Using Eruption Source Parameters and High-Resolution Grain-Size Distributions of the 7.7 ka Cleetwood Eruption of Mount Mazama (Oregon, United States) to Reveal Primary and Secondary Eruptive Processes

Numerical simulations of real-time volcanic ash dispersal forecasts and ensuing tephra hazard assessments rely on field-derived Eruption Source Parameters (ESPs) such as plume height, erupted volume, mass eruption rate and the Total Grain-Size Distribution (TGSD) of particles ejected from a volcano into the atmosphere. Here we calculate ESPs for the similar to 7.7 ka Cleetwood eruption of Mount Mazama (Crater Lake/giiwas, Oregon, United States) that immediately preceded the caldera-forming eruption. We also introduce a novel approach to produce high-resolution grain-size distributions (GSDs) of individual samples over a wide range of particle sizes (0.00035-35 mm) by combining laser diffraction with dynamic image analysis. Detailed field analysis allows us to divide the Cleetwood eruptive sequence into a series of three distinct and consecutive VEI 4 eruptions: Phase 1 (similar to 0.26 km(3)), Phase 2 (similar to 0.98 km(3)) and Phase 3 (similar to 0.20 km(3)). Phase 2 was the most intense with a plume height of similar to 19 km and an average mass discharge rate of similar to 3.1 x 10(7) kg s(-1). Its TGSD yields a fractal dimension D similar to 3.1, like other similar eruptions. All twelve high-resolution GSDs produced in this study exhibit two systematic breaks in slope from a power-law relationship at similar to 0.125 and similar to 0.510 mm. These breaks in slope create three segments: S1 (<0.125 mm), S2 (0.125-0.510 mm), and S3 (>0.510 mm) that can be fit by power-law relationships with fractal dimensions of D1 = 2.5 +/- 0.2, D2 = 0.5 +/- 0.1, and D3 = 3.6 +/- 1.1, respectively. Together with ESPs and detailed componentry, D values at various locations give insight into magma fragmentation and tephra transport. We find that D1 values are positively correlated with the median grain-size and are similar to values found in rapid decompression magma fragmentation experiments. We infer that D1 values reflect the size distribution of the primary products of magma fragmentation and could thus be used to infer the potential energy at fragmentation. We interpret the relatively low values of D2 to an increase in dense components due to particle rafting. Our work shows that comparing high-resolution GSDs at several locations on the dispersal axis can further constrain primary and secondary eruptive processes, which prove crucial to improving tephra hazard assessments and dispersal forecasting.