PALEOMAGNETISM

The natural remanent magnetization (NRM) of archive-half sections were initially measured and then remeasured after alternating-field (AF) demagnetization at selected levels. Sections obviously affected by drilling disturbance were not measured. Core 202-1233A-1H was demagnetized at peak fields of 10, 15, and 20 mT. Cores 202-1233B-1H through 3H were demagnetized at 15 and 20 mT. Cores 202-1233B-4H through 11H were demagnetized at 20 and 25 mT. All other cores (202-1233C-1H through 13H, 202-1233D-1H through 13H, and 202-1233E-1H through 8H) were demagnetized at 25 mT only.

Initial NRM intensities were very high, ranging from 0.2 to 3.25 A/m (Fig. F17). The high NRM intensities are due largely to a drill string magnetic overprint. This overprint, characterized by steep positive inclinations (averaging +59° in Hole 1233B), was substantially reduced (typically by 75%) by demagnetization at peak AF fields as low as 10 mT. NRM intensities, though substantially reduced, were still quite strong (averaging near 0.1 A/m) after 25-mT AF demagnetization. A few intervals that will be discussed below had much lower intensities (1 x 10^-3 A/m). AF demagnetized inclinations (20 or 25 mT), aside from a few intervals, averaged approximately -52° (Fig. F18). This value is close to the expected inclination for an axial geocentric dipole (-59°) at this site latitude (~40°S) and is an indicator that most of the drill string overprint was removed by 25-mT AF demagnetization. Declinations demagnetized at 25 mT also appear to remove the drill string overprint, whereas prior to demagnetization declinations were biased toward 0° or 360°, which is also consistent with a drill string overprint.

The strength of the NRM intensity is probably related to two factors. First, these sediments are composed of homogenous fine-grained terrigenous material of an Andean and Coastal Range provenance with minor biogenic components (see "Lithostratigraphy") that contains abundant detrital Fe oxides. Second, prior to demagnetization, the magnetic field imparted by the drilling process results in a strong ubiquitous magnetic overprint. In most instances at Site 1233, this overprint is removed. However, for some cores, the drilling overprint is much larger (higher NRM intensity prior to demagnetization) and incompletely removed even after demagnetization (Fig. F18). This harder overprint is present during times when the APCT tool was used (see Lund et al., this volume). It is thought that this larger and harder overprint results from the longer time that the core barrel sits in the mud when this tool is being used.

Figure F18 displays the paleomagnetic field directions for Holes 1233B, 1233C, 1233D, and 1233E after AF demagnetization at 25 mT. For the upper few cores the high positive inclinations associated with the drilling overprint are removed by 20 mT. Below that, an additional increment of AF demagnetization (25 mT) is required to recover inclinations closer to expected values (approximately -59°). Negative inclinations dominate the sediments in all cores. Near the bottom of each hole, positive inclinations probably associated with the drilling overprint are still observed in narrow intervals. We were unwilling to demagnetize these sediments further so as to preserve the NRM for shore-based paleomagnetic measurements. The drilling overprint appears to affect inclination more than declination. In Figure F18, the declinations presented have for each core been rotated by assuming their average declination was 0°. Such a rotation is generally consistent with the tensor orientation tool, which was used from the third core down in all holes drilled at this site. The declination variation within individual cores is in most cases consistent with that expected of normal secular variation (<40°). The one exception is an apparent reversal in both inclinations and declinations near 68 mcd. We interpret this to be a magnetic field excursion and discuss it in more detail below. Overall, these data suggest that the sediments at Site 1233 are of normal polarity and Brunhes Chron (0-0.78 Ma) in age.

Shipboard measurements also suggest that a record of paleomagnetic secular variation (PSV) is preserved in Site 1233 sediments (Fig. F18). Initial comparison of the inclination and declination variability from hole to hole shows that this can be correlated (Fig. F19). If shored-based paleomagnetic studies corroborate these measurements and demonstrate that a high-fidelity PSV signal is recorded, then Site 1233 would provide the first pre-Holocene PSV record in the Southern Hemisphere from the eastern South Pacific to South American region. Currently the only published PSV data from this region come from Holocene lakes of Argentina. A summary of PSV data for the last 3000 yr together with historical observations of the geomagnetic field at the same location are shown in Figure F20A (from Lund and Constable, unpubl. data). The paleomagnetic results in Figure F20A have been dated with several bulk radiocarbon dates (calendar year corrected), and the estimated ages are consistent with the historical observations. The PSV variability at Site 1233 is shown in Figure F20B. The chronology for this interval has been determined from several accelerator mass spectrometry (AMS) radiocarbon dates from nearby piston core GeoB3313-1 (Lamy et al., 2001), which have been transferred to Site 1233 based on magnetic susceptibility correlation. The two PSV records in Figure F20 can be correlated, and selected PSV features are noted (D1-D4 and I1-I2). The apparent ages of these features are similar (~±200 yr). This comparison provides an independent assessment of PSV data quality and surface age estimates at Site 1233.

An apparent geomagnetic field excursion is recorded between 65 and 70 mcd in Holes 1233B, 1233C, and 1233D. The NRM (25 mT) shows an interval of steep positive inclinations and declinations ~180° opposed to the other declination values within these cores. One example of this record from Core 202-1233D-8H is displayed in Figure F21. NRM intensities are also low within this interval, suggesting that this may represent one of the Brunhes age geomagnetic excursions. Significant shore-based work will be needed to examine this interval and assess its reliability. Based on estimated sedimentation rates for the surficial sediments described above, the Laschamp geomagnetic field excursion (~41 ka) is the most likely candidate for this excursion. The thickness over which this excursion is recorded in Hole 1233B (~2 m) (Fig. F21) suggests that it may represent one of the highest resolution records of this geomagnetic phenomenon ever observed.

Under ideal circumstances, an estimate of the relative changes in past geomagnetic field strength (paleointensity) can be recovered from deep-sea sediments. Relative paleointensity can be estimated by normalizing the demagnetized NRM intensity by a proxy that can account for changes in the remanence-carrying magnetic material in the sediment. Under ideal conditions, the residual should reflect changes in geomagnetic field strength. The sediments at Site 1233 appear to record a stable magnetization that is consistent with what is expected for the site latitude and therefore may provide such a record. In order to assess this potential, the shipboard NRM data were normalized with the whole-core MST-derived magnetic susceptibility data. Though far from being an optimal paleointensity proxy, the susceptibility-normalized NRM can provide a first approximation of whether these sediments can be used for this purpose. Results (Fig. F22) show a general reproducibility in the ratio among four holes drilled at Site 1233. Low ratios in the excursional interval (~65-70 mcd) as well as the overall pattern of variability are consistent with what is known about paleointensity record for the last 150 k.y. Extremely low values at 95.5-99 mcd and 113.5-116 mcd may reflect early sediment diagenesis (see below) and must be considered with caution.

Site 1233 is generally characterized by high values of both parameters, presumably due to the high concentration of detrital magnetic grains in this siliciclastic-rich sediment. However, within two intervals (95.5-99 and 113.5-116 mcd) in the lower part of the record, magnetic susceptibility drops by an order of magnitude, from ~250 to ~35 instrument units, and NRM intensities by almost two orders of magnitude, from ~0.1 A/m to 2 x 10^-3 A/m (Fig. F23). The larger drop in NRM intensity relative to susceptibility suggests that the finer-grained remanence-carrying part of the magnetic fraction has been preferentially reduced and/or dissolved to nonferromagnetic phases relative to the coarser-grained component. The fact that this does not always occur suggests time-dependent sediment diagenesis that may reflect stronger reducing conditions. The anomalous magnetic directions within these intervals (Fig. F18) may be at least partially due to these conditions. Anomalously low NRM and MS values were not observed in the interval recording the excursional magnetic directions at ~65-67 mcd.