The Walvis Ridge at 20°S separates the Angola and Cape Basins in the eastern South Atlantic. Plate tectonic processes dominated the early evolution of the South Atlantic, which caused pronounced differences between the two basins since the Cretaceous because of the development of the Walvis Ridge and Rio Grande Rise as barriers for open-ocean circulation in the northern basins. The Walvis Ridge affects the oceanography of the region to this day and, in conjunction with the structure of the continental margin, is an important element of the Benguela upwelling and current systems.
The fourth working area of ODP Leg 175 covers the Walvis Ridge and the Walvis Basin as part of the Cape Basin. The general structure of the Cape Basin continental margin, which developed during the Cretaceous after the breakup of Gondwanaland, is summarized in several comprehensive papers (e.g., Emery and Uchupi, 1984; Dingle and Robson, 1992). In general, sedimentation at the continental margin was characterized by intense terrigenous deposition during the early opening of the South Atlantic in the Cretaceous. This high sedimentation rate subsequently declined during the Paleogene. Aridification of the continent in the Oligocene and Miocene led to a further decrease in sedimentation rate which, in conjunction with intensified bottom-water currents, resulted in starved conditions in the deeper region of the Cape Basin.
Isopach maps (Gerrard and Smith, 1984) based on borehole data, offshore geology, and seismics related to exploration activities along the coast indicate the presence of only a few elongated Cenozoic depocenters along the margin seaward of the Mesozoic depocenters. The northernmost depocenter, the Walvis Basin, was identified as a potential area to drill expanded sequences of Neogene upwelling sediments and was surveyed during Meteor Cruise M34/1 in January 1996 (Bleil et al., 1996).
The Walvis Ridge and Walvis Basin together represent a depositional realm, which is fed mainly by a combination of diatom-dominated coastal upwelling and oceanic production of biogenic carbonate. Deposition rates of the different components are supposed to be a function of distance to the coast, upwelling intensity, and the location of the different arms of the Benguela Current systems. The site on the Walvis Ridge was located closer to the coast and toward the coastal upwelling cell on the Walvis Ridge than DSDP Site 532 (Hay, Sibuet, et al., 1984). Sites in the Walvis Basin, for which basically no seismic data were available before Meteor Cruise M34/1, were added to the original proposal to reconstruct fluctuations in upwelling activity and to monitor movements of the Benguela ocean current.
An extensive 7-day seismic survey was carried out between 19°30'S and 22°S in water depths from 200 to 2500 m (Fig. 3). The area is located off the most active coastal upwelling areas of the world, and it was assumed that it received significant sediment input during the late Neogene. Variations in thickness and regional distribution of seismic units should reflect the intensities and locations of the major sediment sources. Accordingly, a grid of seismic lines was chosen to allow future two-dimensional analyses. The seismic survey revealed a wide depositional basin characterized by significant subsidence on the upper continental slope, probably also as a result of the high sediment load. This led to a moderately inclined continental slope, which supported rapid hemipelagic deposition without the typical mass-movement events. No indication for slumping was found in Neogene sequences.
Because the seismic survey covered both the proposed sites on the Walvis Ridge and in the Walvis Basin and DSDP Sites 532 and 362, drilled in 1300 m water depth on the crest of the ridge, a seismostratigraphic framework could be developed for the uppermost seismic units based on previous drilling results. Accordingly, a few marker horizons were chosen and could be traced along the lines into the working area to estimate ages and sedimentation rates. Studies of seismic and digital echosounder data near DSDP Site 532 also revealed increasing current intensities, winnowing, and erosion.
A net of 16 seismic lines with a total length of 1550 km was recorded in the working area during Meteor Cruise M34/1 (Fig. 3). About 395 km was located on the crest and southern flank of the Walvis Ridge; 1155 km covered the Walvis Basin. The lines were oriented so that both sediment input from and variations along the coast could be identified from two-dimensional analyses. Sites in the Walvis Basin were chosen on Lines GeoB/AWI 96-015 and 96-017, for which crossing lines were shot. The Walvis Ridge site was originally proposed based on data of Sibuet et al. (1984), which were collected during the presite survey for DSDP Leg 75. High-resolution multichannel seismic data were recorded during Meteor Cruise M34/1. Lines GeoB/AWI 96-020 to 96-022 connect the Walvis Basin with DSDP Sites 532 and 362, and Line GeoB/AWI 96-024 runs across DSDP Sites 532 and 362 along the crest of the Walvis Ridge upslope. Shore-based studies will allow detailed correlations of the seismic data, specifically the reconstruction of areal depositional patterns and accumulation rates.
The area of the seismic survey generally shows similar overall acoustic characteristics. Three sections of seismic lines will be shown across Sites 1081–1083 to illustrate these similarities.
Line GeoB/AWI 96-024 (Fig. 4) reveals a large number of strong reflectors in the upper seismic Unit 1. They cover a depth range of 250 ms two-way traveltime (TWT) at DSDP Site 532, but extend down to 750 ms TWT at Site 1081. Also, the zone of higher seismic amplitudes does not cover the same lithologies but ranges stratigraphically deeper at Site 1081. This may indicate different origins for high amplitudes at both sites, and further, more detailed analyses are required to understand these lateral variations. This seismic unit was not penetrated at Site 1081; therefore, deeper units will not be discussed here (see Sibuet et al., 1984). This unit, however, thins significantly between common depth points (CDPs) 1000 and 3000, and an erosional contact can be identified at ~100 ms TWT. The erosional contact is attributed to bottom currents that are forced to cross the Walvis Ridge and are accelerated, causing winnowing and erosion. Surface sediment cores taken at shallower water depth by the University of Bremen revealed numerous hiatuses. The sedimentary unit above the erosional surface thickens again toward Site 1081 from 100 ms TWT to a thickness >300 ms TWT. There, the sequence appears to be continuous. The sedimentary packages beneath the drilled sequences are moderately affected by faulting. Also, a slump unit could be identified at ~700 ms TWT upslope of Site 1081, but no indication for slumping was found at shallower depth.
Line GeoB/AWI 96-015 (Fig. 5) also shows an upper seismic unit of high-amplitude reflectors to a sub-bottom depth of 600 ms TWT, which was the target for drilling. This unit thins downslope, and again the interval of high seismic amplitude ranges stratigraphically deeper than at greater water depth. Around CDP 5500, a major slump scarp probably of Eocene age was observed, which could also be identified on other lines. Reflectors in the upper seismic unit can be traced to DSDP Site 532 along Lines GeoB/AWI 96-020, 96-021, and 96-022. In particular, according to this preliminary correlation, the Eocene/Oligocene boundary can be placed at about 750 ms TWT at Site 1082. A zone of even higher seismic amplitudes associated with significant scatter energy at shallower depth between CDPs 7200 and 8300 was avoided for drilling.
Line GeoB/AWI 96-017 (Fig. 6) basically reveals the same seismic characteristics as the other seismic lines but also shows a seaward thinning of the upper sedimentary sequence. The general morphology is still slightly affected by the surface created during the Eocene slump event and/or by subsequent movements, but deposition of the upper unit of high seismic-reflection amplitudes is apparently unaffected. A major tectonic feature around CDP 4700, including local subsidence, was avoided for drilling.
Site 1081 is located in 794 m water depth at the northern end of the survey area (Fig. 3) on Line GeoB/AWI 96-024 (CDP 3967). Figure 7 shows a 10-km-long seismic section of Line GeoB/AWI 96-024 across Site 1081. The seismic pattern reflects hemipelagic deposition without major disturbances or faulting. Some indication for slumping was found at 700 ms TWT sub-bottom depth upslope of CDP 4020, but not within the drilling range of 450 m. Seismic amplitudes vary significantly along Line GeoB/AWI 96-024 between DSDP Site 532 and ODP Site 1081 (Fig. 4) and increase near Site 1081. Seismic characteristics remain constant within the drilled depth range and give rise to the development of a detailed seismostratigraphy.
Figure 8 shows a close-up of the seismic section, plotted against sub-bottom depth for a sound velocity of 1500 m/s, for a 1-km-long interval near the drill site. Seismic reflectors are compared with the density log (see "Downhole Logging" section, this chapter). Sharp peaks in the density log indicate dolomitic or carbonate-rich clays, sometimes lithified to different degrees (see "Lithostratigraphy" section, this chapter). Most of these peaks could also be correlated to peaks in GRAPE density or findings of lithified sediment pieces in the core or core catchers, which are not measured with the MST.
Some seismic reflectors can be directly correlated to these layers, but in other cases no seismic energy is found at these depths. A dependency on layer thickness, as derived from logging data (see Fig. 47, "Downhole Logging" section, this chapter), was not observed. Surprisingly, the band of dolomitic layers from 300 to 320 mbsf is not seen in the seismic record. Also, some major lithologic changes seem to be associated with seismic reflectors, but seismic modeling is required to provide a precise reference for high-resolution seismic stratigraphy. Traditional seismic stratigraphy, conducted by identifying major reflectors with pronounced lithologic changes and geologic events, cannot be carried out where early diagenesis has locally overprinted the physical properties, as is observed at Site 1081. A more detailed approach has to be taken and further high-resolution seismic measurements, together with the analysis of frequency-dependent seismic properties (seismic attributes), are required to distinguish between dolomitic layers and other lithologic changes.
Site 1082 is located in 1280 m water depth in the middle of the survey area (Fig. 9) on Line GeoB/AWI 96-015 (CDP 6837). Figure 9 shows a 10-km-long seismic section of Line GeoB/AWI 96-015 across Site 1082. The seismic pattern reflects hemipelagic deposition without major disturbances or faulting. Seismic amplitudes are high for the upper unit and gradually decrease beneath 600 ms TWT sub-bottom depth. Seismic characteristics remain constant within the drilled depth range and give rise to the development of a detailed seismostratigraphy.
Figure 10 shows a close-up of the seismic section, plotted against TWT, for a 1-km-long interval near the drill site. Seismic reflectors are compared with the sound velocity log (see "Downhole Logging" section, this chapter), which was also used to recalculate logging depth to TWT. This modification was necessary because for the first time during the leg, velocities were so high that scale stretching could not be used because of a nonlinear relationship. Peaks in the velocity log are associated with dolomite layers or lithified intervals, and several could be correlated to seismic reflectors. Major lithologic changes have a more pronounced effect on density, which is not shown here but may have only a minor impact on velocity. Therefore, it cannot be expected that all reflectors find expressions in the velocity log. However, the boundary between lithologic Units IB and IC from diatom-rich clay to nannofossil clay, which is associated with a sharp increase in density with depth, is clearly identified at 485 ms in the seismic record.
Site 1083 is located in 2178 m water depth northwest of Site 1082 (Fig. 3) on Line GeoB/AWI 96-017 (CDP 6642). Figure 11 shows a 10-km-long seismic section of Line GeoB/AWI 96-017 across Site 1083. The seismic pattern reflects hemipelagic deposition without major disturbances or faulting and is very similar to the pattern at Sites 1081 and 1082. An unconformity or an intercalated slumped block is observed deeper in the section at 650 ms and is more pronounced farther downslope, from CDP 6700 on. The unit of increased seismic amplitudes is thinner than at Site 1082 and can be correlated in detail between the sites along Lines GeoB/AWI 96-021 and 96-022.
Figure 12 shows a close-up of the seismic section, plotted against sub-bottom depth for a sound velocity of 1500 m/s, for a 1-km-long interval near the drill site. Seismic reflectors are compared with the GRAPE density log (see "Physical Properties" section, this chapter). Physical properties show pronounced and rapid fluctuations at rates that cannot be directly imaged because of the limited seismic frequency content of the data. Some major features of the density core log seem to be associated with reflectors, however, particularly in the interval of higher diatom abundance between 100 and 160 mbsf. No evidence for lithified intervals was found at this site, and amplitudes of reflection are generally lower. Some of the reflectors seem to be traceable to Site 1082, where they are associated with dolomitic layers.