The origins of large volume, compositionally-zoned volcanic eruptions - new constraints from U-series isotopes and numerical thermal modeling for the 1912 Katmai-Novarupta eruption
Abstract
A stratigraphically constrained set of samples from the 1912 Katmai-Novarupta eruption in Alaska have compositions that range from a nearly primitive basalt (at 50.4 wt. % SiO2 and 11.4 wt. % MgO), through basaltic andesite, andesite and dacite to rhyolite with 77.7 wt. % SiO2. The entire major and trace element range can be well modelled by 80-90% closed-system crystal fractionation over a temperature interval from 1279 to 719 C at 100 MPa. Given an erupted rhyolite volume of 7.2 km3 the implied volume of parental basalt is ~ 60 km3. Numerical thermal models suggest that, for wall rock temperatures appropriate to this depth, this volume of magma would cool and crystallise 90% over this temperature interval within a few 10’s kyr. However, 87Sr/86Sr ranges from 0.70339 to 0.70374, (230Th/238U) from 0.97 to 1.05 and (226Ra/230Th) from 2.52 to 1, clearly requiring open-system processes. The effects of crystal fractionation combined with assimilation of the Naknek Formation host sediments can replicate the range of Sr isotopes but the range in U-Th-Ra isotopes is much harder to account for using such models. The variety of (226Ra/230Th) ratios in samples from the basalt to andesite compositional range require that these magmas were generated less than several 1000 years before eruption and had varying initial Ra excesses. Crustal residence times for some dacites are close to 8000 years old whereas the rhyolites appear to be significantly older (50-350 kyr). Thus, the magmas that erupted within 60 hours in 1912 had a wide range of crustal residence times. Nevertheless, because they were emplaced in the same, shallow-level thermal regime, they all evolved along similar liquid lines of descent from parental magmas with similar overall compositions. This lead to a continuum in bulk composition and temperature in the pre-eruption magma chamber system. We conclude that this system was built progressively over time with multiple mafic inputs providing both mass and heat some of which lead to thawing of an older, pre-existing silicic pluton that provided much of the magma mass for the rhyolite. Although this sort of scenario is increasingly envisaged as a common one, it cannot not be deduced from field observations or major and trace element compositions. Only the short-lived isotope data are sensitive to processes such as protracted magma build-up, differentiation, assimilation and eruption.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2009
- Bibcode:
- 2009AGUFM.V42A..01T
- Keywords:
-
- 1000 GEOCHEMISTRY