Magnetic charachterization of surface sediments in the South Atlantic

Surface sediment samples representative for the tropical and subtropical South Atlantic (15°N to 40°S) were investigated by isothermal magnetic methods to delineate magnetic mineral distribution patterns and to identify their predominant Holocene climatic and oceanographic controls. Individual parameters reveal distinct, yet frequently overlapping, regional sedimentation characteristics. A probabilistic ('fuzzy c-means') cluster analysis was applied to five concentration independent magnetic properties assessing magnetite to hematite ratios and diagnostic of bulk and fine-particle magnetite grain size and coercivity spectra. The resultant 10 cluster structures establish an oceanwide magnetic sediment classification scheme tracing the major terrigenous eolian and fluvial fluxes, authigenic biogenic magnetite accumulation in high-productivity areas, transport by ocean current systems, and effects of bottom water velocity on depositional regimes. Distinct dissimilarities in magnetic mineral inventories between the eastern and western basins of the South Atlantic reflect prominent contrasts of both oceanic and continental influences.

Magnetic properties of ODP Site 175-1075

We present sediment magnetic and chemical analysis of cyclic ocean sediments of the upwelling region of the Lower Congo Basin (equatorial Atlantic). We investigated two >100-k.y. intervals from Ocean Drilling Program Site 1075 to analyze the hysteresis properties, sources of magnetic susceptibility, anhysteretic remanent magnetizations, thermomagnetic behavior, and element concentrations of Fe, Ca, Ti, Mn, and K using an X-ray fluorescence (XRF) core scanner. The upper interval was sampled between 14 and 32 meters composite depth (mcd; 0.09-0.21 Ma) and the lower between 141 and 163 mcd (1.31-1.54 Ma) at a resolution of 20 cm, which represents a temporal resolution of 2.0 and 1.3 k.y., respectively. XRF core-scanner data were acquired at 5-cm intervals. The measurements show that ferri(o)magnetic minerals have no significant influence on the cyclicity of the magnetic susceptibility, which is dominated by paramagnetic and diamagnetic minerals and reflects changes of sediment input from the Congo River. The Fe, Ti, K, and Mn concentrations covary with the magnetic susceptibility where high concentrations of these elements correlate with intervals of high susceptibility and low concentrations with intervals of low susceptibility. The Ca counts correlate well with the calcium carbonate concentration but do not show the same cyclicity as the other elements or the susceptibility. With the exception of the Ca concentration, which is significantly higher in the upper interval, and the magnetic grain size, which indicates that less fine grained magnetite is present in the lower interval, no significant differences in the properties of the upper and the lower intervals were detected.

Paleomagnetic, stable oxygen isotope ratios, and U/Th activities of ODP Site 172-1062

Geomagnetic excursions are recognized as intrinsic features of the Earth's magnetic field. High-resolution records of field behaviour, captured in marine sedimentary cores, present an opportunity to determine the temporal and geometric character of the field during geomagnetic excursions and provide constraints on the mechanisms producing field variability. We present here the highest resolution record yet published of the Blake geomagnetic excursion (~125 ka) measured in three cores from Ocean Drilling Program (ODP) Site 1062 on the Blake-Bahama Outer Ridge. The Blake excursion has a controversial structure and timing but these cores have a sufficiently high sedimentation rate (~10cm/ka) to allow detailed reconstruction of the field behaviour at this site during the excursion. Palaeomagnetic measurements of the cores reveal rapid transitions (<500 yr) between the contemporary stable normal polarity and a completely reversed state of long duration which spans a stratigraphic interval of 0.7 m. We determine the duration of the reversed state during the Blake excursion using oxygen isotope stratigraphy, combined with 230Th excess measurements to assess variations in the sedimentation rates through the sections of interest. This provides an age and duration for the Blake excursion with greater accuracy and with constrained uncertainty. We date the directional excursion as falling between 129 and 122 ka with a duration for the deviation of 6.5±1.3 kyr. The long duration of this interval and the fully reversed field suggest the existence of a pseudo-stable, reversed dipole field component during the excursion and challenge the idea that excursions are always of short duration.

Magnetic mineralogy and geochemistry of ODP Holes 137-504B and 140-504B samples

Leg 140 of the Ocean Drilling Program deepened Hole 504B to a total depth of 2000.4 m below seafloor (mbsf), making it the deepest hole drilled into ocean crust. Site 504, south of the Costa Rica Rift, is considered the most important in-situ reference section for the structure of shallow ocean crust. We present the results of studies of magnetic mineralogy and magnetic properties of Hole 504B upper crustal rocks recovered during Legs 137 and 140. Results from this sample set are consistent with those discussed in Pariso et al. (this volume) from Legs 111, 137, and 140. Coercivity (Hc) ranges from 5.3 to 27.7 mT (mean 12 mT), coercivity of remanence (HCR) ranges from 13.3 to 50.6 mT (mean 26 mT), and the ratio HCR/HC ranges from 1.6 to 3.19 (mean 2.13). Saturation magnetization (JS) ranges from 0.03 to 5.94 * 10**-6 Am**2, (mean 2.52 * 10**-6 Am**2), saturation remanence (JR) ranges from 0.01 to 0.58 * 10**-6 Am2 (mean 0.37 * 10**-6 Am**2), and the ratio JR/JS ranges from 0.08 to 0.29 (mean 0.16), consistent with pseudo-single-domain behavior. Natural remanent magnetization (NRM) intensity ranges from 0.029 to 7.18 A/m (mean 2.95 A/m), whereas RM10 intensity varies only from 0.006 to 4.8 A/m and has a mean of only 1.02 A/m. Anhysteretic remanent magnetization (ARM) intensity ranges from 0.04 to 6.0 A/m, with a mean of 2.46 A/m, and isothermal remanent magnetization (IRM) intensity ranges from 0.5 to 1683 A/m, with a mean of 430.7 A/m. Volume susceptibility ranges from 0.0003 to 0.043 SI (mean 0.011 SI). In all samples examined, high-temperature oxidation of primary titanomagnetite has produced lamellae or pods of magnetite and ilmenite. Hydrothermal alteration has further altered the minerals in some samples to a mixture of magnetite, ilmenite, titanite, and a high-titanium mineral (either rutile or anatase). Electron microprobe analyses show that magnetite lamellae are enriched in the trivalent oxides Cr2O3, Al2O3, and V2O5, whereas divalent oxides (MnO and MgO) are concentrated in ilmenite lamellae.

Sedimentology and geochemistry of ODP Site 184-1144

We analyzed sediment from Ocean Drilling Program (ODP) Site 1144 in the northern South China Sea to examine the weathering response of SE Asia to the strengthening of the East Asian Monsoon (EAM) since 14 ka. Our high-resolution record highlights the decoupling between continental chemical weathering, physical erosion and summer monsoon intensity. Mass accumulation rates, Ti/Ca, K/Rb, hematite/goethite and 87Sr/86Sr show sharp excursions from 11 to 8 ka, peaking at 10 ka. Clay minerals show a shorter-lived response with a higher kaolinite/(illite + chlorite) ratio at 10.7-9.5 ka. However, not all proxies show a clear response to environmental changes. Magnetic susceptibility rises sharply between 12 and 11 ka. Grain-size becomes finer from 14 to 10 ka and then coarsens until ~7 ka, but is probably controlled by bottom current flow and sealevel. Sr and Nd isotopes show that material is dominantly eroded from Taiwan with a lesser flux from Luzon, while clay mineralogy suggests that the primary sources during the Early Holocene were reworked via the shelf in the Taiwan Strait, rather than directly from Taiwan. Erosion was enhanced during monsoon strengthening and caused reworking of chemically weathered Pleistocene sediment largely from the now flooded Taiwan Strait, which was transgressed by ~8 ka, cutting off supply to the deep-water slope. None of the proxies shows an erosional response lasting until ~6 ka, when speleothem oxygen isotope records indicate the start of monsoon weakening. Although more weathered sediments were deposited from 11 to 8 ka when the monsoon was strong these are reworked and represent more weathering during the last glacial maximum (LGM) when the summer monsoon was weaker but the shelves were exposed.

Optimized approach to decision fusion of heterogeneous data for breast cancer diagnosis.

As more diagnostic testing options become available to physicians, it becomes more difficult to combine various types of medical information together in order to optimize the overall diagnosis. To improve diagnostic performance, here we introduce an approach to optimize a decision-fusion technique to combine heterogeneous information, such as from different modalities, feature categories, or institutions. For classifier comparison we used two performance metrics: The receiving operator characteristic (ROC) area under the curve [area under the ROC curve (AUC)] and the normalized partial area under the curve (pAUC). This study used four classifiers: Linear discriminant analysis (LDA), artificial neural network (ANN), and two variants of our decision-fusion technique, AUC-optimized (DF-A) and pAUC-optimized (DF-P) decision fusion. We applied each of these classifiers with 100-fold cross-validation to two heterogeneous breast cancer data sets: One of mass lesion features and a much more challenging one of microcalcification lesion features. For the calcification data set, DF-A outperformed the other classifiers in terms of AUC (p < 0.02) and achieved AUC=0.85 +/- 0.01. The DF-P surpassed the other classifiers in terms of pAUC (p < 0.01) and reached pAUC=0.38 +/- 0.02. For the mass data set, DF-A outperformed both the ANN and the LDA (p < 0.04) and achieved AUC=0.94 +/- 0.01. Although for this data set there were no statistically significant differences among the classifiers' pAUC values (pAUC=0.57 +/- 0.07 to 0.67 +/- 0.05, p > 0.10), the DF-P did significantly improve specificity versus the LDA at both 98% and 100% sensitivity (p < 0.04). In conclusion, decision fusion directly optimized clinically significant performance measures, such as AUC and pAUC, and sometimes outperformed two well-known machine-learning techniques when applied to two different breast cancer data sets.