From flow to furnace: Low viscosity of three-phase lavas measured at Kīlauea 2018 eruption conditions

Melt composition, temperature, and crystallinity are often seen as the three most important characteristics driving lava rheology, which controls eruptive behavior. Traditional methods of measuring the viscosity of crystallizing basalts often yield different mineral characteristics to natural samples and are typically bubble-free. To quantify the viscosity of basalts inclusive of bubble and crystal cargo, we developed a new technique to measure high-temperature three-phase isothermal lava viscosity and applied it to samples from the 2018 eruption of Kīlauea. This new experimental technique begins at subliquidus temperatures, preserving original phenocrysts. A short experimental duration allows for the retention of most of the original bubble population (19%−31% vs. 36% in the original lava) and accurate replication of crystal textures from field samples, as documented in quenched postexperiment samples. The observed rheological behavior in these experiments, conducted at syneruptive temperatures (1150−1105 °C) and strain rates (0.4−18 s⁻¹), should therefore be representative of the lava flows. We measured average viscosities of 116 Pa·s at 1150 °C to 167 Pa·s at 1115 °C, i.e., only 10%−25% higher than calculated liquid viscosities at those temperatures, and a maximum of 1800 Pa·s at 1105 °C. These results are much lower than viscosity measured in traditional bubble-free experiments, which plateaued at ∼14,000 Pa·s at 1115 °C. Our results suggest the effect of bubbles in three-phase magmas may be greater than predicted by models based on two-phase bubbly liquids, and this effect must be included in realistic lava flow rheology models. The method proposed here supplies a framework for providing the necessary experimental constraints.