GP34A-01 INVITED 16:00h
The Concept of Linear Unmixing - a Key for Interpreting Environmental Magnetic Signals
In recent years, environmental magnetism has become a powerful examination method in many research fields. The interpretation of the results, however, requires careful analysis since many simultaneously occurring physical and chemical processes depending on each other and acting over different scales in time and space, may be involved. In order to resolve this complexity, the concept of linear unmixing has become an important empirical interpretation method, which can be used also for quantitative analyses. The basic assumption -- the coexistence of endmembers -- is almost always satisfied. Three examples - ranging from palaeomagnetism to environmental pollution monitoring - will be given, illustrating the importance of the linear unmixing concept. {\it Palaeomagnetism} Continuous records of the palaeomagnetic field are accessible via the measurement of the natural remanent magnetisation of stratified sediments. As the sediment acts as a non-linear filter, the field record may be distorted causing misinterpretation. This distortion depends on the influence of the environment on the magnetic mineralogy of the sediment, which is -- in the case of loess sediments -- composed of coexisting pedogenic and detrital endmembers with specific magnetic properties. Their variable contribution determines the degree of distortion. Transitional polarity records in loess/palaeosol sediments may not always reflect real geomagnetic field behaviour but rather lithological variations. {\it Climate} Proxy records, e.g. magnetic low-field susceptibility, are {\it en vogue} to characterise sediment stratigraphies. They reflect a climatic signal, but in essence, the proxy material has acted as a filter, and the climatic information is resolved just roughly. Using the linear unmixing concept, proxy data can be decoded and valuable information is accessible via the stratigraphic variation of endmembers, (e.g. detrital, pedogenic, biogenic). A profile from a gleyic luvisol has been unmixed resulting in four different remanence components. Their variation through the soil profile allows conclusions about formation and alteration of magnetic phases during the soil development, hence throwing some light on the pedogenic processes. {\it Pollution} Linear mixing (or unmixing) is able to recognise and to quantify natural and anthropogenic components. Several air filters collecting dust from roads with variable pollution degree were investigated. Two remanence components were observed in all filters. One component, which is directly related to the exhaust pipe emissions of motor vehicles, can be used for an empirical quantitative estimate of the mass contribution of exhaust emissions. Independent chemical cross-calibration confirms the magnetic measurements as a fast and inexpensive method for systematic pollution monitoring of wide areas with passive sampling methods.
GP34A-02 16:15h
Dust and Nutrients in Modern and Late Quaternary Ecosystems of Central Colorado Plateau Drylands
The inputs of rock-derived nutrients to ecosystems, along with pedogenic and geomorphic processes that distribute nutrients in substrates, are critical aspects of ecosystem dynamics. Stratigraphic, geochronologic, magnetic, chemical, biogeochemical, and textural studies of late Quaternary surficial deposits (including modern and buried soils) on the central Colorado Plateau reveal important ecological roles for far-traveled dust and document the redistribution of deposited dust across the landscape through geomorphic processes. In vast upland settings underlain by nutrient-poor Paleozoic and Mesozoic eolian sandstone, alternating episodes of dune activity and soil formation during the late Pleistocene and Holocene have produced dominantly sandy deposits that support grass and shrub communities. In these deposits, isothermal remanent magnetization (IRM) acquired at 0.3 Tesla provides a measure of eolian dust abundance. Magnetic and petrographic studies indicate that IRM is carried by silt-size Ti-magnetite and is unaffected by pedogenic growth or destruction of iron oxides. Such magnetite is absent in local bedrock, and it thus represents far-traveled dust. Magnetite (dust) abundance, redistributed by modern slope processes, increases systematically down shallow-gradient hillslopes, and it correlates strongly with potential nutrients (e.g., K, Na, P, Mn, Zn). Magnetite and potential nutrients also covary with depth in auger holes (2-3 m) and soil pits in dune-crest, dune-swale, and other geomorphic settings. Magnetite abundance further correlates strongly with elements (e.g., Ti, Li, As, Co, Ni, Th, La, Sc) that are geochemically stable in these environments. Amounts of magnetite and associated potential nutrients vary greatly (as much as 5x) in deposits accumulated over the past ~20-30 ka. Magnetite abundance and its relations to particle-size distribution, hematite abundance, and elemental abundances change sharply in sediment corresponding to 15-10 ka and 5-4 ka, and to deposition over the last century. These changes likely reflect shifts in source regions, the most recent one perhaps in response to human activity in and beyond the Colorado Plateau. Changes in dust accumulation in these surficial deposits during the late Quaternary are important to modern ecosystem dynamics because some plants today tap nutrients deposited as long ago as about 12-15 ka and because eolian dust at depth influences soil-moisture capacity.
GP34A-03 16:30h
Determination of Relative Contributions from Marine and Terrestrial Sediment Sources in the Cariaco Basin using a Magnetic Mixing Model
The Cariaco Basin, an anoxic basin off the coast of Venezuela, acts as a depocenter for sediments of both marine and terrestrial origin. Sedimentation rates for the upper 170 m of sediment recovered during ODP Leg 165 at Site 1002 average 35 cm/k.y. and are fairly constant over the entire 600-k.y.-long interval cored. We have used a magnetic mixing model to identify and quantify the contributions to the sediment at this site from marine and terrestrial sources during the past 27,000 years. Our approach is based on a broad array of environmental magnetic parameters. Using the appropriate statistical methods, we first determine the extent to which different pairs of parameters are interrelated. In the case of the Cariaco Basin, many of the parameters can be shown to vary coherently. We then use the extremal values of each particular magnetic parameter to determine the magnetic signature of two hypothetical source materials. A magnetic mixing model is used to calculate the contribution at any point in the core from each of the sources. The magnetic signatures of the two hypothetical sources can be compared to actual source materials that have contributed sediment to the basin. For the Cariaco core, recent geochemical studies have shown that variations in Ti and Fe content correlate with precipitation and that wetter intervals are associated with more riverine (terrigenous) input and higher Ti and Fe content while dryer intervals are associated with more marine (biogenic) input and lower Ti and Fe content. The relative contributions from our two hypothetical sources, as determined by the magnetic mixing model, show a good correlation with variations between higher and lower Ti and Fe content. This suggests that we have successfully determined the magnetic signatures of the marine and terrestrial components in the sediment.
GP34A-04 16:45h
Holocene Variations in the Strength of the North Atlantic Deep Water, a Magnetic Approach
Precise documentation of the evolution of the thermohaline circulation (THC) during the Holocene period is raising an increasing interest among the community. Model studies suggest that the strength of the deep water current in North Atlantic is tightly linked to the formation rate of deep-water via the meridional transport of ocean heat. The understanding of the natural variability of the THC during the entire Holocene in North Atlantic is therefore of critical interest for the understanding of its future evolution under the effect of global greenhouse-gas warming. We report on detailed magnetic analyses of 9 Holocene marine sequences distributed from northern to southern Gardar and Bjorn drifts (south of Iceland, 53°N to 61°N), the Gibbs fracture zone (52°S) and from the Eirik drift (south of Greenland). The average sedimentation rate of these sequences varies between 20 and 80 cm/kyr. The cores have been taken on board the R.V. Marion Dufresne of the IPEV during the IMAGES P.I.C.A.S.S.O cruise in June 2003 except for one taken in 1977 with the R.V. J. Charcot. The magnetic parameters indicate that during the Holocene, the magnetic fraction is composed of magnetite with uniform mineralogy and grain size distribution in the pseudo-single domain range (a few micrometers). On the other hand, the concentration varies in space and in time on both long and short terms. In space, the amount of magnetite decreases, together with the magnetic grain size from North to South along the Gardar and Bjorn drifts illustrating a progressive deposition of the magnetic fraction derived from the basaltic Iceland-Faeroe province along the path of the Iceland-Scotland overflow water branch of the NADW. In time, a long-term decrease is observed in the amount of magnetites transported to the studied sites from the northern basaltic province of Iceland and the Faeroe islands. This decrease is not quantitatively compensated by the increase in the carbonate fraction. We interpret this decrease as being directly related to a decrease in the strength of the bottom current through Holocene. The short term features will be discussed for the sequences dated by 14C and compared to other records.
GP34A-05 INVITED 17:00h
Magnetic fabrics elucidate past climatic and environmental conditions
Loess and paleosol deposits, worldwide, have provided an important medium from which paleoclimatic conditions and paleoenvironments could be reconstructed. Within the field of environmental magnetism, the most fruitful data have been time series of various common magnetic parameters, such as, but not exclusively, bulk susceptibility and its frequency dependence, magnetic coercivity, remanent magnetizations and coercivity of remanence. Until some recent efforts by others in China and Siberia and our own in Alaska, the potential of the anisotropy of magnetic susceptibility (AMS) had been underestimated. In non-lithified poorly consolidated sediments offering very little textural differentiation, as is the case in Central Alaskan loess deposits, AMS is an indispensable tool. Loess intervals having undergone post-depositional deformations cannot provide reliable paleoclimatic data yet environmental conditions leading to the preserved deformations inferred from their magnetic fabrics do impose chronological constraints. Furthermore, loess intervals preserving magnetic fabrics of aeolian deposition enable the reconstruction of relative time series of paleowind directions from their magnetic lineations. These relative time series provide additional chronological constraints and suggest a paleoclimate glacial - interglacial cycle control over surface air circulation in Central Alaska. Other common environmental magnetic parameters as well as geochemical data could not identify post-depositional deformations or paleowind directions in the loess deposit studied herein. This underscores the resolving power magnetic anisotropy has and can have in non-lithified poorly consolidated sediments.
GP34A-06 17:15h
A Possible Magnetite/Maghemite Electrochemical Battery in the Magnetotactic Bacteria
Magnetotactic bacteria produce membrane-bound chains of magnetic minerals (magnetosomes) within their cells. The purposes of magnetosomes, which include orientation in a redox gradient (1), are not fully characterized. A better understanding will assist in interpreting the magnetofossil record. Magnetotactic bacteria are known to make magnetosomes of either magnetite or greigite, both of which are mixed valence minerals with cubic inverse spinel structures. Vali and Kirschvink (2) proposed that magnetite producers might transform their magnetosomes into a third ferrimagnetic spinel, maghemite, to generate energy: a sort of electrochemical battery. Simple calculations suggest that, by transforming magnetite to maghemite in oxidizing waters and back to magnetite in reducing waters, a bacterium could access ~90 mV of additional redox potential. The energy necessary to traverse a cm-scale redox gradient and tap this energy is less than the additional energy provided by complete oxidation of a single magnetosome. We are currently testing the battery hypothesis by several methods. Initial experiments involved innoculation of dense suspensions of Magnetospirillum magneticum AMB-1 into an oxygen gradient medium. After several days, samples were collected from the high-oxygen top and low-oxygen bottom of the medium and freeze-dried. Low temperature magnetic properties were measured to test for magnetosome oxidation, the presence of magnetite chains, and the size of the magnetic particles. Variation in the strength of the Verwey transition indicated that magnetosomes from high-oxygen conditions were more oxidized than those from low-oxygen conditions, as the magnetosome battery hypothesis would predict. The Moskowitz test (4, 5) was used to test for the presence of chains. While the strength of the test weakens with magnetosome oxidation, results from both high- and low-oxygen conditions suggest the presence of intact chains. Both low-temperature and hysteresis properties indicate that the magnetosomes remained as single-domain grains, rather than forming the superparamagnetic rims typical of abiotic oxidation. Initial results therefore support the magnetosome battery hypothesis. If a magnetosome battery does exist, then magnetofossils may indicate the presence of sharp redox gradients and provide information about chemical stratification in ancient environments. Magnetite/maghemite batteries would reflect more oxidizing conditions than greigite/mackinawite batteries; assuming the magnetofossil record can be extended through most of the Precambrian, greigite-derived magnetofossils may be expected to dominate the magnetofossil record of the Archean, before the rise of free oxygen. Some of the maghemitization observed in many magnetofossils may reflect biologically-controlled oxidation rather than diagenetic oxidation. 1. R. B. Frankel, D. A. Bazylinski, M. S. Johnson, B. L. Taylor, Biophys. J. 73, 994 (1997). 2. H. Vali, J. L. Kirschvink, in Iron biominerals R. B. Frankel, R. P. Blakemore, Eds. (Plenum Press, New York, 1990) pp. 97-115. 3. A. R. Muxworthy, E. McClelland, Geophys. J. Intl. 140, 101 (2000). 4. B. M. Moskowitz, R. B. Frankel, D. A. Bazylinski, EPSL 120, 283 (1993). 5. B. P. Weiss et al., EPSL 224, 73 (2004).
GP34A-07 17:30h
Oxygen isotopes as biogenic criteria for nanocrystal of magnetite?
Recent geochemical investigations have shown that physical and chemical properties of nanometer-sized objects do significantly differ from those of millimeter- and micrometer- sized particles. In sediments and sedimentary rocks, magnetite crystals of nanometer size are often neogenic. They can be formed either by a biogenic process or by inorganic reactions. It is therefore of primary importance to understand how magnetite nanocrystals can be formed to define quantitative tools to differentiate biotic from abiotic processes for paleoenvironmental reconstructions. A series of carefully controlled laboratory experiments was carried out to calibrate the oxygen isotopic fractionation between abiotitc nano-magnetites and water at low temperature. We found a 2% progressive increase in the magnetite-water oxygen isotope fractionation factor between 5 and 70 ° C. Our results stand in contrast to previous experimental data on the oxygen isotope composition of biogenic magnetite nanoparticles in which the fractionation factor was found to decrease with increasing temperature. We propose that this different dependence can be used to determine the abiotic/biotic contribution to the remanent magnetization of sediments if the temperature of formation is known.
GP34A-08 17:45h
Accurate Granulometry Of As-Substituted Nanophase Ferrihydrite Particles Using Static And Dynamic Magnetic Properties
Ferrihydrite typically occurs in the nanometer size range and is suspected to be the precursor common to diagenetic formation of goethite, hematite, and magnetite in near surface environments. Identifying and estimating the relative grain-size of ferrihydrite (Fh) can be achieved rapidly, albeit qualitatively, by low temperature magnetic susceptibility (Guyodo et al., 2003). We have now undertaken a program of cross-validated quantitative granulometry of Fh using multiple magnetic techniques and M\"{o}ssbauer spectral analysis between 4.2 and 300K. Fh is antiferromagnetic, but a superimposed magnetic moment can result from uncompensated Fe$^{3+}$ spins. Here we present results from synthetic ferrihydrite with varying amounts of arsenic substitution (0, 1, and 10 wt %). Static and dynamic magnetic techniques were used to elucidate the magnetic behavior of these samples and estimate the grain size. First, ZFC/FC magnetization curves obtained by cooling in zero field (ZFC) and then measuring the induced magnetization at stepwise increasing temperatures, from 2 K to 300 K, in a small applied field (B=5 mT). The sample was again cooled in the same small field (FC), and FC magnetization curves were obtained by measuring induced magnetization at stepwise increasing temperatures. Second, the samples were cooled, with an applied field of 50 mT and susceptibility was measured in stepwise increasing fields (50 mT) with changing frequency (1, 10 and 100 Hz) and in the temperature range 10-300 K. Finally, M\"{o}ssbauer spectra were obtained at temperatures from 4.2 K up to 300 K. Samples measured at 4.2 K displayed only sextets, indicating a complete magnetic order. A progressive size-dependent thermal unblocking was observed by the collapsing hyperfine field on warming until the spectrum was represented only by a doublet. The three techniques indicated very close values for grain size estimates, (5.5 nm for samples Arsenic-free and 5.0 nm for samples with Arsenic) and are also in agreement with our transmission electron microscopy (TEM) observations. Our success demonstrates the feasibility of accurate granulometry of ferrihydrite by low temperature magnetic and M"{o}ssbauer techniques.