Testing paleointensity determinations on recent lava flows and scorias from Miyakejima, Japan

Still no consensus has been reached on paleointensity method. Even the classical Thellier method has not been fully tested on recent lava flows with known geomagnetic field intensity based on a systematic sampling scheme. In this study, Thellier method was applied for 1983, 1962 and 1940 basaltic lava flows and scorias from Miyakejima, Japan. Several vertical lava sections and quenched scorias, which are quite variable in magnetic mineralogy and grain size, provide an unparalleled opportunity to test paleointensity methods. Thellier experiments were conducted on a completely automated three-component spinner magnetometer with thermal demagnetizer 'tspin'. Specimens were heated in air, applied laboratory field was 45 microT, and pTRM checks were performed at every two heating steps. Curie points and hysteresis properties were obtained on small fragments removed from cylindrical specimens. For lava flows sigmoidal curves were commonly observed on the Arai diagrams. Especially the interior part of lava flows always revealed sigmoidal patterns and sometimes resulted in erroneously blurred behaviors. The directions after zero-field heating were not necessarily stable in the course of the Thellier experiments. It was very difficult, for the interior part, to ascertain linear segments on Arai diagrams corresponding to the geomagnetic field intensity at the eruption. Upper and lower clinker samples also generally revealed sigmoidal or upward concave curves on Arai diagrams. Neither lower nor higher temperature portions of the sigmoids or concaves gave the expected geomagnetic field intensities. However, there were two exceptional cases of lava flows giving correct field intensities: upper clinkers with relatively low unblocking temperatures (< 400 deg.C) and lower clinkers with broad unblocking temperature ranges from room temperature to 600 deg.C. A most promising target for paleointensity experiments within the volcanic rocks is scoria. Scoria samples always carry single Curie temperatures higher than 500 deg.C, and the ratios of saturation remanence to saturation magnetization (Mr/Ms) of about 0.5 are indicative of truly single-domain low-titanium titanomagnetite. Unambiguous straight lines were always observed on Arai diagrams covering broad temperature ranges like the lower clinker samples, and the gradients gave the expected field values within a few percent errors. Thellier experiments applied for the recent lava flows did not successfully recover the expected field intensity from most samples. No linear segment was recognized or incorrect paleointensity values were obtained from short segments with limited temperature ranges. In Thellier or other types of paleointensity experiments laboratory alteration is checked in details, but if a sample once passed the alteration check, the TRM/NRM ratios of any limited temperature or field ranges were accepted as reflecting paleointensity. Previously published paleointensity data from lava flows should include much of such dubious data. Generally lava flows are not suitable for paleointensity determinations in light of its large grain-size and mixed magnetic mineralogy, except for scoria and clinker.