Abstract : A study of melt inclusions (MIs) has allowed us to follow the dissolved volatile evolution in the magmatic system of Hekla volcano during the melt differentiation (fractional crystallization). This work has allowed us to predict the “expected” volatile concentrations, which correspond to the volatile contents in the basaltic icelandite melt just prior to eruption. Such approach avoids underestimation of the pre-eruptive volatile contents measured in MIs and improves the estimates on the volatile mass released into the atmosphere. This approach has suggested that the magma carried 0.1 Mt of HCl, 0.2 Mt of HF and 3.8 of SO2 into the atmosphere during the 2000 Hekla eruption. Moreover, the Plinian phase of this eruption emitted a mixed tephra-gas plume that was scavenged by snowstorms, offering the unique opportunity to better study the geochemistry of this explosive gas phase. The study of these snow samples has shown that enrichment of most of volatile trace elements is linked to a degassing process as halides and/or sulfates. More surprisingly, refractory elements are also enriched in the gaseous phase of Hekla and they are best explained by the non-stoechiometric partial dissolution of tephra by F-rich volcanic gases in the eruptive plume.
A study based on MIs has confirmed the hypothesis emitted by Walker et al. (1993) in which differentiation of tholeiitic magmas of Masaya occurs at shallow depth from a "dry" magma also characterized by its homogeneous composition. Moreover, chemical and physical characterizations of volcanic aerosols by SEM and chemical analyses of the eruptive phase of Masaya have allowed us to conclude that trace elements are mostly degassed and transported in Masaya's plume as chlorides, but also as sulfates and/or chloro-sulfates. Volatile trace element contribution to the atmosphere has also been estimated showing that the volcanic plume of Masaya is an important source of atmospheric pollution on the regional scale.