MAGMA SOURCES, MIXING, AND SULFUR DEGASSING DURING THE 2018 KĪLAUEA FISSURE ERUPTION INFERRED FROM MINERAL AND MELT INCLUSION GEOCHEMISTRY

On April 30, 2018, the collapse of Kīlauea’s Pu῾u ῾Ō῾ō vent preceded downrift magma propagation and then fissure eruption in the Lower East Rift Zone (LERZ). Furthermore, magma withdrawal from summit reservoir(s) caused the Halema’uma’u lava lake to drain and the summit caldera to partially collapse. The 2018 rift eruption was Kīlauea’s largest in >150 years and resulted in profound impacts to southeastern Hawai’i.

After two weeks of eruption, lavas from LERZ fissures transitioned from viscous, evolved basalt to higher temperature and more fluid lavas resembling Pu῾u ῾Ō῾ō and summit magma. By May 28, LERZ activity focused at fissure 8, building a 55 m tephra cone and feeding a high effusion rate lava channel. Vigorous activity continued until abruptly ceasing in early August 2018. The high effusion rates were accompanied by prodigious degassing, with SO₂ emission rates among the highest measured worldwide.

We collected rapidly quenched tephra at different stages of the eruption and conducted electron microprobe analyses on minerals, melt inclusions, and matrix glasses to investigate mineral compositions and melt volatile contents. Mixed proportions and compositions of olivine, clinopyroxene, and plagioclase reflect magmas sourced from shallow and deep summit reservoirs that hybridized with stored ERZ magmas during transit. Olivine crystals include very primitive summit-sourced grains (Fo_85–90) and differentiated ERZ-sourced grains (Fo_75–80). Many olivine display disequilibrium dissolution textures and compositional zoning due to mixing of these magma types.

Olivine-hosted melt inclusions contain from 400–1300 ppm sulfur, indicating variable degassing histories prior to inclusion entrapment. In quenched golden pumice from fissure 8, melt embayments contain 700–900 ppm S, whereas matrix glass contains 200–500 ppm S. The difference between S in melt embayments and matrix glass indicates that up to 75% of the magmatic S was degassed during extrusion at fissure 8, while at least 20–25% remained in the lava and degassed during surface flow. The temporal variation in erupted melt and crystal compositions is further compared to variations in SO₂ emission rate data. Finally, we calculate a total inferred sulfur budget required to explain the exceptionally high SO₂ emission rates measured during this eruption.