DISCUSSION

The first-order seismic correlations between drilled Sites 1118, 1109, and 1115 can be seen on Figure F2. Dolerite basement was intersected at Sites 1109 and 1118 (Red unit), whereas samples recovered from below ~2200 ms at Site 1115 record a history of forearc basin sedimentation (Tan unit). An unconformity above these units is onlapped successively by the Orange, Yellow, and Blue units, intersected at Sites 1109 and 1115. Increasing subsidence tilted these units to the south, producing a margin slope that received hemipelagic sediments at the same time as turbidites lapped northward and resulted in the differential sedimentation recorded by the Gold through Teal units. Submarine channels (Fig. F1) formed during this time and selectively eroded parts of these units, as seen in Figure F2 between Sites 1109 and 1115.

No samples were recovered at Site 1118 between 0 and 205.0 mbsf, as this section was drilled without coring. Age estimates can be extrapolated into this interval by correlation with Site 1109. A pair of reflectors (Teal 1 and Teal 2) in the middle of the Teal unit can be directly correlated (Fig. F2) between Site 1118 (Fig. F3) and Site 1109 (Fig. F4). At Site 1109, the reflectors are at 3086 and 3114 ms; their causative events are 22 ms higher. The time-to-depth function derived for this site indicates that the corresponding depths are 79.63 and 102.08 mbsf. Interpolation of the results of Takahashi et al. (this volume) assigns ages of 1.23 and 1.32 Ma. At Site 1118, these reflectors are at 3180 and 3214 ms, corresponding to an event at 3162 and 3196 ms. The depth-to-time function for this site assigns depths of 60.83 and 87.53 mbsf, which, assuming that the reflectors are time correlative as elsewhere, can be assigned the ages of 1.23 and 1.32 Ma (Tables T3, T4).

At Site 1118, the first prominent reflector in the Light Green (LG) unit (Fig. F3; LG 1) is directly correlatable to Site 1109 (Fig. F4; LG 1). This reflector is well reproduced in the synthetic data for each site. The synthetic seismogram for Site 1118 shows that this reflector is the result of the top of a prominent horizon at 204 mbsf distinguished by an abrupt increase in velocity and density. Logging and sparse samples define this interval as continuing to 255.5 mbsf and having a lithology that is predominantly silty at the bottom, grading upward into a formation of primarily clay (Shipboard Scientific Party, 1999d). The MCS data over this depth interval show that the bottom of this unit, as well as some internal structure, may also be resolved. The vertical seismic resolution (/4) indicated by the source wavelet is ~19 m, indicating that the entire unit should be resolvable. At Site 1109, the LG-1 reflection results from the top of a unit at 219 mbsf with unusually high velocity and density. A contribution to the reflector may also arise from a velocity and density peak at 210 mbsf. From logging data, this unit is interpreted to comprise a thin carbonate layer grading upward into thick sand characterized by high gamma ray counts. Samples of unusually dark sand with some fragments of wood were recovered from this interval, with the presence of volcanic material implied from a very high thorium/uranium ratio (Shipboard Scientific Party, 1999b). The bottom of the unit is at 234 mbsf. The MCS and synthetic data only detect the top of this unit, with no internal structure recorded. Simple interpolation of the results of Takahashi et al. (this volume) date this reflector as 1.74 Ma at Site 1118 and 1.76 Ma at Site 1109 (Tables T3, T4). This horizon is closely time correlative and formed in a middle bathyal setting at both sites. The associated MCS reflector is continuous between the two sites, despite the varying lithology.

Reflectors above the seismic Blue unit cannot be correlated directly between Sites 1109 and 1115 (Fig. F2). Some, such as LG 1, onlap the slope to the north. Others are partially or wholly removed by channel erosion. For these units, therefore, we use the detailed chronostratigraphy to identify time-correlative reflectors at Site 1115 (Fig. F5). For example, the 1.76-Ma reflector at Site 1109 (Fig. F4) can be correlated with the top of the Olduvai Chron (1.77 Ma), seen at 90.5 mbsf at Site 1115 (Fig. F5) (Shipboard Scientific Party, 1999c).

Reflectors above the Blue unit record sedimentation during subsidence of the basin from upper to middle bathyal depths. The Gold, Dark Green (DG), Light Green, and Teal units successively onlap to the north in a shingled pattern (Figs. F2, F3, F4). The DG and LG units are thicker by a factor of three at Site 1118 relative to Site 1109. The onlap does not represent an unconformity in time but rather a lateral and vertical change in depositional style from basin turbidites to hemipelagic drape (Shipboard Scientific Party, 1999a).

The top of the Gold unit is a continuous reflector that can be traced between Sites 1118 and 1109. At Site 1118, the reflector occurs at 700 mbsf, below an increase in density and velocity values from 680 mbsf. At Site 1109, the synthetic seismogram shows that the reflector originates from a segment of the reflection coefficient preceding a similar density and velocity increase and is probably caused by a local velocity low at ~350 mbsf. Interpolation of the results of Takahashi et al. (this volume) date the top of the Gold unit as 3.28 and 3.22 Ma at Sites 1118 and 1109, respectively (Tables T3, T4). This reflector can be correlated to Site 1115 using a sample containing a ⁴⁰Ar/³⁹Ar age of 3.23 ± 0.08 Ma that was found at 243.43 mbsf (Lackschewitz et al., in press) just below a reflector at 1850 ms at Site 1115 (Fig. F5; Top Gold). Interpolation of sedimentation rates independent of this age (Takahashi et al., this volume) dates this reflector as 3.23 Ma. It is associated with a downhole variation of velocities and densities very similar to that observed at Site 1118.

The top of the Blue unit (Top Blue; Figs. F4, F5) is dated by foraminifers as 3.5 Ma at Site 1109 and by interpolation between ⁴⁰Ar/³⁹Ar measurements as 3.41 Ma at Site 1115 (Takahashi et al., this volume). At Site 1109, the Blue unit is equivalent to the lower part of lithologic Unit VI. The photoelectric log data for the lower part of this unit (Shipboard Scientific Party, 1999b) show that it can be distinguished on the basis of increasing carbonate content, resulting in a gradual increase in velocity. At Site 1115, the bottom of the Blue unit coincides with lithologic Unit IV, a calcareous sandy silty claystone. A rapid increase in carbonate content at the bottom of Unit IV, as shown by the photoelectric effect log (Shipboard Scientific Party, 1999c), coincides with an increase in density and velocity.

The top of the Yellow unit at Site 1115 coincides with an increase in density and velocity at the base of lithologic Unit IV. Peaks in velocity and density within the Yellow unit are shown by the photoelectric effect log to be caused by local increases in the carbonate content (Shipboard Scientific Party, 1999c). At Site 1109, a gradual increase in density and velocity, with a localized density low at ~540 mbsf, marks the top of the Yellow unit. At both sites, the top of the Yellow unit correlates with a change from upper-bathyal to neritic environments. At Site 1115, the base of the unit may shoal to a marine lagoonal or nearshore wave-worked environment (Shipboard Scientific Party, 1999c). The age of the top of the unit (Figs. F4, F5; Top Yellow) is estimated at 3.8 Ma (3.74 Ma at Site 1109 and 3.87 Ma at Site 1115) (Takahashi et al., this volume).

The top of the Orange unit at Site 1109 is coincident with a sharp decrease in velocity and density, largely due to a decrease in carbonate content associated with brackish and freshwater deposits. Local spikes in velocity may be due to goethite and carbonate concretions (Shipboard Scientific Party, 1999b). At Site 1115, the top of the Orange unit is poorly defined in the MCS data. A velocity drop similar to that at Site 1109 is present, although less sharply developed. At Site 1109, this unit represents a downward transition from a brackish water lagoonal to a subareal environment, with an intervening period of freshwater deposition. At Site 1115, a marine lagoonal environment grades downward into a possible fluvial or beach environment (Shipboard Scientific Party, 1999b, 1999c). At Site 1109, no dates were obtained within the Orange unit or below. The top of the Orange unit is dateable only as older than 5.23 Ma. The results of Takahashi et al. (this volume) imply an age of 5.39 Ma for the top of the Orange unit at Site 1115 (Table T5; Top Orange).

At Site 1115, a conglomerate with sandstone and siltstone is the cause of a large peak in density and velocity between 565.7 and 571.9 mbsf. The associated spike in the reflection coefficient creates a reflection in the synthetic seismogram at ~2220 ms that is coincident with the top of the Tan unit. The thin unit causing this reflection probably marks local sedimentation prior to the onset of rifting in this region (Taylor, Huchon, Klaus, et al., 1999). A nannofossil identified from sediments just above this unit was dated as 8.6 Ma (Shipboard Scientific Party, 1999c). The Tan unit records the last stages of the depositional history of a forearc basin (Taylor, Huchon, Klaus, et al., 1999) that does not have an analog at Sites 1118 and 1109.

At Sites 1109 and 1118, the sharp peaks in reflection coefficient starting at ~740 and 860 mbsf, respectively, derive from the igneous conglomerates and the underlying dolerite basement. The resultant reflectors in the synthetic seismograms and MCS data mark the top of the Red unit. This unit represents the Paleocene ophiolitic basement that has analogs on the Papuan mainland (Baldwin et al., 2000) and in commercial wells to the west (Tjhin, 1976; Francis et al., 1987) and is implied to underlay the forearc basin sediments encountered at the bottom of Site 1115.