectors. The thickness varies between 100 and 160 ms two-way traveltime (TWT; ~80 to ~120 m). The base of the unit reveals a transition from a completely transparent interval to higher reflection amplitudes. The sedimentary column appears to be heavily faulted. Digital echosounder data, however, indicate only minor vertical offsets of a few meters. The seismic unit commonly cuts across reflectors; the unit represents a zone of modified reflector amplitudes.
Seismostratigraphic Unit 2 shows higher reflection amplitudes, which are still weaker than the surface reflector. The base of the unit is more affected by the undulating deeper structures than its top, which may be caused by differential subsidence and higher sediment accumulation in the "topographic" lows. Accordingly, the thickness varies significantly from 220 to 350 ms TWT (~160 to ~260 m). Because of the pronounced changes in reflectivity, it cannot be excluded that either the top or the base of Unit 2 is an unconformity. Seismic Unit 2 is split, being separated by one or two thin transparent bands (one on Line GeoB/AWI 93-001 and two on Line GeoB/AWI 93-002).
After the safety panel review, only 200 m of drilling was allowed for each of the proposed sites in the Lower Congo Basin area. Therefore, deeper seismostratigraphic units could not be reached by drilling during Leg 175. At the greater depths, evidence was found both in multichannel seismic and digital echosounder data for the existence of gas, gas hydrate, fluid migration, and microfaulting through the entire area. The identification of seismic units based on average reflection amplitudes can be affected by these components. Amplitudes show a strong lateral variation and in many cases do not reflect lithologic boundaries. Commonly, nearly vertical transparent zones are seen in the vicinity of small faults and surface depressions, which in some cases appear as pockmarks.
The basic assumption that the margin is not receiving sediment via distributary channels of the Congo Fan could be confirmed by the seismic survey. Only hemipelagic deposits were found, and slumps or debris flows seem to be absent in the upper sedimentary column. The sedimentation rate appears to be nearly constant within the area and is more influenced by local effects such as subsidence and diapirism.
Seven potential drill sites had been identified in the Upper Congo Fan area north of the Congo Canyon in water depths between 1400 and 3000 m. Six were finally proposed and were approved for drilling down to 200 mbsf. They all represent a similar depositional environment at varying distances to the shelf break and at different water depths and positions with respect to the Congo River plume. Cross profiles were shot for five sites.
Site 1075 is located at the northwestern corner of the survey area, farthest from the river mouth (common depth point [CDP] 9750 of Line GeoB/AWI 93-002) in 2995 m water depth. A crossing single-channel seismic line was shot during the approach to the site with the onboard equipment of the JOIDES Resolution to assure the absence of high-amplitude layers indicative of potential gas accumulations. No evidence was found for gas accumulation. Figure 7 shows a 10-km-long seismic section of Line GeoB/AWI 93-002 in the vicinity of Site 1075. Weak reflectors were enhanced by normalization of amplitudes to a constant value. Therefore, the transition between the low reflective seismic Unit 1 and the higher reflective Unit 2 at 150 ms TWT does not appear as pronounced. The figure clearly shows that numerous (weak) reflectors can be found in the upper, mostly transparent unit (Fig. 6). Most reflectors are coherent in the section, but have interruptions that are attributed to small faults. Site 1075 was located between two of those faults to avoid a discontinuous section. The transition to seismic Unit 2, however, is marked by a reflector of higher amplitude. Most faults can be traced down to the base of Unit 2.
Figure 8 shows a close-up of about 1-km length from the vicinity of the drill site in comparison with the wet bulk density profile for Hole 1075A, plotted against sub-bottom depth for a sound velocity of 1500 m/s. Seismograms are plotted as wiggle traces with gray-scaled amplitudes as background. The lateral amplitude changes can be clearly identified, as well as the transition between seismic Units 1 and 2 at ~85 mbsf. Numerous reflectors can be correlated with changes in wet bulk density. Velocity variation is expected to be <40 m/s (<3%) and therefore can be disregarded as the main origin of acoustic impedance changes, compared with density variations of as much as 20%. An explanation for the mostly transparent upper seismic unit from the density log is not evident, although changes are generally small in the upper unit and increase with depth.
It is clear from the comparison that the main frequency of the seismic signal is not sufficiently high to resolve individual changes in the downhole physical properties and, therefore, changes in interference from consecutive layers occur. Calculation of synthetic seismograms from calibrated and spliced GRAPE data sets is required for more detailed analyses. The required thorough editing and quality control can only be carried out on shore.
Site 1076 is the shallowest site in the transect, with a water depth of 1403 m at the southeastern edge of the survey area. Figure 9 shows a 10-km-long section of seismic Line GeoB/AWI 93-001 close to the drill site at CDP 22,990. From this site, the hemipelagic sequence can be correlated to the other sites. At greater depth, disturbed intervals, hyperbolic echoes, and vertical amplitude anomalies can be identified in the unmigrated seismic section at a distance of several kilometers. The reflectors appear coherent, however, within 1 km around the drill site on the scale of this seismic image.
At the drill site, the two seismic units within a mostly undisturbed sequence can be clearly distinguished with a transition at 160 ms TWT. Reflectors of seismic Unit 2 show a toplap with Unit 1, which might be caused either by an erosional contact or by an overprint of reflection amplitudes in Unit 1.
The close-up section of 1-km length in the vicinity of the drill site is shown in Figure 10. A transition between the mostly transparent seismic Units 1 and 2 occurs between 110 and 140 mbsf, with lateral variation of reflection amplitudes. A comparison with the wet bulk density profile derived from index properties measurements (see "Physical Properties" section, "Site 1076" chapter, this volume) allows the assignment of sharp density changes to several reflectors, although a unique solution cannot be provided at this time because of the limited resolution of the data set and the limited quality of the available core log data. Further refinement will be achieved with the calculation of synthetic seismograms in shore-based studies.
Biostratigraphic data (see "Biostratigraphy and Sedimentation Rates" section, "Site 1076" chapter, this volume) revealed a break in continuous deposition at about 120 mbsf, which could be associated with the observed toplap interval. Alternatively, a nearby fault zone with a 10-m vertical offset could explain a missing interval at Site 1076. Because of the low reflection amplitudes in the upper seismic unit, further processing is required to elucidate the seismic features at the depth of the identified hiatus.
Site 1077 lies between Sites 1075 and 1076 on seismic Line GeoB/AWI 93-001 at CDP 17,000 in 2381 m water depth. This site was used specifically to investigate the origin of the acoustic transparent interval (perhaps indicating the presence of gas hydrate) with a limited suite of downhole logging tools (see "Downhole Logging" section, "Site 1077" chapter, this volume) and a detailed high-resolution sampling for interstitial water analyses across the crucial interval.
Figure 11 shows a 10-km-long section of seismic Line GeoB/AWI 93-001 around the drill site. Seismic Units 1 and 2 are clearly identified with a transition at 120 ms TWT. Besides several coherent reflectors of several kilometers extent within Unit 1, a strong reflector that was not found at the other two sites is present at 40 ms TWT. The transitional zone between both seismic units is characterized by a completely transparent band of ~10-ms TWT thickness.
Figure 12 shows a close-up of 1-km length in the vicinity of the drill site compared with a density profile derived from downhole logging, which is available beneath 76 mbsf. This data set is in very good agreement with GRAPE density logs measured on whole cores, although GRAPE density reveals pronounced minima at core breaks beneath 50 mbsf because of gas expansion. The wet bulk density profile derived from index properties measurements (see "Physical Properties" section, "Site 1077" chapter, this volume) shows a gradual increase down to 100 mbsf, with less scatter than beneath this depth. However, these density changes alone probably cannot explain the distinct difference in reflection amplitudes. The existence of gas hydrate at layers within the section was tested as a possible explanation for higher reflection amplitudes by detailed high-resolution pore-water analyses (see "Inorganic Geochemistry" section, "Site 1077" chapter, this volume). No evidence for gas hydrate was found. The gas content in the sediment, which was first observed in high concentration at ~50 mbsf, will be further investigated as a potential cause for a decrease in amplitudes.
Downhole logging information from density, velocity, and temperature also confirmed the absence of massive clathrate layers. Although the instruments were operating at their limits of resolution in these sediments with high porosity, the derived density and velocity profiles compare well with core and discrete measurements and can be further used for precise calculation of synthetic seismograms.