This content will become publicly available on January 29, 2027

Title: Viscosity and Flow Dynamics of the Rhyolitic Lava of Mount Arteni, Armenia: Insights from Numerical Modeling of Emplacement and Effusion Rates
The Arteni Volcanic Complex lies within the Ararat Plain, within theAragats Volcanic Province (AVP) of western Armenia, and is noted for itsremarkable diversity of obsidian types. K-Ar dates show EarlyPleistocene ages of 1.26 to 1.60 Ma, while fission track dating yieldsslightly younger ages of 1.17 to 1.38 Ma. The Arteni Volcanic Complexwas formed through successive eruptions from two adjacent centers: MetsArteni and Pokr Arteni, both dome-shaped rhyolitic volcanoes. Thecomplex consists of rhyolitic domes, dome-coulees, necks, fissures,extrusions, and dikes—typical of high-viscosity silicic volcanism. The Arteni rhyolitic lava flow extends a remarkable 8 km from the summitarea of Mets Arteni. Covering an area of ~15 km², itsvolume is estimated at 0.6 km³. The unit consists of a basal layer ofpyroclastics and crystalline rhyolitic-dacitic rocks, followed bycrystal-clear obsidian, and capped by vesicular water-rich perlite. Theobsidian typically contains between 0.06 and 0.2 wt% H₂O, while theoverlying perlite contains up to 4 wt% H₂O. Although rhyolitic melts are usually highly viscous, the Arteni flow’sexcessive length suggests lower viscosity due to temperature,composition and water content during emplacement. Consequently, wemodeled viscosity using 850-970 °C obtained from similar rheologicalcalculations for the Dry Fountain obsidian complex located 22 km northof Yerevan, Armenia. We note that the upper perlite layers of theAragats flow contain 3.5–4 wt% H₂O. If the underlying obsidian formedby rapid degassing under lithostatic pressure applied by the overlyingperlite, the original melt likely had a water content of~3.8 wt%. For this water content and temperature, theviscosity of Arteni rhyolite was 3.86–4.79 log η (Pa·s), much lowerthan anhydrous obsidian (7.45–9.18 log η Pa·s). We suggest that thislow viscosity led to the exceptional length of the rhyolite flow due todepolymerization of melt structure as well as lack of phenocrysts.  more » « less
Award ID(s):
2211680
PAR ID:
10684942
Author(s) / Creator(s):
; ; ; ;
Publisher / Repository:
AGU
Date Published:
Subject(s) / Keyword(s):
Sponsoring Org:
National Science Foundation
More Like this
  1. Abstract The continental crust is produced by the solidification of aluminosilicate‐rich magmas which are sourced from deep below the surface. Migration of the magma depends on the density (ρ) contrast to source rocks and the melt viscosity (η). At the surface, these silica‐rich melts are typically sluggish due to highη > 1,000 Pa s. Yet at their source regions, the melt properties are complexly influenced by pressure (P), temperature (T), and water contents (). In this study, we examined the combinedP‐T‐ effects on the behavior of melts with an albite stoichiometry (NaAlSi3O8). We usedfirst‐principlesmolecular dynamics simulations to examine anhydrous (0 wt % H2O) and hydrous (5 wt % H2O) melts. To constrain thePandTeffects, we exploredP ≤ 25 GPa across several isotherms between 2500 and 4000 K. The melts show anomalousP‐ρrelationships at lowP ∼ 0 GPa and highT ≥ 2500 K, consistent with vaporization. At lithospheric conditions, meltρincreases with compression and is well described by a finite‐strain formalism. Water lowers the melt density (ρhydrous < ρanhydrous) but increases the compressibility, that is, 1/Khydrous>1/KanhydrousorKhydrous < Kanhydrous. We also find that the meltηdecreases with pressure and then increases with further compression. Water decreases the viscosity (ηhydrous < ηanhydrous) by depolymerizing the melt structure. The ionic self‐diffusivities are increased by the presence of water. The decreasedρandηby H2O increase the mobility of magma at crustal conditions, which could explain the rapid eruption and migration timescales for rhyolitic magmas as observed in the Chaitén volcano in Chile. 
    more » « less
  2. Abstract Viscosity of silicate melts governs magma transport and influences mantle dynamics, yet effects of pressure and water on melt viscosity remain poorly understood. Here, we report in situ falling‐sphere viscosity measurements on diopside (Di) melts with 0–3 wt.% H2O along the liquidus up to 7 GPa and 2103 K using synchrotron X‐ray radiography. By incorporating our hydrous melt data into a previously validated model for the dry system, the effects of pressure, temperature, and H2O contents on Di melt viscosity can be satisfactorily captured by the function: whereT*is the homologous temperature,xH2Ois the molar % H2O,η0 = 8.90 (1.50) × 10−8 Pa s,b0 = 3.02 (0.10), andH*(P) = 15.72 (0.03)−0.35 (0.01)·P + 1.07 (0.07) × 10−2·P2−1.19 (0.14) × 10−4 P3, ×10−3 GPa−1. Adding 3 wt.% H2O systematically reduces viscosity by ∼0.7 log units. For both dry and hydrous melts, viscosity along the liquidus decreases monotonically with increasing pressure, suggesting that moderate hydration may not significantly alter the compressional behavior of Di melts. Combining the Di viscosity model with models for feldspar and olivine, we simulated the viscosity of analog basaltic magmas under mantle conditions. Increasing H2O content from 0 to 3wt.% raises mobility of basaltic magma increases by >1 order of magnitude. In hot plume settings, the mobility further increases by a factor of 30 relative to typical ambient mantle. Assuming a simple percolation model, the increased mobility corresponds to faster melt ascent in mantle plumes that could, in part, explain the voluminous magmatism of large igneous provinces. 
    more » « less
  3. Abstract Two distinct types of rare crystal-rich mafic enclaves have been identified in the rhyolite lava flow from the 2011–12 Cordón Caulle eruption (Southern Andean Volcanic Zone, SVZ). The majority of mafic enclaves are coarsely crystalline with interlocking olivine-clinopyroxene-plagioclase textures and irregular shaped vesicles filling the crystal framework. These enclaves are interpreted as pieces of crystal-rich magma mush underlying a crystal-poor rhyolitic magma body that has fed recent silicic eruptions at Cordón Caulle. A second type of porphyritic enclaves, with restricted mineral chemistry and spherical vesicles, represents small-volume injections into the rhyolite magma. Both types of enclaves are basaltic end-members (up to 9.3 wt% MgO and 50–53 wt% SiO 2 ) in comparison to enclaves erupted globally. The Cordón Caulle enclaves also have one of the largest compositional gaps on record between the basaltic enclaves and the rhyolite host at 17 wt% SiO 2 . Interstitial melt in the coarsely-crystalline enclaves is compositionally identical to their rhyolitic host, suggesting that the crystal-poor rhyolite magma was derived directly from the underlying basaltic magma mush through efficient melt extraction. We suggest the 2011–12 rhyolitic eruption was generated from a primitive basaltic crystal-rich mush that short-circuited the typical full range of magmatic differentiation in a single step. 
    more » « less
  4. The molar abundance ratio of H2 Om /OH of rhyolitic porous pyroclasts, obsidian pyroclasts, and flow obsidians from the 1060CE Glass Mountain eruption at Medicine Lake Volcano (USA) were obtained by Diffuse Reflectance Infrared Fourier Transform spectroscopy. Samples were also analyzed for their total water content, [H2 Ot ], and isotopic composition, 𝜔D, via Temperature Conversion Elemental Analysis (TC/EA). Porous clasts exhibit a H2 Om /OH ratio of 0–4.63 (and a [H2 Ot ] measured by Giachetti et al. (2020) of 0.34–1.2 wt%) that is positively correlated with their porosity and [H2 Ot ], contrary to the low [H2 Ot ] and low H2 Om /OH ratios of both obsidian pyroclasts (0.32–0.72 wt.%, 0.14–0.63) and flow obsidians (0.04–0.07 wt.%, 0.24–0.28). These results confirm the interpretations of Giachetti et al. (2020) that porous clasts were rehydrated for about 1000 years after the eruption via diffusion of overwhelmingly molecular, meteoric water in the matrix-glass, whereas obsidian pyroclasts and flow obsidians are essentially dense and thus rehydrate poorly. Analyses were also conducted on two size fractions of flow obsidian (∱ 63 εm and 63–250 εm) that were ground either (1) immediately before analysis or, (2) about nine years prior to analysis and kept in closed containers ever since. Results show that the ∱ 63 εm ground samples gained 0.06–0.24 wt% of meteoric water in just nine years, with a H2 Om /OH ratio increasing from 0.32–0.54 when measured immediately after crushing, to 3.35–4.64 when measured nine years later. For these smaller particles, even 24 h of heating at 130 ⋛ C under vacuum before analysis is insufficient to remove all the water gained by rehydration. We thus recommend the use of coarser powders (∲ 63 εm) and longer pre-analysis heating time under vacuum (∲2 days) for more reliable [H2 Ot ], H2 Om /OH, and 𝜔D measurements on obsidian samples. Given the thinness of the glass in between vesicles in porous pyroclasts and thus their ability to quickly rehydrate, total water content obtained via analysis of the bulk material (e.g., by TC/EA, Loss on Ignition, Karl Fischer titration) must be interpreted together with 𝜔D and/or H2 Om /OH data to evaluate the extent of rehydration, even for relatively young samples. 
    more » « less
  5. Rhyolitic melt that fuels explosive eruptions often originates in the upper crust via extraction from crystal-rich sources, implying an evolutionary link between volcanism and residual plutonism. However, the time scales over which these systems evolve are mainly understood through erupted deposits, limiting confirmation of this connection. Exhumed plutons that preserve a record of high-silica melt segregation provide a critical subvolcanic perspective on rhyolite generation, permitting comparison between time scales of long-term assembly and transient melt extraction events. Here, U-Pb zircon petrochronology and 40 Ar/ 39 Ar thermochronology constrain silicic melt segregation and residual cumulate formation in a ~7 to 6 Ma, shallow (3 to 7 km depth) Andean pluton. Thermo-petrological simulations linked to a zircon saturation model map spatiotemporal melt flux distributions. Our findings suggest that ~50 km 3 of rhyolitic melt was extracted in ~130 ka, transient pluton assembly that indicates the thermal viability of advanced magma differentiation in the upper crust. 
    more » « less