Site 999 is located ~800 m east of Shotpoint 4780 of IG2901 MCS Line CT1-12a, which was obtained on 28 and 29 May 1978 aboard the Ida Green (Bowland and Rosencrantz, 1988; Bowland, 1993). Line CT1-12 (of which CT1-12a is a short segment) continues for hundreds of kilometers to the southwest of Site 999 crossing the Mono Rise and the Costa Rican Fan and continues to the northeast of Site 999 into the eastern Colombian Basin (Fig. 1). The seismic source consisted of two 1500-in3 Bolt Inc. air guns fired at ~116-m intervals. Seismic data were received using an array of 24 hydrophone groups, each 100 m in length. Maximum hydrophone group offset ranged from 2405 to 2638 m; thus the length of the receiving array is comparable to the water depth of the Kogi Rise (~2800 m).
IG2901 12-fold stack data were provided by the University of Texas Institute for Geophysics. Processing included demultiplexing, trace editing, common depth point sorting, semblance velocity analysis spaced at 5-km intervals, and 12-fold stacking. Seismic data in this paper are displayed using band-pass filtering (5-50 Hz) and automatic gain control (AGC) with a 500-ms window.
The seismic stratigraphy of portions of the Colombian Basin, including the area of Site 999 on the Kogi Rise, has been discussed by Bowland and Rosencrantz (1988), Bowland (1993), and Sigurdsson, Leckie, Acton, et al., (1997). The seismic stratigraphy of the Site 999 area in particular has been examined by Bowland (1993) and Sigurdsson, Leckie, Acton, et al. (1997) from the analysis of MCS reflection Profile IG2901 CT1-12a (Fig. 2).
The elevated basement of the Kogi Rise is interpreted as the top of thick oceanic plateau crust, which may have experienced subsequent tectonic uplift prior to significant sediment accumulation (Bowland and Rosencrantz, 1988; Sigurdsson, Leckie, Acton, et al., 1997). Prominent, semicontinuous, dipping basement reflections are interpreted as multiple flow units and interbedded sediment representing the later stages of plateau formation and may be comparable in origin and age to the multiple basalt flow units of Campanian age (~77 Ma) recovered at Site 1001 (Sigurdsson, Leckie, Acton, et al., 1997). The maximum age for basement is unlikely to exceed the age of the widespread and contemporaneous igneous activity determined from radiometric dating (88-90 Ma) of on-land, circum-Caribbean igneous sections (Duncan et al., 1994; Sinton et al., 1997) and biostratigraphic ages (Turonian-Coniacian) of Deep Sea Drilling Project (DSDP) Leg 15 basement cores in the Venezuelan Basin and Beata Ridge, Caribbean Sea (Donnelly et al., 1973).
The bathymetry of volcanic basement strongly controls depositional character in the Colombian Basin, where basement highs such as the Mono Rise and Kogi Rise remained isolated from the large influx of terrestrial turbidites that filled in basement lows with up to 6 km of sediment from Eocene through Miocene time. A complete, almost 1067-m-thick record of largely pelagic/hemipelagic sediments from the Maastrichtian to the Pleistocene was recovered at Site 999 (Sigurdsson, Leckie, Acton, et al., 1997).
Bowland (1993) identified three principal seismic units in the sedimentary section, all of which drape over the presumed oceanic plateau basement of the Kogi Rise. Seismic Unit CB5 overlying volcanic basement extends from 4.556 s TWT to 4.936 s TWT at Site 999 and consists of continuous, high-amplitude reflections concordant with basement.
Seismic Unit CB4 extends from 4.066 s TWT to 4.556 s TWT and consists of two subunits separated at 4.4 s TWT, both of which maintain a uniform thickness in the vicinity of Site 999. Both subunits (CB4A and CB4B) have a hummocky mounded to chaotic and disrupted seismic facies. The upper subunit, CB4A, displays a high-amplitude upper boundary followed by a relatively transparent interval. The thin lower subunit, CB4B, is characterized by chaotic, discontinuous, and relatively high amplitude reflections throughout the interval.
Seismic Units CB3 and CB2 are not present at Site 999 and are restricted to portions of the Colombian Basin west of Mono Rise. These seismic units are interpreted as turbidite deposits that thin by onlap onto Unit CB5 on the western flank of Mono Rise (Bowland, 1993).
Seismic Unit CB1 extends from seafloor (3.776 s TWT) to 4.066 s TWT. This unit has a sheet-drape form, and a relatively reflection free seismic character.
Bowland (1993) reported interval velocities for these seismic units based on the average of 48-50 semblance-type velocity analyses at 5-km intervals along MCS Line CT1-12. We use the reported range of interval velocities to estimate a range of sedimentary thicknesses overlying volcanic basement at Site 999 (Fig. 2). Seismic Unit CB1 with interval velocities of 1.7 ± 0.1 km/s results in an estimated 240-270 m unit thickness. Applying the range of interval velocities of 1.8 ± 0.1 km/s for CB4A results in an estimated 284-317 m unit thickness (524-587 mbsf total). Interval velocities at CB4B are 2.6 ± 0.4 km/s, resulting in a unit thickness of 172-234 m (695-821 mbsf total). Finally, applying the range of interval velocities of 3.0 ± 0.4 km/s results in an estimated 494- to 646-m-thick CB5 section and a depth to volcanic basement ranging from 1190 to 1467 mbsf (Fig. 2).
Compressional velocity data measured on split cores by the Digital Sonic Velocimeter from 0 to 105.93 mbsf and with a Hamilton Frame apparatus from 107.43 to 1066.4 mbsf are available over the entire interval cored at Site 999 (table 18, on CD-ROM, Shipboard Scientific Party, 1997). Corrections to this data for in situ stress by empirical relations described by Urmos et al. (1993) were applied and are available from 0 to 796 mbsf (table 18, on CD-ROM, Shipboard Scientific Party, 1997). Uncorrected laboratory velocity data are used to calculate traveltimes over each depth increment to produce a plot of TWT vs. depth (Fig. 3). The resultant depth vs. TWT data are tied to TWTs picked from MCS data and average velocities are determined (Fig. 2).
Downhole measurements of velocity and density data used in this report were acquired during a single run of the tool string known as the "Quad Combo," and are available (on the Log and Core Data CD-ROM in Shipboard Scientific Party, 1997). The Quad Combo tool string included, from top to bottom, the telemetry cartridge, natural gamma spectroscopy, long-spaced sonic (LSS), compensated neutron, lithodensity, and dual induction resistivity, creating a tool string 33 m long. The sonic tool, in the middle of the tool string, sampled from 525.8 mbsf to as deep as 1041.65 mbsf (total cored depth = 1066.4 mbsf). The LSS consists of two transmitters spaced 2 ft apart, located 8 ft below two receivers, also spaced 2 ft apart. Velocity is calculated from measured traveltime between transmitter/receiver pairs over a known distance (i.e., 8, 10, 10, and 12 ft). During the downward trip into Hole 999B, the far-spaced transmitter (12-ft spacing between transmitter and receiver) in the LSS tool was determined to be malfunctioning. Postcruise examination of velocity data indicated that only velocities from the 8- and 10-ft spacing traveltime data were valid (see Log and Core Data CD-ROM in Shipboard Scientific Party, 1997). The difference in traveltimes between the 10- and 8-ft receiver/transmitter spacing (known as DTLN) were used to calculate the velocities. The major pitfall of using only some of the transit times to compute velocity, rather than all the possible combinations of transmitter receiver spacings, is the lack of compensation for borehole disturbances, which is achieved by averaging the difference between each pair of transit times. The result is a DELTA-T in µs/ft.
Density data were recorded from 551 to 1055 mbsf and were edited for unrealistic values postcruise (Shipboard Scientific Party, 1997, Site 999). A constant value of 1.9 g/cm3 (density value at 551 mbsf) was used to extend the density data up to 525.8 mbsf, corresponding to the first sonic velocity measurement. An examination of the velocity and density data vs. resistivity data from three different measurements of varying resolution and borehole penetration reveal a remarkable correlation throughout the entire hole, indicating that borehole effects are not adversely affecting the traveltime or density measurements (Fig. 4).