THE IMPACTS OF LULLS AND PEAKS IN ERUPTION RATE ON LAVA FLOW PROPAGATION

Variable eruption rates during lava flow emplacement can complicate lava flow predictability. Recent studies have used wax experiments with pulsatory source flow rates to investigate flow inflation, breakouts, and tube formation (e.g., Rader et al. (2017), EPSL, 477; Peters et al. (2022), JVGR, 432). Using polyethylene glycol (PEG) analog wax experiments, we consider the role of a lull and a peak in eruption rate (Q) on lava flow propagation. We address the influence of unsteady Q on three emplacement modes common to lava flows: breakouts (on the surface and at flow margins), inflation, and lava tubes. We conducted 30 experiments using a programmable pump to inject dyed PEG wax into a chilled bath (~ 0° C) in a tank with a roughened base at a slope of 0°. The experiments were divided into two conditions: lull (a 50-second decrease in Q) and peak (a 50-second increase in Q). For each experimental run, an initial 150 cm³ of wax was erupted at a Q varying between 1–6 cm³/s. After the initial injection of wax, there was a 50-second period during which Q either halved (lull) or doubled (peak). After 50-seconds, an additional 150 cm³ of wax was erupted at the initial 1–6 cm³/s. Experiments were conducted across five morphology regimes corresponding with faster or slower cooling rates. Our results indicate the following: (1) during lulls in Q if the crust is strong, then breakouts will be limited and localized on the flow surface and at the margin, respectively. Inflation is likely and tubes are possible. Flow expansion may not accelerate when Q increases after the lull. (2) If there is a lull and the crust is weak, then breakouts will be limited at the surface but widespread at the margin. Inflation and tubes are unlikely and flow expansion will be coupled with Q at the vent (except at very low Q). (3) If there is a peak in Q and the crust is strong, surface and marginal breakouts will be widespread. Inflation is likely, but tubes are not. Flow expansion will correlate with Q at the vent. (4) If there is a peak and the crust is weak, surface breakouts are unlikely and/or limited and marginal breakouts are widespread. Inflation is very unlikely. Tubes do not form. Simple sheet flows are possible and flow expansion is coupled to Q at the vent. These results have implications for process-based statistical modeling and reconstructing historical eruption conditions.