We used continuous measurements of physical properties of the cores: the MST aboard JOIDES Resolution (Sigurdsson, Leckie, Acton, et al., 1997) and the split core multisensor core logger (MSCL, GEOTEK™, Surrey, UK; Weaver and Schultheiss, 1990; Gunn and Best, 1998) from the Geosciences Department of Bremen University, which we temporarily took to the Bremen ODP repository. The MSCL includes instruments that measure magnetic susceptibility and density. Magnetic susceptibility data were collected at 1 cm intervals using a Bartington™ point sensor, which is more accurate for measuring half cores and gives much higher spatial resolution than the loop sensor. Densities were also determined from GRAPE measurements recorded at 1 cm intervals (Boyce, 1976).
Shipboard GRAPE density values (obtained by MST) do not match the magnitude of the shipboard wet bulk density data (index properties of individual samples) below 400 meters below seafloor (mbsf) at Site 999 (Sigurdsson, Leckie, Acton, et al., 1997). The cores at these depths do not fill the liner because of the increased induration of the sediments. The gamma-ray attenuation processing aboard ship assumed a core diameter corresponding to a filled liner; therefore, the resulting GRAPE density values are too low (Sigurdsson, Leckie, Acton, et al., 1997). In order to compare shore-based and shipboard GRAPE measurements, the shore-based GRAPE density values are normalized to the shipboard GRAPE density data.
The chemical element composition of the cored material was analyzed using a new XRF core scanner. The XRF core scanner is a non-destructive analysis system for scanning the surface of archive halves of cores. The instrument was developed and built at the Netherlands Institute for Sea Research (NIOZ, Texel). The general method and some calibration procedures are described by Jansen et al. (1998).
Here we present the specifications and procedures of the system at the Geosciences Department of Bremen University and how we adapted it to the LPTM cores. The XRF core scanner is installed within a standard 20-ft container to allow easy transport. Both the cover and the core fit system are pneumatically activated. The central sensor unit consists of a molybdenum X-ray source (3-50 kV) and a Peltier-cooled PSI detector (kevex™) with a 125 µm beryllium window and a multichannel analyzer with a 20 eV spectral resolution (Fig. 2). The whole system is computer controlled. The scanner electronics allow precise positioning capabilities and include an integrated safety interlock mechanism. Our system configuration (X-ray tube energy, detector sensibility) allows the analysis of elements from potassium (K, atomic number 19) through strontium (Sr, atomic number 38; 20 kV X-ray voltage). The analyses are performed at predetermined positions and counting times. The measurement unit (X-ray source, detector) is moved along the Z axis, the plastic prism is lowered on the core surface (covered by special foil) during analysis; a slit defines the dimensions of the irradiated core surface (here: 1 cm2). To avoid loss of energy because of scattering in air, the area of analysis is flushed by helium (between prism and detector, covered by a condom, and within the prism). The core is moved along the X axis (Fig. 2). The XRF data are collected at 1 cm intervals over a 1 cm2 area, and test run calibration resulted in the use of 15 s count time and an X-ray current of 0.15 mA to obtain statistically significant data of the elements we were interested in (e.g., K, Ca, Fe, Ti, Mn, Sr). In this paper we present Ca and Fe intensity data, which are highly correlated to the physical and chemical properties measured both downhole and in cores. The relatively indurated LPTM cores were cut by a saw, and therefore, already had an ideally flat core surface, which is needed for successful XRF analysis. Most of the rotary-drilled cores are broken into several centimeter to decimeter long pieces. Special effort was taken in preparing the cores carefully by bringing all the single pieces up to one level. The response of the elements also depends on the wavelength of the fluorescent radiation; the penetration depth of the XRF analysis is on the order of tenths of millimeters (Ca) and hundredths of millimeters (Fe) deep. Therefore, special accuracy was needed for cleaning cutting residues from the core surface. The acquired XRF spectrum for each measurement is processed by the kevex™ software Toolbox©. Background subtraction, sum-peak and escape-peak correction, deconvolution, and peak integration are successively applied. The resulting data are basically element intensities in counts per second. Element concentrations (e.g., in percent or parts per million) are not directly available, but by comparisons with data from standard chemical analyses from discrete samples, these counts can be converted to element concentrations (Jansen et al., 1998).
Downhole logging measurements of the borehole wall at Sites 999 and 1001 were accomplished by a suite of different instruments including the long-spaced sonic tool (sonic velocity), natural gamma-ray tool (natural gamma-ray activity), lithodensity tool (bulk density), the dual induction-spherically focused resistivity tool (three different measurements of resistivity), and the FMS tool (described below). In addition, dry weight fractions of major oxides were derived at Site 999 using data from the geochemical tool, and at Site 1001, magnetic polarity reversals were obtained using data from the geological high-sensitivity magnetic tool. A complete description of individual logging instruments is given in the Ocean Drilling Program Manual (Borehole Research Group, 1990) and in Sigurdsson, Leckie, Acton, et al. (1997), for Leg 165 downhole measurements in particular. FMS data and data processing are detailed below because of the introduction of "calibrated" FMS data and the importance of FMS data to core log integration.
The FMS produces high-resolution images of the resistivity character of the borehole wall that can be used for detailed sedimentological and/or structural interpretations (Ekstrom et al., 1986; Bourke et al., 1989; deMenocal et al., 1992). The FMS tool comprises 16 electrode "buttons" on four orthogonal pads that are pressed against the borehole wall. The electrodes are spaced 2.5 mm apart and are arranged in two diagonally offset rows of eight electrodes each. A focused electrical current flows between electrodes and variations in current intensity are measured. These measurements reflect resistivity variations, are recorded every 0.25 cm, have a 0.5 cm vertical resolution, and have a maximum penetration depth of 25 cm. In contrast, conventional downhole measurements are recorded every 15 cm and have vertical resolutions ranging from 15 cm to 2 m (deMenocal et al., 1992). Thus, the sampling rate for FMS is 60 times greater than conventional logs, with resolution capabilities 30-400 times greater than conventional logs. This fine-scale resolution makes FMS the downhole measurement that is most comparable in scale to measurements made on whole core and individual samples, as well as providing visual comparisons to core photographs.
The FMS tool string contains a general purpose inclinometry tool (GPIT) that orients the measurements relative to magnetic north through the use of a magnetometer and records variation in uphole speed through the use of an accelerometer. Standard processing utilizes information from the GPIT and converts variations in current intensity as recorded on 16 individual electrodes on each pad into spatially oriented, variable-intensity color or gray-scale images of the borehole wall. The FMS images reveal relative variations in borehole resistivity but do not provide a direct quantitative value of formation resistivity.
Resistivity in this environment is primarily controlled by electrolytic conduction of pore fluids and/or cation exchange on the surfaces of clays and other conductive minerals (Bourke et al., 1989). Thus, images of resistivity variations reflect changing chemical and physical properties of the borehole such as porosity, mineralogy, induration, grain size, and chemistry of pore fluids.
The standard presentation of FMS data as variable-intensity gray-scale images are often visually similar to split cores or black and white photographs of the cores. Although visual correlations can be striking, they can not easily be quantitatively interpreted. In addition, all other core measurements are presented as a digital depth series of discrete data values, which are not always easily correlated to gray-scale images. In order to expand the use of FMS data as a link between core and all other discrete measurements, FMS relative resistivities are scaled to the absolute resistivity measured with the shallow spherically focused resistivity log (SFLU), and a digital depth series of scaled FMS resistivity values are extracted.
Scaling of FMS data to measurements of absolute formation resistivity was accomplished at the Laboratoire de Mesures en Forage, IMT, Marseille, France, using the proprietary Schlumberger™ software module known as "BORSCA©." Scaled FMS data are displayed as a single "wiggle" trace of resistivity (ohm-m) with depth calculated every 1.27 cm, and represent the sum of scaled values from 16 electrodes on one of four orthogonal pads. This electrode averaging and use of one pad is appropriate given the flat-lying strata imaged with the FMS in the depth intervals discussed in this paper. Scaled (calibrated) FMS data are compared to "raw" FMS data over the same depth interval in Figure 3.
The images from four tool pads cover ~22% of the borehole wall for each pass; however, the FMS images presented in this paper are shifted closer together for display purposes and thus appear to cover a greater percentage of the borehole (e.g., Fig. 4, Fig. 5, Fig. 6). Calipers on the FMS tool provide precise measurements of borehole diameter in two orthogonal directions. Use of the FMS is restricted to hole diameters <38 cm (15 in). All depth intervals presented in this study have hole diameters between 28 and 35 cm.
FMS data in this paper are displayed as a combination of standard gray-scale images and the calibrated trace from one of four pads. In the gray-scale FMS images, black represents the least resistive values and white the most resistive ones. For the cores examined in this paper; calcareous horizons are characterized in the image by light gray to white; intermediate grays reflect calcareous-clayey mixed sediments; distinct volcanic ash layers show up as dark gray to black sharp-bounded thin layers, and clay-rich intervals (not necessarily altered ash) are also characterized by black, low resistivity intervals. Disseminated chert and/or zones of silicification may also be evident as irregularly shaped white areas. The FMS trace acts as a visual "link" between the FMS image and all other downhole and core measurements, which are also presented as discrete values with depth. The FMS data play an important role in reconstructing the sequences with incomplete recovery.
Depths for downhole measurements data are initially determined by the length of wire suspended below the rig floor adjusted for the position of a particular tool on the tool string. Wireline length is then calibrated to the position of the end of the drill pipe, which is at a known depth below the dual-elevator stool on the rig floor, which in turn, is at a measured elevation above mean sea level. The "bottom felt depth" representing the first contact of the drill string with seafloor is used to convert core and log depths to mbsf. Downhole measurements made during different trips into the open hole are depth calibrated to each other using a natural gamma-ray spectrometry tool (NGT), which is included on every tool string, including the FMS, as a common correlative measurement. Downhole measurements presented in this paper are those found on the CD-ROM in Sigurdsson, Leckie, Acton, et al. (1997), which have been depth shifted by the Lamont-Doherty Earth Observatory Borehole Research Group (LDEO-BRG). Regardless of the absolute depth value, relative thicknesses and spacing of features displayed on FMS and other downhole measurements should be identical within the limits of varying vertical resolution capabilities.
Although both core and log depth values are given relative to drill-string measurements, numerous factors including incomplete core recovery result in depth mismatches between core and downhole measurements. Depths (mbsf) given in this paper are from the FMS tool string; thus, where depth discrepancies are found, core depths are shifted to match FMS downhole measurements.