No abstract

Crystals within volcanic rocks record geochemical and textural signatures during magmatic evolution before eruption. Clues to this magmatic history can be examined using crystal size distribution (CSD) studies. The analysis of CSDs is a standard petrological tool, but laborious due to manual hand-drawing of crystal margins. The trainable Weka segmentation (TWS) plugin in ImageJ is a promising alternative. It uses machine learning and image segmentation to classify an image. We recorded back-scattered electron (BSE) images of three volcanic samples with different crystallinity (35, 50 and ≥85 vol. %), using scanning electron microscopes (SEM) of variable image resolutions, which we then tested using TWS. Crystal measurements obtained from the automatically segmented images are compared with those of the manual segmentation. Samples up to 50 vol. % crystallinity are successfully segmented using TWS. Segmentation at significantly higher crystallinities fails, as crystal boundaries cannot be distinguished. Accuracy performance tests for the TWS classifiers yield high F-scores (>0.930), hence, TWS is a successful and fast computing tool for outlining crystals from BSE images of glassy rocks. Finally, reliable CSD’s can be derived using a low-cost desktop SEM, paving the way for a wide range of research to take advantage of this new petrological method.

To assess whether magma ascent rates control the style of volcanic eruption, we have studied the petrography, geochemistry and size distribution of microlites of plagioclase and pyroxene from historical eruptions from Tongariro, Ruapehu and Ngauruhoe volcanoes located in the southern Taupo Volcanic Zone, New Zealand. The studied deposits represent glassy andesitic and dacitic tephra shards from the Mangamate, Mangatawai, Tufa Trig and the Ngauruhoe tephra formations, ranging in age from 11,000 years BP to 1996 AD. Covering a range in eruption styles and sizes from Strombolian to Plinian, these samples provide an excellent opportunity to explore fundamental volcanic processes such as pre-eruptive magma ascent processes. Our quantitative petrographic analysis shows that larger microlites (> 30 µm) display complex growth zoning, and only the smallest crystals (< 30 µm) have formed during magma ascent in the conduit. Using a combination of orthopyroxene geothermometry, plagioclase hygrometry, and MELTS modelling, we show that these microlites nucleated at maximum pressures of 550 MPa (c. 16.5 km) from hot andesitic magmas (1010-1130 ˚C) with low H2O content (0-1.5 wt%). Size distributions of a total of > 60,000 microlites, involving 22 tephras and 99 glass shards, yield concave-up curves, and the slopes of the pyroxene microlite size distributions, in combination with well-constrained orthopyroxene crystal growth rates from one studied tephra, indicate microlite population growth times of ∼3 ± 1 days, irrespective of eruption style. These data imply that microlites form in response to cooling of melts ascending at velocities of < 5 cm s-1 prior to H2O exsolution, which only occurs at < 33 MPa. Maximum magma ascent rates in the upper conduit, calculated using the exsolution of water during final decompression, range between 3 to 12 m s-1, i.e. at least an order of magnitude lower than the hypersonic vent velocities typical of Vulcanian or sub-Plinian eruptions (up to 400 m s-1). This implies that magma ascent from depths of an average of 4 km occurs in dykes, and that vent velocities at the surface are controlled by a reduction of conduit cross-section towards the surface (e.g. dyke transiting to cylindrical conduit).

Low elevation flank eruptions represent highly hazardous events due to their location near, or in, communities. Their potentially high effusion rates can feed fast moving lava flows that enter populated areas with little time for warning or evacuation, as was the case at Nyiragongo in 1977. The January–March 1974 eruption on the western flank of Mount Etna, Italy, was a low elevation effusive event, but with low effusion rates. It consisted of two eruptive phases, separated by 23 days of quiescence, and produced two lava flow fields. We describe the different properties of the two lava flow fields through structural and morphological analyses using UAV-based photogrammetry, plus textural and rheological analyses of samples. Phase I produced lower density (∼2,210 kg m−3) and crystallinity (∼37%) lavas at higher eruption temperatures (∼1,080°C), forming thinner (2–3 m) flow units with less-well-developed channels than Phase II. Although Phase II involved an identical source magma, it had higher densities (∼2,425 kg m−3) and crystallinities (∼40%), and lower eruption temperatures (∼1,030°C), forming thicker (5 m) flow units with well-formed channels. These contrasting properties were associated with distinct rheologies, Phase I lavas having lower viscosities (∼103 Pa s) than Phase II (∼105 Pa s). Effusion rates were higher during Phase I (≥5 m3/s), but the episodic, short-lived nature of each lava flow emplacement event meant that flows were volume-limited and short (≤1.5 km). Phase II effusion rates were lower (≤4 m3/s), but sustained effusion led to flow units that could still extend 1.3 km, although volume limits resulted from levee failure and flow avulsion to form new channels high in the lava flow system. We present a petrologically-based model whereby a similar magma fed both phases, but slower ascent during Phase II may have led to greater degrees of degassing resulting in higher cooling-induced densities and crystallinities, as well as lower temperatures. We thus define a low effusion rate end-member scenario for low elevation effusive events, revealing that such events are not necessarily of high effusion rate and velocity, as in the catastrophic event scenarios of Etna 1669 or Kilauea 2018.