xviii, 179 p. : ill. (some col.). Includes 3 video files. / Mafic arc volcanoes have eruption styles that range from explosive to effusive. In a broad sense, eruption style is controlled by the rate of magma supply to the vent. In this dissertation I examine relationships between eruption rate and style in two separate studies: (1) an investigation of ongoing activity at NW Rota-1, a submarine volcano in the Mariana arc, and (2) a morphologic study of the Collier Cone lava flow field in the Central Oregon Cascades. The eruptions of NW Rota-1 range from effusive to moderately explosive; eruptions are effusive when mass eruption rate (MER) is low and explosive when MER increases. The explosivity of submarine eruptions is suppressed by seawater because of increased hydrostatic pressure, rapid cooling, and the high viscosity of water relative to air (which limits expansion). The combination of seawater and relatively low MERs limit pyroclast deposition to within meters to tens of meters of the vent. In fact, many pyroclasts fall back into the vent and are recycled. Evidence for recycling includes microcrystalline inclusions within erupted pyroclasts and elevated Cl and Na concentrations in matrix glass. Enrichment of Cl and Na suggests that seawater assimilation provides a geochemical signature of recycling. Recycling is limited to low MER explosive eruptions and is not observed in either effusive lava or deposits from high MER explosions. Direct observations of eruptions allow measurements of eruption rate. However, it is more challenging to estimate MERs of eruptions that were not observed. To address this problem, I develop and test methods of constraining the eruption rate (and duration) of the c. 1600 year old Collier Cone lava flow using the flow morphology. To quantify flow morphology I combine field observations with GIS analysis of Lidar-derived digital topography. Channel dimensions constrain emplacement rates; dominant wavelengths and amplitudes of surface folds constrain spatial and temporal changes in flow rheology. Three videos of eruption activity accompany this dissertation as supplemental files. This dissertation includes both previously published and unpublished co-authored material. / Committee in charge: Dr. Katharine V. Cashman, Chair; Dr. Joshua J. Roering, Member; Dr. Paul J. Wallace, Member; Dr. Patricia F. McDowell, Outside Member; Dr. William W. Chadwick, Outside Member

Comprehension of eruptive histories is critical in understanding the evolution of magmatic systems at arc volcanoes and may supply evidence to the petrogenesis of intermediate and evolved magmas. Within the 300 ka eruptive history of Mount St. Helens, Washington, the Kalama Eruptive Period, 1479- ~1750 CE was bracketed by interludes of quiescence (Hoblitt et al., 1980) and thus likely represents an entire eruptive cycle within a span of 300 years. Study of the magmatic evolution during this short time period provides key information regarding inputs and the plumbing system of Mount St. Helens. This research aims to enhance comprehension of processes leading to the petrogenesis of intermediate magmas by providing whole rock and phase geochemical data of an eruptive cycle, thereby providing constraints on the magmatic evolution of the Kalama Eruptive Period. The eruptive sequence is divided into early, middle and late subperiods. The early Kalama began with two dacitic plinian eruptions and continued with smaller eruptions of dacite domes (64.4-66.5 wt% SiO₂) that included quenched mafic inclusions (53.7-57.7 wt% SiO₂). The middle Kalama signified the onset of basaltic andesite and andesite eruptions ranging between 55.5-58.5 wt % SiO₂. Subsequently, summit domes that began as felsic andesite (61-62.5 wt% SiO₂) and transitioned to dacite (62.5-64.6 wt% SiO₂) dominated the late Kalama. Previous work on Kalama-aged rocks suggests magma mixing is an integral process in their production. Compositions and textures of crystal phases, in addition to the presence of xenocrysts in middle and late Kalama rocks, confirm mechanical mixing of magmas likely produced many of the sampled compositions. New petrographic observations were integrated with new whole rock and phase EMP and LA-ICP-MS data and the known stratigraphy in order to constrain the magmatic and crustal components active during the Kalama Eruptive Period. New findings include: 1. Two populations of quenched mafic inclusions, one olivine-rich and one olivine-poor, are identified from the early Kalama based on mineralogy, textures, and major and trace element chemistry. Major element modeling shows crustal anatexis of plutonic inclusions found in early Kalama dacites could produce the felsic magma source of the olivine-poor population. The olivine-rich population incorporated cumulate material. 2. Four distinct lava populations erupted during the early part of the middle Kalama (X lavas), including two found exclusively in lahar deposits: M-type lahars are the most mafic, B-type lahars are more mixed, the Two Finger Flow was previously grouped with other middle Kalama-age lavas, and the X lava (in situ) has unique geochemical and textural character. X tephras likely correlate with the lavas. 3. There were at least three mafic source contributions at Mount St. Helens during the eruptive period: the parent to the X deposits, the cumulate material in the olivine-rich QMIs, and the calc-alkaline parent to the MKLV and SDO. The magma reservoir at Mount St. Helens has been modeled as a single, elongate chamber (Pallister et al., 1992). Multiple coeval basaltic or basaltic andesite parents fluxing into the magmatic system beneath the volcano could indicate a more complex magma chamber structure. / Graduation date: 2011