The recently updated Global Paleomagnetic Database provides an opportunity to construct a preliminary model of long-term magnetic polarity reversal behavior for the last 570 My. After filtering for quality, a sub-set of data from this global database is used to estimate relative reversal frequency, and the results show the expected long intervals of low reversal rate during the Cretaceous Normal and Permo-Carboniferous Reversed Superchrons. In addition to the established superchrons, these new results indicate a long period of very low reversal rate in the Ordovician, which may be a previously unidentified superchron, and an additional short period of anomalously low reversals at the Jurassic/Triassic boundary. If the Ordovician reversal rate anomaly is a single polarity superchron, it represents the third major change in reversal rate within the Phanerozoic, with approximately 200 My between the CNS, the PCRS and the new Ordovician reversed interval. INTRODUCTION Long-term changes in geomagnetic field behavior, including large-scale variations in polarity reversal rate, are believed to be indications that conditions in the core and lower mantle have undergone fundamental change, specifically in the seismicallydefined D" transition zone at the bottom of the mantle [McFadden and Merrill, 1984; Loper and McCartney, 1986; Courtillot and Bess, 1987]. Two long periods without polarity reversals have been established for the Phanerozoic, the best documented being the Cretaceous Normal Superchron [the CNS; Helsley and Steiner, 1969; Irving and Pullaiah, 1976], and the Permo-Carboniferous Reversed Superchron [the PCRS; McElhinny, 1971; Irving and Pullaiah, 1976; Divertere and Opdyke, 1991]. These fundamental changes in reversal rate have been correlated with true polar wander [Courtillot and Besse, 1987], mantle plume activity (Larson, 1991a; Larson and Olson, 1991), global climate change [Larson, 1991b] and biological extinction events [Courtillot, 1990], although relationships between geomagnetic field changes and surface phenomena are presently a subject of considerable debate. Early investigations of field behavior used compilations of global paleomagnefic data to examine polarity "bias" as a function of time [McElhinny, 1971; Irving and Pullaiah, 1976]. Although not a direct measure of the reversal rate, periods of high polarity bias were attributed to long single polarity

No abstract

The magnetization of Pacific seamounts shows a surprisingly large variation when viewed as a function of time. Similar to normal upper oceanic crust, the basalts that form the external shell of volcanic seamounts have a high average magnetization immediately after formation. This high magnetization decays rapidly with time, resulting in a factor of 5 reduction in intensity in less than 500,000 years. Because submarine volcanoes appear to consist of the same basalts that form the upper part of the seafloor, low temperature oxidation of the magnetic mineral titanomagnetite is assumed to be responsible for the reduction in magnetization for both types of crust. Some large oceanic volcanoes are of sufficient volume that the upper capping rocks erupt subaerially and submerge completely due to thermal subsidence only after several million years. These capping rocks can be locally abundant (i.e., Hawaii), are exposed to a different post formation environment, and can have a magnetic mineralogy that is different from submarine volcanoes. In spite of these differences, the magnetization of subaerially erupted rocks of the Hawaii/Emperor Chain also may show a similar rapid reduction in intensity with age. Finally, seamounts of Cretaceous age show a factor of 2 higher average magnetization when compared to Tertiary seamounts, similar to the previously observed increased magnetization of Cretaceous seafloor basalts recovered from Deep Sea Drilling Project/Ocean Drilling Project cores. This increase in Cretaceous magnetization appears to be reflected in the increased amplitude of magnetic anomalies in some, but not all of the ocean basins, prior and subsequent to the Cretaceous Normal Superchron. The processes potentially responsible for the elevated Cretaceous magnetization values include a high geomagnetic field intensity, large-scale variations in whole rock major element geochemistry of the extrusive rocks, or a systematic acquisition of a secondary magnetization during the Cretaceous Normal Superchron. [Johnson and Merrill, 1973; Marshall and Cox, 1971a and . to insure internal consistency, we limited our study to the 1971b; Johnson and Atwater, 1977]. h• addition to the examination of seamounts from the Pacific basin. progressive decrease of magnetization as young oceanic crust ages, normal seafloor basalts also show a anomalous high Seamount Magnetization Variability Unlike normal oceanic seafloor, seamounts have a wide range of initial formation conditions and are exposed to a variety of post formation environments. For simplicity, we divided these 13,701 13,702 JOHNSON ET AL.: PACIFIC SEAMOUNT MAGNETIZATION volcanoes into two categories:(1) those that formed entirely beneath the sea surface, and (2) those that are of sufficient size that the f'mal formation stages occur subaerially. Owing to the difference in eruptive environments, the relevant initial magnetic minerals for these two classes of seamounts are different, and the post formation alteration history for the two types of rocks is also likely to be distinct....