Congratulations to Meredith Calogero who defended her dissertation on Thursday, January  14, 2021

Advisor:s Eric Hetland & Rebecca Lange

Abstract:

 High-silica rhyolites (>75 wt% SiO2) are the most differentiated magmas on Earth. Though rare in subduction volcanism, it erupts in large volumes (100–1000s km3) in extensional settings. Understanding their generation is crucial as rhyolite formation substantially reconstitutes the continental crust, and their eruptions have hazardous global implications. Two supervolcano quantity eruptions of high-silica rhyolite have occurred in the Western U.S. within 1 Ma, and this thesis focuses on Long Valley caldera, where ~800 km3 rhyolite has erupted in the last 2.2 Ma at various eruption rates, including the catastrophic caldera-forming >600 km3 Bishop Tuff eruption of zoned rhyolite in under a week. The second-most eruptive magma type in LV is basalt. Numerous numerical modeling efforts in the literature have addressed the formation of silicic magmas in response to basalt influx. This study differs in that we present a high-resolution numerical model both spatially (~1 m) and temporally (<day), tracking crustal-scale and local sill-scale processes in response to basalt influx in the form of random sill emplacements.

 
          In Chapter 2, the crustal-scale response of basaltic sill emplacements into thick (32 km) granitoid crust is observed, and we identify the importance of emplacement rate and geometry (i.e. initial depth interval for sill emplacements) for substantial heating. We assume a finite amount of H2O in each basaltic sill and recognize the timescale for volatile release at ~75% crystallization. At the sill-scale we observe the formation of transient wall-rock partial melts. The timescales of both processes are dependent on local ambient temperature, and broadly overlap, within 10–100s of years after sill emplacement. We hypothesize a mechanism for secondary melt migration, wherein the ubiquitous scattered aplite dikes found pre-existing throughout the Sierra Nevada batholith, which forms the basement underneath Long Valley, become the site of melt-filled fractures. 


          In Chapter 3, the consequence of secondary transport of rhyolitic melts, composed of partial melt contributions from wall-rock granitoid and previously solidified mafic sills and interstitial rhyolite from crystallizing basalt, is explored. Melts are allowed to ascend if two criteria are met in an updated version of the numerical model: 80% crystallization of the sill center, and enough cumulative partial melt (i.e. thickness) within the wall rock adjacent to the basaltic sill to be captured by model resolution. Additionally, we add compositional tracking of Sr to address a paradox observed between the eruptive LV rhyolites and basalts. The basalts are rich in Sr (1000–2000 ppm), whereas the eruptive rhyolites are extremely depleted (<25 ppm). 


          At the emplacement conditions presented (50 m/5 kyr into initial 20–30 km), results depict anticipated warming of the crust around ascended rhyolite sills, and a cooling of the deep crust, both ~60ºC in magnitude. For 3 km of basalt emplaced, ~1.7 km of rhyolite was generated, with ~30% contribution from basaltic sources. This extent of rhyolite generation is possible because of H2O contribution from the crystallizing basalt, and neither fluid-absent wall-rock granitoid, nor basaltic sources alone could produce such extensive rhyolite melt. We observe a drop in Sr concentration in the secondary rhyolites (~380 ppm) which is notably elevated relative to eruptive LV rhyolites, highlighting the necessity of multiple stages of crustal reworking, further exemplified through simplified calculations for Sr and Nd partitioning. Significant progress is made towards developing a predominantly sub-solidus multi-stage process for differentiated rhyolite generation.