Cyclostratigraphic and magnetostratigraphic constraints on the duration of the CAMP

Olsen, P E, polsen@ldeo.columbia.edu, Lamont-Doherty Earth Observatory, Rt 9W, Palisades, NY 10964 United States, Kent, D V, dvk@rci.rutgers.edu, Lamont-Doherty Earth Observatory, Rt 9W, Palisades, NY 10964 United States & Department of Geological Sciences, Rutgers University, Piscataway, NJ 08854-8066 United States.

The duration of what may be the largest known flood basalt, the Central Atlantic Magmatic Province (CAMP), has major implications for the magmatic processes intimately related to one of the largest known continental rift provinces formed just antecedent to the break-up of central Pangea. While the reported scatter in radiometric dates from the CAMP has been decreasing in recent years (e.g. Marzoli, 1999; Hames et al., 2000), there is still a broad range spanning tens of millions of years, centered on about 200 Ma. The cited high precision of many of these dates might lead one to conclude that the scatter has some relevance to the duration of the CAMP, but independent means are needed to assess their accuracy, as opposed to analytical uncertainties. One independent test is provided by the cyclostratigraphy of lacustrine strata surrounding and interbedded with the CAMP in eastern North America This cyclostratigraphy extends from the Newark basin of the northeastern US, where it was first worked out (Olsen et al. 1996), to Nova Scotia and Morocco and constrains the outcropping extrusive event to less than 600 ky in duration starting roughly 20 ky after the Triassic-Jurassic boundary, and to within one pollen and spore zone and one vertebrate biochron. However, 3/4 of the geographical extent of CAMP still has no cyclostratigraphic control. Another independent means of assessing the duration of the CAMP is provided by its paleomagnetic character. Normal paleomagnetic polarity is a virtual signature of the CAMP, and this information combined with the Newark astronomically calibrated geomagnetic polarity time scale (Kent and Olsen, 1999; Olsen and Kent, 1999), new additional constraints from the Hartford basin, and the published magnetostratigraphy for the Paris Basin (Yang et al., 1996) allows a probabilistic assessment of the reality of the published scatter in CAMP dates. Using the age probability function for all CAMP rocks presented by Marzoli et al. (1999) as a theoretical sampling function for the paleomagnetic record, it is clear that the expected frequency of normal polarity for the Late Triassic and early part of the Jurassic should be about 0.5, or 50 percent. It is not until this sampling function is compressed to 5 my or less that we can reach the observed greater than 95 percent normal polarity, for this probability distribution. Based on this probabilistic argument, it is likely that the reported overall scatter in CAMP ages, which can occur within cyclostratigraphically constrained short intervals of time, reflects distortions of the actual crystallization ages by as yet poorly documented geological processes. This would imply that the cited estimates of analytical precision are often over an order of magnitude better than the accuracy of the dates. Based on our analysis, the existing paleomagnetic data from both intrusive and extrusive rocks suggest emplacement of nearly all the CAMP within an interval of less than 3 my of nearly entirely normal polarity, most likely during (Newark basin astronomically tuned GPTS) chron E23n and/or E24n. The very few examples of reversed magnetizations tantalizingly suggest that some CAMP activity may have occurred just prior to the Triassic-Jurassic boundary in chron E23r.

Hames, W.E. et al. 2000. Geology 28:859-862; Kent, D.V. \& Olsen, P.E. 1999. J. Geophys. Res. 104:12,831-12,841; Marzoli, A. et al., 1999. Science 284:616-618; Olsen, P.E. et al., 1996. Mus. North. Ariz. Bull. 60;11-22; Olsen, P.E. and Kent, D.V., 1999. Phil. Trans. Roy. Soc. Lond. (A), 357;1761-1787; Yang, Z. et al., 1996. J. Geophys. Res., 101:8025-8042.