Introduction of the LTD-DHT Shaw method and its application to historical and older volcanic rocks

Since the 1970s, Thellier and Shaw methods have been most commonly used to determine absolute paleointensity of the geomagnetic field. The principles behind these methods are the same in using thermal remanent magnetization but very different in demagnetization techniques and reliability checks. For ideal samples (e.g. assemblage of non-interactive single-domain particles without alteration due to laboratory heating), the Thellier method is regarded as valid and robust. However, recent studies have revealed that natural rock samples are not always ideal material for the Thellier method: i.e. some historical lava flows give systematically high paleointensities (e.g. Yamamoto et al. 2003; Mochizuki et al., 2004). Though some improvements have been proposed, methods other than the Thellier-type are useful for crosschecking the reliability of paleointensity. We have worked on improvement of the Shaw-type method originally proposed by Shaw (1974). At present, the most advanced version is the LTD-DHT Shaw method (Tsunakawa and Shaw, 1994; Yamamoto et al., 2003): it utilizes the anhysteretic remanent magnetization (ARM) correction, the double heating test (DHT) for checking the ARM correction applicability, low temperature demagnetization (LTD) and thermal remanence acquired by heating in vacuum. During the past few years, we have successfully applied the LTD-DHT Shaw method to several historical lava flows in Japan and Hawaii. In recent years, the application has been extended to older volcanic rocks in order to obtain datasets independent of the Thellier type. From the preliminary analysis of the new dataset, mean virtual dipole moments (VDMs) for the last five million years are calculated to be 3.70 [10 22 Am2] (N = 34) for stable geomagnetic filed and 1.17 [10 22 Am2] (N = 9) for transitional field. These values are lower than those obtained from the Thellier data set.