Episodic crystallization in young explosive eruptions in Dominica, Lesser Antilles, revealed by U-Th dating of zircons

Abstract

In the field, pumice clasts were taken from four unconsolidated ignimbrite deposits and a welded tuff in the Roseau Valley (Fig. DR1). The sample location distribution was controlled by exposures, often limited to road cuts, but spans the length of the Roseau Valley. The most distal sample was from a 20 m thick deposit in the city's Botanical Gardens (King's Hill), not previously documented, ~8 km from the summit of Micotrin. The basal St. Aromont sample is from a 50 m thick pyroclastic sequence previously described by Howe et al. (2014) as part of the Link Flow sequence, ~8 km from Micotrin. The Casso sample was located on the southern ridge flank of the Roseau River, 6 km from Micotrin at an elevation of 250 m. The Casso deposit is a ~15 m thick pyroclastic deposit, dominated by ash and has not previously been studied. The Trafalgar sample was from a 2 m boulder of welded tuff at the base of Trafalgar Falls, a ~200 m thick sequence, which is the most proximal deposit to the proposed vent, ~2 km from Micotrin.

Evidence for an Active, Transcrustal Magma System in the Last 60 ka and Eruptive Degassing Budget (H2O, CO2, S, F, Cl, Br): The Case of Dominica

“…More recently, Smith et al (2013), Howe et al (2015Howe et al ( , 2014, and Boudon et al (2017) proposed a stratigraphy of the explosive eruptions that occurred in the central part of Dominica. Howe et al (2014) and Frey et al (2018) dated ignimbrites by U-Th/He and U-Th zircon ages, respectively, whereas Boudon et al (2017) proposed a detailed chronostratigraphy based on a series of 14 C dates. The storage conditions of the magma at the origin of the large ignimbrites from the central part of Dominica have been determined on the basis of volatile contents preserved in melt inclusions (Balcone-Boissard et al, 2018) and through phase-equilibrium experiments (Solaro et al, 2019).…”

Zircon and Melt Extraction From a Long‐Lived and Vertically Extensive Magma System Underneath Ilopango Caldera (El Salvador)

“…The nature of multiple peaks in zircon age spectra remains controversial, and explanations range from magma assembly from different magma batches to remobilization of solidified parts of the magma reservoir where zircon crystallized at different times (Bacon & Lowenstern, 2005; Charlier et al., 2005; Miller & Wooden, 2004; Reid & Vazquez, 2017; Schmitt et al., 2003, 2010; Storm et al., 2011). Multiple zircon crystallization peaks could also be an artifact of the low number of analyses (<30 analyses; Frey et al., 2018), and confident identification of distinct zircon growth episodes would require the identification of age gaps, either from truncation of internal growth zones due to intermittent resorption or from sharp age increments in depth profiling analyses (Frey et al., 2018; Reid & Vazquez, 2017; Storm et al., 2012). Although the total number of crystals analyzed for individual TBJ units in this study was 38 and 72 for rim and core ages, respectively, the number of analyses in the <100 ka range is smaller, and therefore apparent modes within the <100 ka interval may lack statistical significance.…”

“…Multiple zircon crystallization peaks could also be an artifact of the low number of analyses (<30 analyses; Frey et al, 2018), and confident identification of distinct zircon growth episodes would require the identification of age gaps, either from truncation of internal growth zones due to intermittent resorption or from sharp age increments in depth profiling analyses (Frey et al, 2018;Reid & Vazquez, 2017;Storm et al, 2012).…”

Supplemental Material: High-precision geochronology requires that ultrafast mantle-derived magmatic fluxes built the transcrustal Bear Valley Intrusive Suite, Sierra Nevada, California, USA

“…Each of these samples contain two populations of three to nine grains that have weighted mean ages offset by gaps of 150-340 kyrs. Although multi-modal zircon populations are occasionally observed in young erupted systems (Barboni et al, 2016;Kern et al, 2016;Frey et al, 2018), these types of distributions are quite rare in CA-ID-TIMS studies of intrusive systems. However, we believe that the observed distributions are robust: given the tight clustering of ages within each population, these distributions would be exceedingly difficult to produce through random sampling of a continuous distribution and would similarly require remarkable coincidences for the older population in each sample to be composed of grains with inherited cores that all have the identical core to rim volume ratios and initial uranium contents necessary to generate overlapping ages within uncertainty.…”