Paleointensity in ignimbrites and other volcaniclastic flows

Ash flow tuffs (ignimbrites) are common worldwide, frequently contain fine-grained magnetite hosted in the glassy matrix, and often have high-quality ⁴⁰Ar/³⁹Ar ages. This makes them attractive candidates for paleointensity studies, potentially allowing for a substantial increase in the number of well-dated paleointensity estimates. However, the timing and nature of remanence acquisition in ignimbrites are not sufficiently understood to allow confident interpretation of paleointensity data from ash flows. The remanence acquisition may be a complex function of mineralogy and thermal history. Emplacement conditions and post-emplacement processes vary considerably between and within tuffs and may potentially affect the ability to recover ancient field intensity information. To better understand the relevant magnetic recording assemblage(s) and remanence acquisition processes we have collected samples from two well-documented historical ignimbrites, the 1980 ash flows at Mt. St. Helens (MSH), Washington, and the 1912 flows from Mt. Katmai in the Valley of Ten Thousand Smokes (VTTS), Alaska. Data from these relatively small, poorly- to non-welded historical flows are compared to the more extensive and more densely welded 0.76 Ma Bishop Tuff. This sample set enables us to better understand the geologic processes that destroy or preserve paleointensity information so that samples from ancient tuffs may be selected with care. Thellier-type paleointensity experiments carried out on pumice blocks sampled from the MSH flows resulted in a paleointensity of 55.8 μT +/- 0.8 (1 standard error). This compares favorably with the actual value of 56.0 μT. Excluded specimens of poor technical quality were dominantly from sites that were either emplaced at low temperature (<350°C) or were subject to post-emplacement hydrothermal alteration. The VTTS experienced much more wide-spread low-temperature hydrothermal activity than did MSH. Pumice-bearing ash matrix samples from this locality are characterized by at least two magnetic phases, one of which appears to carry a chemical remanent magnetization. Paleointensities derived from the second phase give results that vary widely but which may be correlated with degree of hydrothermal alteration or hydration. Preliminary data from the Bishop Tuff suggests that vapor-phase alteration at high (>600°C) temperatures does not corrupt the paleointensity signal, and additional data will be presented which explores this more fully.