Three-dimensional magnetic stripes require slow cooling in fast-spread lower ocean crust.

Earth's magnetic field is recorded as oceanic crust cools, generating lineated magnetic anomalies that provide the pattern of polarity reversals for the past 160 million years1. In the lower (gabbroic) crust, polarity interval boundaries are proxies for isotherms that constrain cooling and hence crustal accretion. Seismic observations2-4, geospeedometry5-7 and thermal modelling8-10 of fast-spread crust yield conflicting interpretations of where and how heat is lost near the ridge, a sensitive indicator of processes of melt transport and crystallization within the crust. Here we show that the magnetic structure of magmatically robust fast-spread crust requires that crustal temperatures near the dike-gabbro transition remain at approximately 500 degrees Celsius for 0.1 million years. Near-bottom magnetization solutions over two areas, separated by approximately 8 kilometres, highlight subhorizontal polarity boundaries within 200 metres of the dike-gabbro transition that extend 7-8 kilometres off-axis. Oriented samples with multiple polarity components provide direct confirmation of a corresponding horizontal polarity boundary across an area approximately one kilometre wide, and indicate slow cooling over three polarity intervals. Our results are incompatible with deep hydrothermal cooling within a few kilometres of the axis2,7 and instead suggest a broad, hot axial zone that extends roughly 8 kilometres off-axis in magmatically robust fast-spread ocean crust.

Inferred time- and temperature-dependent cation ordering in natural titanomagnetites.

Despite years of efforts to quantify cation distribution as a function of composition in the magnetite-ulvöspinel solid solution, important uncertainties remain about the dependence of cation ordering on temperature and cooling rate. Here we demonstrate that Curie temperature in a set of natural titanomagnetites (with some Mg and Al substitution) is strongly influenced by prior thermal history at temperatures just above or below Curie temperature. Annealing for 10(-1) to 10(3) h at 350-400 °C produces large and reversible changes in Curie temperature (up to 150 °C). By ruling out oxidation/reduction and compositional unmixing, we infer that the variation in Curie temperature arises from cation reordering, and Mössbauer spectroscopy supports this interpretation. Curie temperature is therefore an inaccurate proxy for composition in many natural titanomagnetites, but the cation reordering process may provide a means of constraining thermal histories of titanomagnetite-bearing rocks. Further, our theoretical understanding of thermoremanence requires fundamental revision when Curie temperature is itself a function of thermal history.