No abstract

[1] Near-bottom magnetic measurements on board submersible Nautile were carried out on the Mid-Atlantic Ridge 21°40 0 N segment, and deep-sea geomagnetic vector anomalies along 19 dive tracks were obtained by applying the processing method for shipboard threecomponent magnetometer data. A forward modeling technique using short-wavelength components of the anomalies arising from local topography and vertical motion of the submersible was designed to estimate the absolute magnetization intensity of the seafloor. In the vicinity of the spreading axis a considerable number of magnetization estimations are reliably confirmed by the high correlation between observed and modeled anomalies, whereas less reliable estimations are obtained off-axis, probably because the sediment buries the basement topography. The natural remanent magnetization (NRM) measured on basalt samples collected during these dives is compared with the magnetization from anomalies. Though both results give a similar range of magnetization intensity, no correlation is confirmed between them, possibly because the magnetization from anomalies represents laterally averaged seafloor magnetization, whereas the NRM has variations at the scale of individual pillow or lava pile. Equivalent magnetization inverted from the sea-surface magnetic anomalies shows axial magnetization increases significantly from the segment center to the segment ends. However, the results of eight dives conducted near the spreading axis at different locations along the segment show much less variation in magnetization intensity along the axis. We ascribe the high equivalent magnetization at segment ends to preferential serpentinization of peridotite near the segment ends and the associated formation of magnetite. The results of three across-axis transects composed of 15 dives running in the spreading direction can be consistently interpreted as recording geomagnetic paleointensity variations during the Brunhes epoch. Although magnetization lows are generally correspondent to periods of low paleointensity, they show deeper drop than predicted from the paleointensity variation.Citation: Honsho, C., J. Dyment, K. Tamaki, M. Ravilly, H. Horen, and P. Gente (2009), Magnetic structure of a slow spreading ridge segment: Insights from near-bottom magnetic measurements on board a submersible,

Separation of segments in two groups ("hotter" and "colder") based on AMBA shows different types of variations among segment, with grouped amplitudes at segment centers and scattered amplitudes at segment ends for hotter segments and the opposite for colder segments. These observations support shallow fractionation and the presence of serpentinized bodies in the vicinity of discontinuities as the major processes which control the axial magnetic anomMy amplitude variations.

Summary During cruise Tammar (May 1996) of R/V Nadir and submersible Nautile, 66 basaltic rock samples were collected along two long cross‐sections encompassing the oceanic crust between 0 and 1.5 Myr at the centre of the mid‐Atlantic ridge segment located at 21°N. This dense sampling provides a means to investigate natural remanent magnetization (NRM) variability with age using rock magnetic studies (high‐field and k–T experiments, palaeofield intensity determinations by the original Thellier method) together with the good geophysical and geological knowledge of this peculiar segment. NRM intensities range from 1.3 to 25.4 A m−1 but do not display any rapid exponential decay with age, as expected. Despite the scatter, they seem to present short‐wavelength variations consistent on both flanks of the two lines. Because of the relative uniformity in grain size (i.e. single domain, SD), ulvospinel content (i.e. x = 0.6) and amount of magnetic minerals in Tammar samples, the observed across‐axis NRM intensity variations may be due either to oxidation degree variations or to geomagnetic field intensity changes. Curie temperatures display no increase towards the flanks nor a clear relationship with NRM intensities, suggesting that oxidation degree is not the major process controlling the NRM variations. Half of the collected samples, either fresh or highly altered, provide good‐quality palaeointensity determinations. Depending on the alteration degree of samples, two major types of NRM/TRM diagrams are observed. For relatively fresh materials, only one line segment and the major part of NRM is used for field intensity calculations. For moderately or highly altered specimens, two line segments are observed with the first most probably corresponding to thermal demagnetization of the original remanence, and the second to the product resulting from chemical modification (i.e. inversion). Despite the difference in the amount of NRM that can be thermally demagnetized, samples of similar age but of different non‐stoichiometric degree give coincident palaeofield strengths. Palaeofield values, from 15 to 62 µT (i.e. VADM from 3.4 to 13.6 × 1022 A m2), are in good agreement with the global palaeointensity database, and display coherent short‐wavelength undulations with age, consistent (assuming a 1‐km‐wide neo‐volcanic zone and steady, symmetrical spreading) with the relative palaeointensity record deduced from sediment core analysis (Guyodo & Valet 1999). Comparison between NRM and field intensity variations with age reveals a remarkable coincidence between both signals, suggesting that the NRM intensities are more sensitive to the field intensity than to the alteration degree. These observations suggest that magnetic particles are titanomaghemites, generated during the initial magma cooling and storing a durable record of the geomagnetic field intensity. The type of remanence is either (1) a chemical remanent magnetization that closely mimics the properties of the initial NRM or (2) a thermoremanent magnetization (the maghemit...