Title: INVESTIGATION OF A LARGE-VOLUME, MAFIC IGNIMBRITE AT LLAIMA VOLCANO, CHILE: HARNESSING THE POWER OF PETROLOGICAL SAMPLING ACROSS DIVERSE SCALES

Abstract: In this dissertation, I investigated the magma compositions and conditions within the magma system that generated the unusual, highly explosive mafic Curacautín eruption of Llaima volcano, Chile. The Curacautin ignimbrite (Ci) is 3.5-4.5 km3 DRE basaltic to basaltic andesite ignimbrite. A new geographically and stratigraphically extensive sampling study allowed us to improve the overall geochemical characterization of the Ci, evaluate petrologic processes that generated the Ci magma, and construct a conceptual model of the pre-eruptive magmatic system and an eruptive framework for the Ci. Using the new and expansive geochemical whole rock compositions, I modeled binary mixing between the two ends of the whole rock range in trace element ratio space. To determine the source of these endmember compositions, I modeled the trace elements associated with flux mantle melting systematics. I characterized the trace element composition of the slab-derived component enriching the ambient mantle. I determined that the observed endmember trace element ratio compositions can be explained by a high extent of melting (25%) producing the lower trace element ratio end member and a lower f (10%) melt producing the higher Nb/Yb endmember. Using olivine compositions coupled with melt inclusion trace element and volatile compositions, I determined that the higher Nb/Yb magma composition was stored shallowly, cooled, and experienced a more significant degree of crystallization. The lower Nb/Yb ratio magma originated hotter and deeper in the crust and moved to a shallower storage region long enough to trap the shallower degassed magma composition but quickly erupted before pre-existing melt inclusions could begin re-equilibrating (less than a few hours). I attempted to correlate outcrops widely distributed around the volcano chemically. While there is geochemical variation stratigraphically and geographically, the correlation of outcrops remains difficult. I expanded on the geochemical investigation by assessing the distribution of whole rock pyroclast compositions in several key sampling horizons. I determined that the two endmember magmas were initially erupting from separate conduits in the east and west. As the eruption progressed, the conduits widened until they eventually collapsed into each other, resulting in the mingling of the two compositions observed as primarily deposited in the north. These findings have significant implications for the interpretation of explosive eruption deposits. This research shows the importance of a geographically and stratigraphically expansive geochemical investigation because focusing on an outcrop from the east, north, or west would have led to different interpretations of the magmatic and eruptive framework.

Advisor: Dorsey Wanless

Co-Advisor: Brittany Brand

Committee Members: Mark Schmitz, Ben Andrews, and Monica Hubbard