DYNAMICS OF CARBON-DIOXIDE EMISSIONS, CRYSTALLIZATION, AND MAGMA ASCENT - HYPOTHESES, THEORY, AND APPLICATIONS TO VOLCANO MONITORING AT MOUNT ST-HELENS

Measurements of CO2 fluxes from open-vent volcanos are rare, yet may o ffer special capabilities for monitoring volcanos and forecasting acti vity. The measured fluxes of CO2 and SO2 from Mount St. Helens decreas ed from July through November 1980, but the record includes variations of CO2/SO2 in the emitted gas and episodes of greatly increased fluxe s of CO2. We propose that the CO2 flux variations reflect two gas comp onents: (a) a component whose flux decreased in proportion to 1/root t with a CO2/SO2 mass ratio of 1.7, and (b) a residual flux of CO2 cons isting of short-lived, large peaks with a CO2/SO2 mass ratio of 15. We propose two hypotheses: (a) the 1/root t dependence was generated by crystallization in a deep magma body at rates governed by diffusion-li mited heat transfer, and (b) the gas component with the higher CO2/SO2 was released from ascending magma, which replenished the same magma b ody. The separation of the total CO2 flux into contributions from know n processes permits quantitative inferences about the replenishment an d crystallization rates of open-system magma bodies beneath volcanos. The flux separations obtained by using two gas sources with distinct C O2/SO2 ratios and a peak minus background approach to obtain the CO2 c ontributions from an intermittent source and a continuously emitting s ource are similar. The flux separation results support the hypothesis that the second component was generated by episodic magma ascent and r eplenishment of the magma body. The diffusion-limited crystallization hypothesis is supported by the decay of minimum CO2 and SO2 fluxes wit h 1/root t after 1 July 1980. We infer that the magma body at Mount St . Helens was replenished at an average rate (2.8 x 10(6) m(3) d(-1)) w hich varied by less than 5% during July, August, and September 1980. T he magma body volume (2.4-3.0 km(3)) in early 1982 was estimated by in tegrating a crystallization rate function inferred from CO2 fluxes to maximum times (20 +/- 4 years) estimated from the increase of sample c rystallinity with time. These new volcanic gas flux separation methods and the existence of relations among the CO2 flux, crystallization ra tes, and magma body replenishment rates yield new information about th e dynamics of an open-vent, replenished magma body.