GEOPHYSICAL RESEARCH LETTERS, VOL. 23, NO. 5, PAGES 475-478, MARCH 1, 1996 Paleomagnetic field intensity recorded in from the East Pacific Rise, the last 69 KA V. Mejia, N.D. Opdyke, and M.R. Perfit Department of Geology, University of Florida, Gainesville, Florida submarinebasaltic glass Abstract. Paleointensity determinations were obtained from Goldstein et al., 1994)the ages estimated based on distance basaltic glass recovered from the northern East Pacific Rise from the axis may be as much as 36 Ka too old. using the Thelliers' (1959) method as applied by Pick and Tauxe (1993a). The samples are up to a69 Ka old based on their Experimental procedure distance from the ridge axis and U-series disequilibria The Thelliers' experiment was carried out within a shielded measurements. The results show a general increase in room, using a cryogenic magnetometer and a solenoid tha paleointensity of the field from 69 Ka to the present with values produces an internal field of 40 mT. Samples of aphyric to only ranging from 50% higher than the present field to as low as 26% slightly phyric basaltic glass weighing < 1 g were embedded in of the present dipole field. The average dipole moment is 8.32 sea salt cylindrical pellets (f = 2.5 cm) to allow accurate x 10 22 Am 2 which is similar to the present axial dipole repositioning in the solenoid and in the magnetomet moment. The results are in general agreement with other Thermal demagnetization was carried out in preheated oven and absolute and relative paleointensity records. lasted 35 minutes for temperatures < 250 ø and 40 minutes for Introduction Pre-Holocene absolute paleointensity records obtained from lava flows are rare, therefore the variations in paleointensity of the Earth's magnetic field for pre-Holocene times are very temperatures > 250 ø. The field-off cooling steps were performed by removing the samples from the oven and placing them in the confines of a mu-metal tube and cooling them using a stream o air provided by a fan. Cooling times were on the order of 15 minutes. poorly known. Holocene paleointensity records have been Data evaluation obtained from ceramics and pottery, but archeological materials are unavailable from pre-Holocene times. Submarine basaltic The statistical treatment of paleointensity data was applie glass (SBG) has been proposed by Pick and Tauxe (1993a, b) as as defined by Coe et al. (1978). Unsuccessful results are usually an excellent material for paleointensity studies notonly due to magnetochemical changes taking place during heating because of the high quality data obtained from this material but especially if they occur during early heating steps so that th because of its wide occurrence in the world ocean basins in both available data points of the Arai plot (ARM-TRM diagram) are time and space. insufficient to yield reliablepaleointensity data. The In this study, a record of paleointensity of the Earth's paleointensity data presented here passed the following magnetic field was obtained from basaltic glass collected using stringent acceptance criteria: (a) linearity of Arai plot'(b) the submersible ALVIN mainly from the East Pacific Rise (EPR) PTRM checks were considered to be successful if the value of the at 9.5 ø - 10øN (Perfit et al., 1994). Several samples were also measured TRM and the remeasured TRM (pTRM check) were studied from the eastern Galapagos rift and the southern Juan de within 5% of the mean value and/or if the remeasured TRM and Fuca ridge (Smith et al., 1994). The Thellier and Thellier (1959) the best fit line-derived TRM agreed within +_5%. This avoids method as applied by Pick and Tauxe (1993a) was used in thiscomparing the remeasured TRM with a single measurement study in a slightly modified form. which involves several possible errors (Thellier, 1959); (c) q factor (Coe et al., 1978) > 5; (d) a single trend of the magnet Age of the samples vector toward the origin in the Orthogonal plot; and (e) k>300 Seven of the samples have been dated using the 235U-231pa and MAD<15 ø (Kirschvink, 1980). disequilibrium method (Goldstein et al., 1993) which provides Restfits model crustal residence ages (magma chamber plus eruption ages). Although the precision of these dates is very high based Paleointensity on a recently erupted sample known age only on analytical uncertainties, the errors in eruption ages can be as great as 8 Ka because of uncertainties of the initial The measurement of paleointensity on a basaltic glass 231pa/235U ratios and in 231pa half-life. All 231pa ages are recovered from very young-looking pillow mounds in the axis concordant with 230Th ages to within the estimated errors of of the Southern Juan de Fuca Ridge (Smith et al., 1992) was each method (Goldstein et al., 1993). The ages of the samples carried out (figure 1). This sample, ALV 2431-1, is believed to are also inferred from the spreading rate (5.5 cm/yr) and the have been erupted sometime between 1981 and 1987 based on distance from the ridge axis. However because of the occurrence repeat SeaBeam Surveys, camera tows and observations from of off-axis volcanism along the EPR (Perfit et al., 1994 and ALVIN (Chadwick et al., 1991 and Chadwick and Embley, 1994). The paleointensity result was 5.6% greater than the Copyright 1996 by the American Geophysical Union. IGRF value for the site at the sea surface for 1985. This is considered good agreement since the value of the field at the sea Paper number 96GL00018 floor is probably greater than at the sea surface. Similarly, Pick 0094-8534/96/96GL-00018503.00 and Tauxe (1993a), have determined the intensity of lavas 475