Alisha Coote (supervised by Phil Shane), University of Auckland (EQC funded project 16/U733)
Young basalt volcanoes in the Kaikohe-Bay of Islands area (KBI), Northland, are New Zealand’s ‘forgotten’ volcanoes having received little scientific investigation despite new eruptions being capable of impacting regional agriculture, tourist and aviation industries. They are considered to be monogenetic volcanoes, formed in one episode, and subsequent activity occurring in a new location. The most recent eruptions in the KBI field occurred 43,000 years ago, indicating the field is dormant but not extinct. The KBI basalts contain large (up to 10 mm), late-forming crystals of plagioclase indicating a period of pre-eruption shallow subterranean magma storage. This study aimed to use the geochemical record contained in plagioclase, clinopyroxene and olivine crystals in lava to quantify magma storage duration and causes of eruption, and to investigate whether the magmas temporarily stall in the crust prior to erupting. This would increase the likelihood of detecting precursor phenomena in future eruptions.
Based on geochemical and isotopic zonation, most the plagioclase and clinopyroxene crystals in the basalts are a diverse cargo. Sr-isotope ratios demonstrate various mantle and crust sources for the plagioclase crystals. Clinopyroxene-melt equilibrium demonstrates periods of crystallisation at depths of about 15 km and >27 km. Many of the crystals comprise a relic core formed in a basaltic forerunner magma or from country rock, surrounded by a rim zone that grew from new ascending magma. We found that ascending magmas interact or mix with earlier magmas residing in the crust, and some magmas fail to erupt. This is a remarkable contrast to the traditional models for small-volume intra-plate basalt volcanoes (e.g. Auckland volcanic field) that assume rapid formation and ascent of magma to eruption without significant crust interaction. We envisage the magma system to be a mush column through the crust where magmas stall at various depths. This petrological model fits well with geophysical data that suggests a low-velocity zone of partial melt occurs beneath the volcanoes at 10-19 km, and the Moho occurs at >27 km.
Studies highlight that during future activity, periods of temporary magma storage and crystallisation could occur at rheology barriers in the crust, thus prolonging the interval of eruption precursor activity, and increasing the potential for detecting seismic or degassing precursor events. Future magma eruptions and failed eruptions (intrusions) at the Kaikohe-Bay of Islands volcanic field could be associated with seismicity at 10-19 km and ~28 km, where the major crustal rheology contrasts occur. The likelihood of intrusion unaccompanied by eruption may also be greater than previously expected for such volcanoes.
Late Quaternary basalts erupted in the Kaikohe-Bay of Islands area provide an opportunity to explore the ascent history of small volume magmas in an intra-continental monogenetic volcano field, and hence, improve our understanding of potential future precursor phenomena. To achieve this goal, we investigated the formation and growth history of phenocrysts (crystals) in the basalts. The plagioclase phenocrysts represent a diverse crystal cargo. Most of the crystals have a rim growth that is in equilibrium with the host basalt rock. The resorbed cores of the crystals variously formed in more differentiated or more primitive melts. The relic cores have 87Sr/86Sr ratios that are either mantle-like (~ 0.7030) or crustal-like (~ 0.7040 to 0.7060), indicating some are antecrysts formed in melts fractionated from plutonic basaltic forerunners, while others are true xenocrysts from greywacke basement and/or Miocene arc volcanics.
Clinopyroxene phenocrysts also record magma mixing and crystal entrainment in the crust. Like the accompanying plagioclase, many have a rim overgrowth which is in equilibrium with the host rock, but have a resorbed core that crystallised in a more silicic magma. These crystals record mafic recharge, presumably the trigger to eruption. Crystal-melt equilibria indicate that the clinopyroxene formed at a narrow temperature range (1095-1200 ºC), but wide pressure range (150-870 MPa). Most formed at 300-600 MPa (~11-23 km depth), and a subordinate population formed at 700-900 MPa (>27 km depth). These depths coincide with major seismic velocity contrasts at a zone of partial melt (10-19 km) and the Moho (~28 km) inferred from geophysical data. Thus, buoyancy or rheology contrasts in the crust temporarily slow magma ascent and promote periods of melt crystallisation and the assimilation of antecrystic and xenocrystic components.
It is envisaged that intrusive basaltic forerunners produced a zone where various degrees of crustal assimilation and fractional crystallization occurred. The erupted basalts represent subsequent mafic recharge of this system. This crystallization history contrasts with traditional concepts of low-flux basaltic systems where rapid ascent from the mantle is inferred. From a societal perspective, staged magma storage and crystallisation beneath some basalt intra-plate fields increases the likelihood of detecting pre-eruption geophysical phenomena that could act as signals to pending eruptions.