Paleointensity experiments using alternating field demagnetization

Experiments for absolute paleointensity involve a large number of heatings and thus increase the probability of mineralogical changes which considerab ly reduce the success rate. Magnetic granulometry also plays an important r ole. In the presence of multidomain grains thermal demagnetization of natur al remanent magnetization (NRM) is not linearly related to acquisition of t hermal remanence (TRM) in the laboratory, so that the NRM-TRM diagrams are characterized by two slopes, neither of them providing a correct field dete rmination. After briefly reviewing most current techniques for paleointensi ty and various attempts to improve the results, the present paper investiga tes approaches relying on alternating field (a.f.) demagnetization. Measure ments have been performed on at least eight samples from seven contemporary flows from Hawaii. Classical thermal experiments were also conducted with poor success. In contrast, 30% of the NRM-TRM curves Obtained after a.f. de magnetization were perfectly linear over the entire spectrum of coercivitie s and thus appropriate to determine a paleointensity. This difference was c aused by the possibility of measuring a much larger number of samples. The mean field is less than 10% higher than the present field and in better agr eement for a subset of samples with NRM and TRM heated at 300 degrees C pri or to a.f. demagnetization, These values are acceptable within the uncertai nties inherent in any study of absolute paleointensity. Thus a.f. technique s can be used with confidence provided that the whole spectrum of coercivit ies is involved in the field determination and exempt from any correction. This condition is necessary and sufficient to demonstrate the absence of al teration after heating. Because of the rapidity of such experiments, an app roach combining a.f. and thermal experiments is certainly the best way to i ncrease considerably the number of field determinations per lava flow and h ence the overall quality of paleointensity studies. (C) 2000 Elsevier Scien ce B.V. All rights reserved.