Conduit processes in crystal-rich dacitic magma and implications for eruptive cycles at Guagua Pichincha volcano, Ecuador

Abstract

Stratovolcanoes are commonly characterised by cyclic eruptive activity marked by transitions between dome-forming, Vulcanian, Subplinian and Plinian eruptions. Guagua Pichincha volcano (Ecuador) has been a location of such cyclicity for the past ~ 2000 years, with Plinian eruptions in the first and tenth centuries AD (Anno Domini/after Christ), and CE (Common Era) 1660, which were separated by dome-forming to Subplinian eruptions, such as the recent 1999–2001 eruption. These cycles are therefore a prominent example of effusive-explosive transitions at varying timescales. Here, we investigate the reasons for such shifts in activity by focusing on degassing and outgassing processes within the conduit. We have coupled a petrophysical and textural analysis of dacites from the CE 1660 Plinian eruption and the 1999–2001 dome-forming/Vulcanian eruption, with different percolation models in order to better understand the role of degassing on eruptive style. We demonstrate that the transition from dome-forming to Plinian activity is correlated with differences in phenocryst content and consequently in bulk viscosity. A lower initial phenocryst content and viscosity is inferred for the Plinian case, which promotes faster ascent, closed-system degassing, fragmentation and explosive activity. In contrast, dome-forming phases are promoted by a higher magma viscosity due to higher phenocryst content, with slower ascent enhancing gas escape and microlite crystallization, decreasing explosivity and yielding effusive activity. Resumen Los estratovolcanes se caracterizan comúnmente por presentar actividad eruptiva cíclica, marcada por transiciones entre erupciones formadoras de domos y erupciones de tipo Vulcanianas, Subplinianas y Plinianas. El volcán Guagua Pichincha (Ecuador) ha dado lugar a tal ciclicidad durante los últimos ~ 2000 años, con erupciones Plinianas tanto en los siglos Primero y Décimo, como en el año 1660, las cuales estuvieron intercaladas por erupciones formadoras de domos y de tipo Subplinianas, tal como ocurrió durante la erupción reciente de 1999–2001. Estos ciclos son, por lo tanto, ejemplos destacados de transiciones eruptivas de tipo efusiva-explosiva a escalas de tiempo variadas. En este trabajo, investigamos las razones de tales cambios de actividad enfocándonos en procesos de exsolución y pérdida de gases del magma en el conducto (desgasificación en sistemas cerrado y abierto). Hemos acoplado análisis petrofísicos y texturales tanto de dacitas de la erupción Pliniana de 1660, como de la erupción formadora de domos/Vulcaniana de 1999–2001, junto con diferentes modelos de percolación, para así comprender mejor el rol de la exsolución de volátiles en el estilo eruptivo. Demostramos que la transición desde una actividad efusiva formadora de domos a una Pliniana está correlacionada con diferencias en el contenido de fenocristales y, subsecuentemente, con la viscosidad total del magma. Un contenido inicial menor de fenocristales y una menor viscosidad se infiere para el caso Pliniano, lo que promueve un ascenso más rápido, desgasificación en sistema cerrado, fragmentación y finalmente actividad explosiva. Por el contrario, las fases formadoras de domos son promovidas por una viscosidad mayor debido a un contenido mayor de fenocristales, con ascenso más lento promoviendo a su vez el escape de gases y la cristalización de microlitos, disminuyendo la explosividad y produciendo actividad efusiva.

Pre‐Eruptive Outgassing and Pressurization, and Post‐Fragmentation Bubble Nucleation, Recorded by Vesicles in Breadcrust Bombs From Vulcanian Activity at Guagua Pichincha Volcano, Ecuador

“…(d) Groundmass porosity versus radial distance to the crust. Note that the connected vesicle networks identified in this study straddle the range of percolation threshold (from the percolation model of Giachetti et al, 2019) and connected porosities in the bomb interiors, both corrected for phenocryst content, from previous studies (Colombier et al, 2022;Wright et al, 2007). 2023) recently presented additional textural and compositional evidence for such cyclic pressurization of the shallow conduit.…”

“…These values straddle or are in excess of the range of percolation thresholds (the transition from an isolated/coalesced vesicle network to system spanning connectivity and onset of permeability, hereafter noted Φ c ) in the groundmass that have been discussed for these breadcrust bombs by Colombier et al. (2022) based on the model of Giachetti et al. (2019) (Figure 6d).…”

“…Porosity evolution with radial distance shows that connected areas are mostly found at a groundmass porosity >0.5 and up to ∼0.8 (except one sample at lower values due to vesicle stretching in this area) (Figure 6d). These values straddle or are in excess of the range of percolation thresholds (the transition from an isolated/coalesced vesicle network to system spanning connectivity and onset of permeability, hereafter noted Φ c ) in the groundmass that have been discussed for these breadcrust bombs by Colombier et al (2022) based on the model of Giachetti et al (2019) (Figure 6d). In addition, groundmass porosity values in the bomb interiors fall just above this porosity threshold (Wright et al, 2007).…”

“…We note that most of the VND data for the breadcrust bombs exceed values from the literature for other silicic eruptions (see Jutzeler et al, 2016 for a compilation). Colombier et al (2022) also reported high values for VND at 10 8 mm −3 for pumice from Plinian activity at Guagua Pichincha. These high values can be better explained by the methodology used in our study and in Colombier et al (2022) in which we focus on localized values of the VND and not average values across heterogenous samples.…”

“…Colombier et al (2022) also reported high values for VND at 10 8 mm −3 for pumice from Plinian activity at Guagua Pichincha. These high values can be better explained by the methodology used in our study and in Colombier et al (2022) in which we focus on localized values of the VND and not average values across heterogenous samples. However, we emphasize that one distribution with four magnifications and four different areas of the sample at the highest magnification also yielded high vesicle number densities of 10 7 mm −3 .…”