The March--April 1986 eruption of Augustine Volcano, Alaska, provided an opportunity to directly measure the flux of gas, aerosol, and ash particles during explosive eruption. Most previous direct measurements of volcanic emission rates are on plumes from fuming volcanoes or on very small eruption clouds. Direct measurements during explosive activity are needed to understand the scale relationships between passive degassing or small eruption plumes and highly explosive events. Conditions on April 3, 1986 were ideal: high winds, clear visibility, moderate activity. Three measurements were made: (1) an airborne correlation spectrometer (Cospec) provided mass flux rates of SO2; (2) treated filter samples chemically characteized the plume and (3) a quartz crystal microcascade impactor provided particle size distribution. Atmospheric conditions on April 3 caused the development of a lee wave plume, which allowed us to constrain a model of plume dispersion leading to a forecast map of concentrations of SO2 at greater distances from the vent.

On April 3, 1986, the emission rate of SO2 at Augustine was 24,000 t/d, one of the largest direct volcanic rate measurements yet recorded with a Cospec. The results, coupled with analytical results from samples simultaneously collected on filters, allow us to estimate HCl emissions at 10,000 t/d and ash eruption rate at 1.5¿106 t/d. Based on other data, this ash eruption rate is about 1/50 of the maximum rate during the March--April activity. Filter samples show that the gas:aerosol proportions for sulfur and chlorine are about 10:1 and 4:1, respectively. By contrast, measurements of Augustine's plume, together with ground-based gas sampling in July 1986 when the volcano was in a posteruptive fuming state, are 380 t/d SO2 and approximately 8000 t/d HCl with no ash emission. The observations of large Cl releases at Augustine support the Cl abundance conclusions of Johnston (1980) based on study of melt inclusions in the 1976 lavas. The results reinforce the need for more measurements during eruptions and for better understanding of scaling of volcanic emissions of various eruptive components. ¿ American Geophysical Union 1988