An Insight into the Origin of Fluidal Obsidian Pyroclasts from a Basaltic Fissure Eruption, Ascension Island, South Atlantic

Basaltic eruptions are the most common manifestations of volcanism on Earth. Pyroclast textures can provide insight into the nature of fragmentation, storage and ascent of magma. Progressing our understanding of eruption dynamics by investigating textural, eld and geochemical data of pyroclasts will signicantly contribute to hazard mitigation models. Unusual, dense obsidian pyroclasts co-erupted with scoria from a small-volume monogenetic ssure eruption on Ascension Island, South Atlantic. The obsidian pyroclasts vary from spherical/sub-spherical glass beads (<1-5 mm in diameter), resembling Pele's tears and spheres, to larger spatter-like bombs (>40 cm in diameter) found within scoria ramparts built up either side of the ssure. They account for <<0.005 vol.% of the total ejected material. Fluidal morphologies imply that the clasts had unusually low melt viscosities on eruption and were therefore capable of owing upon impact. Field relationships indicate separate magmas underwent no mixing during ascent but agglutinated post-eruption to form an agglomerate of obsidian, scoria and lithic clasts. In-situ quantitative geochemical analyses reveals two distinct magmas; the obsidian is rhyolitic (74 wt.% SiO2), while the scoria is trachyandesite (54-58.8 wt.% SiO2). Magmatic compositions are distinctly bimodal, with clear separation between obsidian and scoria components exemplied in macrocryst species present; the obsidian contains anorthoclase, while the scoria contains a continuum between andesine to labradorite. It also implies that magma mixing did not occur within the crust/conduit. This is comparable in the wider Ascension Island magmatic suite where there is no evidence for magma mixing. The storage, fragmentation and genesis of the obsidian clasts remains unclear. One possibility is that a shallow pocket of degassed rhyolitic magma was intersected by ascending mac magma in a dike. However, this does not account for the low erupted volume of obsidian clasts and absence of mingling textures. Currently, the preferred hypothesis for their origin is assimilation of water-/gas-poor felsic country rock. Reheating of pre-erupted, felsic country rock to basaltic temperatures would exceed the glass-transition temperature, remobilising glassy clasts, thus invoking highly uidal textures. Similar pyroclasts elsewhere that co-erupted with basalt have examples of large rhyolitic inclusions which show no evidence of re-melting. Ascension lacks any texturally intermediate clast-type, which highlights the unusual nature of these pyroclasts.