Acid gas emissions from Galeras Volcano, Colombia, 1989–1994

Characterization of acid gases discharged by Galeras Volcano have allowed us to define a clear change in the pattern of degassing during 1989 through 1994, probably associated with pressure buildup that ended with the destruction of the dome by an explosive eruption on July 16, 1992. Using Japanese boxes for sampling, it was possible to establish characteristic discharge (absorption) levels for the period of study and to conclude that the absolute discharge of acid gases was declining at least until the middle of 1991. A staggered temporal response in gas compositions, particularly CO2 and St, for the different fumaroles suggests at least three possibilities: (1) magma resurgence, which would increase relative CO2 discharge as a result of new gas contribution and its relatively low solubility in magma; (2) solidification and/or crystallization of magma, first in a zone of higher thermal gradient where there is a greater interaction with the hydrothermal system (Calvache fumarole), and second toward the volcanic-magmatic fumaroles (Deformes and Besolima); and (3) contribution of the hydrothermal system, produced by water saturation and/or temperature increase that would reduce CO2 solubility in water. The similar behavior of the Calvache fumarole and Japanese Box 3, as well as the existence of La Trucha lake in the northeastern moat between the active cone and the amphitheater margin, help us to define the location of the principal hydrothermal system. The abrupt change toward a hydrothermal component in Deformes fumaroles in August 1990, and the variations in chloride concentrations and other cation concentrations in condensate samples, allow us to observe very large changes over short periods in the composition of the fluid phase of the volcanic system. This suggests that the density of sampling should be increased, together with the development of alternative techniques that could reduce risks and augment the resolution in the observation of the chemical changes. On the other hand, the dependence of gas absorption (e.g., CO2) on the sampling period, seen at the Japanese boxes, suggests that fixed sampling intervals would improve this method by reduction of sampling variables.