Paleomagnetic investigations during Leg 206 consisted mainly of routine remanent magnetization measurements of (1) archive-half sections and (2) discrete samples from the working-half sections, with both being completed before and after alternating-field (AF) demagnetization. In addition, low-field magnetic susceptibility was measured on whole-core sections using the MST and on archive-half sections using the AMST. To investigate rock magnetic properties, we also conducted thermal demagnetization, anhysteretic remanent magnetization (ARM), and isothermal remanent magnetization (IRM) experiments on some of the discrete samples.
Measurements of remanent magnetization were made using an automated pass-through cryogenic magnetometer with direct-current superconducting quantum interference devices (DC-SQUIDs) (2-G Enterprises model 760-R). The magnetometer is equipped with an in-line AF demagnetizer (2-G Enterprises model 2G600) capable of producing peak fields of 80 mT with a 200-Hz frequency. The magnetometer is run and data are acquired by a program called LongCore (version 3.3) written by W.G. Mills (ODP) in the LabView programming language. This version of LongCore was last updated during Leg 197. Key parameters used within the program, including calibration constants for the SQUIDs and coil response functions, are given in Table T16.
The natural remanent magnetization was routinely measured every 2.5 or 5 cm along all sedimentary archive-half sections before demagnetization. Most of these sections were also progressively AF demagnetized up to 30 or 40 mT and measured after each step. Details about the measurement intervals and the demagnetization steps used for the sediments are given in "Paleomagnetism" in "The Sedimentary Overburden (Holes 1256A, 1256B, and 1256C)" in the "Site 1256" chapter.
We decided not to attempt to measure most of the archive-half sections from igneous basement cores because the strong intensities and large volumes of the samples often result in spurious measurements in the long-core magnetometer. Repeated tuning of the magnetometer is necessary, which often entails heating the SQUID coils to release trapped magnetic flux. If not constantly monitored and tuned, the additional magnetometer noise that results from measuring the strong-intensity split-core samples can lead to erratic results. We did, however, experiment with several of the sections that contained continuous pieces that had few or no fractures or gaps over their entire 1.5-m section length. For these sections, the sensor velocity on the magnetometer was set at 10 cm/s in order to avoid saturation of the magnetometer electronics. This saturation is due to high variations of the magnetic flux (flux-jump) induced by the rapid motion of highly magnetized material past the sensor. Details about the sections measured, the measurement interval, and the demagnetization steps used for these basalt sections are given in "Paleomagnetism" in "Basement Formed at Superfast Spreading Rate (Holes 1256C and 1256D)" in the "Site 1256" chapter.
The discrete samples were also measured with the cryogenic magnetometer. Typically, samples were demagnetized in steps of 1 to 5 mT from 0 to 80 mT. One to four samples were placed in the discrete sample tray and run as a batch. Throughout the cruise, we only placed samples in the 20-, 60-, 100-, and 140-cm tray positions to ensure that samples were not influenced by the magnetizations of adjacent samples.
Several basalt samples were progressively demagnetized using a Schonstedt Thermal Demagnetizer (model TSD-1), which demagnetizes the rock specimens by heating them to any specified temperature from room temperature up to 800°C and then cooling them in a low magnetic field environment (<10 nT).
To investigate rock magnetic characteristics of some of the discrete samples, ARM and IRM experiments were conducted and then samples were measured using the cryogenic magnetometer. The DTECH AF demagnetizer was used to impart ARM to discrete samples, using a 100-mT peak alternating field and a 0.05-mT DC field. The ARM was imparted along the -z-axis of the samples, which is accomplished by placing the samples in the DTECH sample holder with the sample orientation arrow pointing into the instrument and aligned with the biasing and demagnetizing field. We then progressively demagnetized the ARM at 5-mT increments from 0 to 80 mT using the cryogenic magnetometer. IRM was imparted to the discrete samples by a DC field generated in the ASC (Analytical Services Company) Impulse Magnetizer (model IM-10). We typically measured the remanence of the discrete samples after imparting an IRM of 1000 mT along the -z-axis, which is accomplished by placing the samples in the Impulse Magnetizer sample holder with the sample orientation arrow pointing out of the instrument and aligned with the biasing and demagnetizing field. The samples were then stepwise AF demagnetized at 5-mT increments from 0 to 80 mT using the cryogenic magnetometer. Backfield IRM (BIRM) experiments were conducted by imparting (1) an IRM of 1000 mT along the +z-axis, (2) a BIRM of 100 mT along the -z-axis, and (3) a BIRM of 300 mT along the -z-axis. After each step, the remanence was measured. In addition, we conducted IRM acquisition experiments by first demagnetizing the sample at 150 mT in the DTECH AF demagnetizer and measuring the remanence. Then IRM was imparted at 10, 30, 100, 150, 200, 300, 400, 500, 600, 800, 1000, and 1200 mT, with the remanence measured after each step.
Low-field magnetic susceptibility was measured for all whole-core sections as part of the MST measurements (see "Multisensor Track Measurements" in "Physical Properties"). The MST susceptibility meter (a Bartington model MS2 with an MS2C sensor, a coil diameter of 88 mm, and an operating frequency of 0.565 kHz) has a nominal resolution of 2 x 10-6 SI (Blum, 1997). Susceptibility was determined at 2.5-cm intervals using a 1-s integration time for each measurement and with five measurements at each interval. The "units" option was set on SI units, and the values were stored in the ODP Janus database in raw meter units. To convert to true SI volume susceptibilities, meter units should be multiplied by 10-5 and then multiplied by a correction factor to take into account the volume of material that passed through the susceptibility coils. Except for measurements near the ends of each section, this factor for a standard ODP core is ~0.7 (Blum, 1997). No correction was applied for any figures illustrating magnetic susceptibilities in the "Paleomagnetism" subsections of the "Site 1256" chapter. Hence, the units are given as raw meter values. Magnetic susceptibility of discrete samples obtained from the working half of cores was measured using a Bartington MS2 susceptibility meter with a dual-frequency MS1B sensor. In addition, magnetic susceptibility was measured every 2 cm on the archive half of cores using the point susceptibility probe (Bartington MS2 susceptibility meter with a MS2F sensor) on the AMST.
Even though results from the shipboard cryogenic magnetometer have been compared with many other laboratories and shown to give consistent results, it is useful to check the calibration of the magnetometer against a known standard at the beginning of each leg. We used a standard purchased from Geofyzika that is an 8-cm3 cube with an intensity of 7.62 A/m (moment = 6.096 x 10-5 Am2). All three axes gave results that differ <2% from this standard. In addition, the automated tray positioning was checked by putting the standard at known positions and measuring the tray. The position indicated by the software was found to be accurate to better than ±1 cm, which is reasonable given the stretch in the pulley system used to move the sample boat.
Based on tests conducted during Legs 186 and 200, the background noise level of the magnetometer in the shipboard environment is ~2 x 10-9 Am2 (Shipboard Scientific Party, 2000, 2003). We repeated tests during Leg 206 for an empty split-core tray (also referred to as the sample boat) and for the tray that holds discrete samples. The results were similar in that the x-, y-, and z-axis moments measured on the sample boat before cleaning were less than ±2 x 10-9 Am2 except for two intervals that were only slightly above this (Fig. F13A). After cleaning the sample boat with window cleaner (Windex brand) and demagnetizing it at 80 mT, the moments are all less than ±1 x 10-9 Am2 (Fig. F13B). These results include the drift correction, which only marginally changes the results (Fig. F13). During Leg 206, we always applied a tray correction to the split-core and discrete samples. The tray-corrected data are the measured magnetic moments for a sample minus those measured at the same position for the empty sample boat. The tray values were updated throughout the cruise by remeasuring the tray. The relative size of these values, however, should always be comparable to those shown in Figure F13 for the clean, empty sample boat. When the tray correction is applied to measurements made on a clean, empty sample boat, the moments drop to less than ±2 x 10-10 Am2 (Fig. F14).
Several discrete sample trays (DSTs) are available for use, but we only used the DST provided by 2-G Enterprises. This one is 150 cm long and has holders for discrete samples every 10 cm. The holders alternate between those capable of holding samples with 2-cm x 2-cm bases and those capable of holding samples with 2.6-cm x 2.6-cm bases. We oriented the DST such that the holders with 2-cm x 2-cm bases were at 20-, 40-, 60-, 80-, 100-, 120-, and 140-cm positions when the DST was in the sample boat.
To test the noise in the DST, it was placed in the sample boat, cleaned with window cleaner, AF demagnetized at 80 mT, and then measured every 5 cm in continuous mode. After the drift correction is made and the effect of the sample boat is subtracted, the noise level associated with the DST is similar to that of the cleaned sample boat (cf. Figs. F13B, F15). When measured in discrete mode and when tray-corrected, where the tray correction accounts for the magnetization of the sample boat and the DST, the four sample positions we used (the 20-, 60-, 100-, and 140-cm tray slots) had magnetic moments of <4 x 10-10 Am2.
The machine noise level is sufficiently low that it is of minor significance for most samples measured. The practical noise level is, however, higher by about an order of magnitude from the cleaned, drift-corrected, and tray-corrected examples above. As was apparent when the sample boat was not clean, just a little bit of dirt on the tray puts the noise level in the 10-9 Am2 range. Noise related to dirt on the sample boat or the DST and to the magnetization of the plastic core liners for split-core samples and plastic cubes for sediment discrete samples will at least be in that range. Even with diligence, it is difficult to keep the trays clean, given the amount of core material measured during a typical leg. For long-core measurements, it is not practical to clean and remeasure the empty tray before each new sample. Thus, the correction for the tray magnetization may only improve the accuracy of the measurements marginally for the shipboard environment. Moreover, the split-core sections are in a plastic core liner that has been stored in dusty conditions prior to coring and that resides in a metal core barrel just prior to core collection. Therefore, the noise associated with the core liners will likely be several times greater than the noise associated with the sample boat (i.e., several times greater than ±2 x 10-9 Am2).
We conclude that under favorable conditions the noise level will be approximately ±2 x 10-9 Am2. For a split core, given the large volume of core material within the sensing region of the magnetometer, which is ~100 cm3, the minimum measurable remanent intensities will be greater than ~2 x 10-5 A/m. For discrete samples, which typically have volumes of 6-10 cm3, the minimum measurable remanent intensities are greater than ~4 x 10-4 A/m.
The standard ODP paleomagnetic coordinate system was used. In this system, +x is vertically upward from and perpendicular to the split-core surface of the archive half, +z is downcore, and +y is orthogonal to x and y in a right-hand sense (i.e., it points left along the split-core surface when looking upcore at the archive half) (Fig. F16).
APC Cores 206-1256B-3H through 18H were azimuthally oriented using the Tensor tool. The Tensor tool consists of a three-component fluxgate magnetometer and a three-component accelerometer rigidly attached to the core barrel. The information from both sets of sensors allows the azimuth and dip of the hole to be measured as well as the azimuth of the double-line orientation mark on the core liner. Orientation is not usually attempted for the top two or three cores (~20-30 mbsf) as the bottom-hole assembly (BHA) is insufficiently stabilized by the sediment. The output from the Tensor tool, which contains a variety of angles including the inclination angle of the hole and the magnetic toolface (MTF) angle (Fig. F17) is archived in the Janus database. The inclination angle of the hole is a measure of deviation of the hole from vertical, and the MTF angle gives the angle between the magnetic north and the double-line orientation mark on the core liner. The core liner is always cut such that the double lines are at the bottom of the working half.
Using the ODP coordinate system for the archive and working halves or for samples taken from them, the measured remanent declination can then be corrected to magnetic north by adding the MTF angle and can be further corrected to true north by adding the deviation of magnetic north from true north, the latter of which can be estimated from the International Geomagnetic Reference Field (IGRF) coefficients. The equation is
where,
For hard rocks it is possible to combine images of the exterior of the core with images of the borehole wall from the Formation MicroScanner (FMS) logging tool. To allow for this possibility, the exterior of basalt cores pieces were scanned with the DMT CoreScan, which is a digital camera system that records an image of the whole-core piece as it is rotated (see "Digital Imaging"). Postcruise analysis of the FMS and digital images will allow the reorientation of some distinctive pieces of core to true geographical north.
Oriented discrete samples were taken from the working half of selected sections. For sedimentary intervals, one sample per core was typically collected for shipboard analysis using plastic cubes with a 2-cm x 2-cm x 2-cm exterior dimension and an aluminum extruder. The interior volume of a plastic cube is ~7 cm3. Each plastic cube has an arrow on the bottom face. The sediment is collected in the extruder with a 2-cm x 2-cm cross-section opening (Fig. F16). Usually a few millimeters of sediment is extruded and removed from the bottom of the extruder with a spatula, leaving a flat surface of sediment and just enough sediment in the extruder to fill the plastic cube. The sediment is then carefully extruded into the plastic cube. An orientation mark on the extruder aids in ensuring the upcore orientation of the sample is maintained, and a mark on the plunger in the extruder aids in ensuring that enough sediment is collected to fill the plastic cube. When measuring a sediment sample, we placed the side with the lid down in the tray with the arrow pointing along the -z-axis, or upcore, direction. This makes the discrete sample orientation the same as that of the archive half, allowing for a more direct comparison of the discrete sample measurements with archive-half measurements.
In lithified intervals, we drew an arrow on the split-core face pointing upcore and used the rock saw to cut the sample. For shipboard samples, we typically cut cubes with ~2-cm-long sides (volume = ~8 cm3), although we also collected smaller pieces for analysis of the drilling overprint. When measuring a hard rock sample, we placed the side with the arrow down in the tray with the arrow pointing along the -z-axis, or uphole, direction, which for the hard rock discrete sample makes the orientation the same as that of the archive half.
Several types of secondary magnetization were acquired during coring, which sometimes hampered magnetostratigraphic and paleomagnetic interpretation. The most common was a steep downward-pointing overprint attributed to the drill string. For the Leg 206 cores, we also observed a bias for 0° declinations in archive-half sections, which has been observed during many previous cruises and has been interpreted as a radially inward overprint (Fig. F16). Details of these and other magnetic complexities are presented in the "Paleomagnetism" sections in the "Site 1256" chapter.
In interpreting the data, we avoided using the measurements within 5 cm from the ends of each section, although we saved these values for future studies that might wish to deconvolve the remanence signal. We also remove data from intervals where drilling disturbance, gaps, or other features exist, as these intervals give erroneous results. A table of the intervals removed is presented in the "Paleomagnetism" sections in the "Site 1256" chapter.
The characteristic remanent magnetization (ChRM) was estimated in three ways. When only the highest demagnetization step used was above (or at least nearly above) the field needed to remove the drilling overprint, then the results from the remanence measured after this single demagnetization step were interpreted as the best estimate of the ChRM. When multiple demagnetization steps were measured at fields above that needed to remove the drilling overprint, the ChRM was estimated by principal component analysis (PCA) (Kirschvink, 1980) and from a Fisherian average (Fisher, 1953) of three or more of the stable endpoint directions.
The PCA analysis was conducted using a program that iteratively searches for the demagnetization steps that minimized the size of the maximum angular deviation, which is a measure of how well the vector demagnetization data fit a line. Maximum angular deviation values <10° are typically considered to provide lines that fit the observations well. The program requires that at least three demagnetization steps be used, never uses data from demagnetization steps lower than a user-defined value, and does not require that the best-fit PCA line pass through the origin of the plot (the "free" option of standard PCA). Because the drilling overprint persisted beyond 15-mT demagnetization in most of the intervals, we only used results from 15 mT or higher in the PCA.
For comparison, the program also computes a Fisherian mean of the highest three or four demagnetization steps for each interval. This is referred to as the stable endpoint direction. Typically, only the highest three demagnetization steps are used in the average, unless the mean of these three directions has a precision parameter of <200 (a measure of dispersion), in which case the fourth highest demagnetization step is included. In cases where the precision parameter is <200, the program will first search for outliers and remove them if they lie >10° from the mean. Both the PCA and stable endpoint directions for interpreted intervals are given in tables in the "Paleomagnetism" sections in the "Site 1256" chapter.
Comparison of the stable endpoint with the PCA direction can be useful for indicating where unremoved or partially unremoved magnetization components exist or where progressive demagnetization has been ineffective in revealing linear demagnetization paths. To assist with this comparison, we have computed the angular distance between the two directions and included it in the tables in the "Paleomagnetism" sections in the "Site 1256" chapter.
Where AF demagnetization successfully isolated the characteristic component of remanence, paleomagnetic directions were used to define polarity zones. Given the shallow inclinations recorded by the Site 1256 sediments, we relied heavily on the declinations for polarity determination. Ages for reversals are from the revised Cenozoic timescale of Cande and Kent (1995) (Fig. F12).