Apparent Eruptive Response of Cascades and Alaska–Aleutian Arc Volcanoes to Major Deglaciations

Abstract:

Precise argon ages of Pleistocene eruptive products from Cascades and Alaska–Aleutian arc volcanoes cluster in time following major deglaciations. Compilation of edifice-volume-weighted dates for over 700 lavas from 16 volcanoes are compared to marine oxygen isotope stages (MIS 2–8) of Bassinot et al. (1994, EPSL, v. 126, p. 91-108) and interpreted temperatures from the Vostok ice core (Petit et al., 1999, Nature, v. 399, p. 429-436). To assess relative time-volume relationships we weight the distribution of ages measured at each volcano by its total edifice volume. The abundance of ages scales with the number of mapped eruptive units, and may differ substantially from the true eruptive output. The distribution is also weighted inversely by the number of dates to account for centers with more or fewer dates. Stacked probability density functions yield significant peaks after MIS 6 and MIS 8. Veniaminof, Emmons Lake, Westdahl, Redoubt (Alaska–Aleutian arc), and Adams and Crater Lake (Cascades arc) have apparent eruptive episodes 135-110 ka (early MIS 5), coinciding with rapid warming of the oceans following the MIS 6 glacial. Veniaminof began growing at 250 ka (end MIS 8) and erupted more than 200 km³ of lava in MIS 7. Emmons Lake, Adams, Rainier, and Glacier Peak also have apparent growth peaks (abundant dated units) following MIS 8. Apparent correlation of eruptive episodes with deglaciations may result from depressurization of magmatic systems due to ice retreat resulting in enhanced decompression melting and/or diminished compressive stress on crustal magma reservoirs, poor preservation of lava sequences during glacial maxima, or coincidence. Next steps in this study include (1) more rigorous assessment of eruptive volumes of dated map units, (2) refining ice volume estimates during MIS 2, 6, and 8 at various centers by dating ice marginal lava flows and tuyas and by mapping moraines at selected volcanoes, (3) re-analyzing sequences previously dated by K/Ar to improve precision, and (4) extending the compilation to deglaciation following MIS 2.