Massive volcanic domes on Venus and the mobilisation of crystal mush: insights from the Troodos Ophiolite, Cyprus

Abstract

Venus is widely regarded as Earth’s ‘sister’ planet, given similarities in size, mass, density, chemical composition, and also their proximity. However, there are also striking differences between the two planets. The surface of Venus is dominated by volcanic and tectonised volcanic terrains. Volcanism on Venus is, presumably, largely plume-related due to the absence of evidence for plate tectonic processes. Under the extreme high temperatures and pressures of the Venusian surface, lava flows can extend for hundreds to thousands of kilometres. Steep-sided domes are among the most starkly discernible volcanic landforms on the surface of Venus, given their gigantic size (e.g., 0.12–1.73 km in height, and 8.3–61.8 km in diameter) and peculiar shape, compared to other volcanic features. Those domes are believed to be volcanic edifices, characterised by steep margins, and relatively smooth, flat upper surfaces, exhibiting a circular shape in plan view. Given their distinct morphology, they are important in (1) understanding the range of magmatic processes operating on Venus, and (2) elucidating geological evolution of planets lacking plate tectonics.

Despite their significance, there remains debate regarding the formation mechanisms of these domes. Physical models suggest that Venusian steep-sided domes require eruption of highly viscous liquids to explain their morphologies. This has led several authors to suggest that they represent eruption of SiO2-rich magmas at relatively low temperatures. However, other authors have argued that radar characteristics of steep-sided domes are inconsistent with eruption of very SiO₂-rich materials, and alternatively suggested that they represent eruption of more mafic liquids. In this scenario, high crystallinities may account for unusually high viscosities evidenced by dome morphology. Bulk viscosity of magmas depends on many factors, including composition (e.g., SiO₂ content and extent of polymerisation), volatile content, crystal content, and most importantly temperature. The overall aim of this study is to assess physical models of Venusian steep-sided dome formation by assessing the validity and implications of constraints on lava viscosity which they provide.

Work is divided into two Sections and four chapters. In Section 1 (Chapter 2), I model magma fractionation to calculate the full range of liquid compositions and bulk viscosity of magmas, which is compared to the viscosity thresholds from physical models. A key finding from this work is that high crystal contents are required to account for formation of high viscosity lavas on Venus. In Section 2 (Chapter 3 to 5), I conduct fieldwork to characterise a terrestrial analogue of eruption of high crystallinity picritic lavas in the Troodos Ophiolite, Cyprus. Using geochemical data, petrological and thermodynamic modelling, and field observations, I investigate the formation and eruption mechanisms, parental magma compositions, and viscosities of picrites. Information from this section is then used to further assess steep-sided dome formation on Venus and controls of crystallinity on magma viscosity.

In Chapter 2, I implement thermodynamic modelling using the rhyolite-MELTS model to constrain magma compositions, and magma viscosities based on various viscosity parameterisations, using bulk compositions inferred from Venera 13 (alkaline basalt) and Venera 14 (low alkali basalt) Soviet lander data. Viscosities are then compared to viscosity thresholds from published physical models of Venusian dome formation. Extensive (>85–90%) fractional crystallisation or equilibrium melting processes alone fail to produce magmas with viscosities required to account for steep-sided domes. The presence of H₂O during equilibrium crystallisation substantially modifies magma composition (e.g., SiO₂), and decreases solidus temperature, contributing to an increase of liquid viscosity. However, this effect is less significant than the direct control of H₂O as a network modifier in lowering viscosity. Instead, our results reveal that crystal contents of >60 vol.% are invariably required to produce sufficiently high magma bulk viscosities. Such high crystallinities probably require eruption of crystal-rich magma followed by surface crystallisation, possibly enhanced by degassing and undercooling.

To further constrain the eruption mechanisms of high crystallinity magma, in Chapter 3, I conduct fieldwork to observe field relations and collect a suite of representative high crystallinity picritic bodies (e.g., lava flows and a dome) from the Margi (Mαρκί, sometimes referred to as Marki) region, Troodos Ophiolite, Cyprus. These picrites represent a series of mafic/ultramafic lava flows, with groundmasses ranging from glassy to holocrystalline/vitrophyric texture and typically containing variable, but very high concentrations of up to cm-sized olivine. The crystal content from different picrite bodies can vary from 36 to 66 vol.%. I describe field relations, petrology, and geochemistry of a number of these picrite bodies, including olivine (+spinel and melt inclusion) compositions, erupted glass compositions and whole rock data. These picrites are discrete units within the upper pillow lava sequence, with crystal contents that imply open-system crystal accumulation. Variations in composition and olivine crystal cargo imply that bodies have discrete compositions, and close proximity of picrites to extensional faults supports a model where extrusion of crystal mush onto the palaeo-seafloor was facilitated by rifting and tapping of small magmatic systems. However, olivine-spinel equilibration temperatures imply minimal re-equilibration, favouring a model of hot crystal mush storage in the crust.

In Chapter 4, based on olivine-hosted melt inclusion (MI) data, I use petrological modelling to constrain parental magma compositions, and petrogenesis of high crystallinity lavas. I also perform forward fractional crystallisation and partial melting using rhyolite-MELTS to constrain conditions under which olivine crystallisation occurred. Results are discussed within the framework of glass (e.g., erupted liquids), olivine and spinel compositional data. The key findings are (1) crystal-rich magmas are likely formed by magma recharging and repetitive fractionation and concentration of olivine crystals within magmatic systems at less than 0.4 GPa (i.e., ~15 km depth); (2) at least 40% of olivine phenocrysts within picrites are in equilibrium with liquids in which they have erupted, and that picrites are likely formed both by fractionation of olivine, and remobilisation of xenocrysts; (3) Erupted picrites represent a mixture of variably evolving melts (closed system fractionation) and olivine from a mush-rich magmatic system, with the average Fo content from different picritic bodies ranging between Fo₈₉ to Fo₉₁; (4) Parental magmas to both erupted liquids and olivines had variable compositions, including both variations in primary compositions and due to variable extents of sulfide saturation during early stages of olivine crystallisation.

In Chapter 5, I perform viscosity calculations based on glass compositions, H2O content, crystallinity, and crystal size distribution or morphology measured directly from picrite samples to estimate the viscosity of picritic flows and lavas in Margi. Although these lavas have very high crystal contents, calculated viscosities are relatively low due to the low viscosity of erupted liquids, which is, in turn, a function of mafic compositions, high water contents and high erupted temperatures. As a result, the picrite dome at Margi is not a good analogue for investigating the eruption mechanisms for Venusian domes, although picritic bodies at Margi do provide insight into the formation and eruptibility of high crystallinity magma. I further compile dome morphologies for all terrestrial lava domes from basaltic to rhyolitic compositions and other extra-terrestrial lava domes. Venusian steep-sided domes have a smaller aspect ratio (height/width) than 90% of terrestrial lava domes, indicating that direct comparison between terrestrial domes and Venusian steep-sided domes is challenging given (1) their large difference in volumes and (2) the difference in surface environments between the two planets (e.g., surface temperature, pressure, atmosphere compositions, etc). As such, greater emphasis should instead be placed on improving models of dome formation under Venusian conditions, and especially, on considering the key role of crystallinity, lava cooling rate as well as composition in controlling lava viscosity and dome morphology.