Secular variation in Mexico during the last 40,000 years

Abstract

Paleomagnetic data are available for Mexico from various archives, covering irregularly the last 40,000 years. Data obtained from lake sediments are affected by ambiguous conversion of age to depth, and the paleomagnetic data themselves are not of high quality, as samples were only demagnetized at one field intensity. These lake sediment data are therefore considered as less reliable for construction of a secular variation curve, although they may give some qualitative information for the last 25,000 years. Archeological samples provide data for the period, 100–1200 a.d. Their 14 C ages are some times not well constrained, with unrecoverable analytical and statistical procedures, and often ages were only inferred from the stratigraphic position and the paleomagnetic record. The paleomagnetic record of these materials is of high quality with very low internal dispersion, although no magnetic cleaning techniques have been applied. Variations of inclination and declination with time are hampered by large age uncertainties and thus of limited value for the construction of secular variation curves. Paleomagnetic data from an U–Th dated stalagmite provide reasonable directions for most of the last 1200 years, being of lower quality in those parts with very low remanence intensity. The U–Th age determinations were interpreted in terms of a constant growth-rate, which might be only valid as a first-order approximation. Recent volcanic rocks from the Trans-Mexican volcanic belt provide 19 paleomagnetic directions of reasonable to high quality, and often also absolute paleointensities. Ages have been determined from 14 C analysis of charcoal or paleosols found below the lavas, and extend over the last ∼40,000 years. Therefore, volcanic rocks provide the most adequate archive for the construction of a secular variation curve, but much more volcanic structures need to be dated to obtain a semi-continuous secular variation curve. Comparison of the data from different archives is possible mainly for the last 2500 years and despite the aforementioned problems similar variations of declination and inclination are observed. This holds as well when comparing them to data from Holocene volcanic rocks from western North America. © 2002 Elsevier Science B.V. All rights reserved.