Advances in gas and aerosol monitoring at active volcanoes.

摘要

To improve interpretation of volcanic SO2 flux data, it is necessary to quantify and understand reactions involving SO 2 in volcanic plumes. Uncertainties in the near-source plume chemistry can complicate interpretations of volcanic activity and hazards, petrology, global emission rates and climatic effects of emissions. The main objectives of this research are to quantify SO2 emission rates and calculate SO2 loss rates at volcanoes within different environments (e.g. low altitude-high humidity vs. high altitude-low humidity) using ground based remote sensing techniques. The work, divided in three main parts, uses new techniques to quantify SO2 decrease rates.; We conducted our measurements using mini-UV spectrometers, which are quickly replacing the COSPEC as the instrument of choice for SO2 measurements. In order to apply this technique, which has only been used in volcanology since 2001, we started by using a COSPEC to measure SO2 fluxes at five volcanoes in Guatemala and El Salvador (1999--2002). This work included extensive error analysis, for instrumental and non-instrumental errors, which can be applied to the analysis of SO2 data collected at the target volcanoes, using mini-UV spectrometers. We also described the contributions of these volcanoes, which are the most active in Guatemala and El Salvador, in the context of the global sulfur budget. This resulted in 6--12% of the global budget, which is high and shows that the global budget is probably underestimated.; In the second part we conducted SO2 measurements, using mini-UV spectrometers, of ash-free plumes in 2004 near the vent and at several distances downwind at Soufriere Hills volcano, Montserrat (SHV) and Lascar volcano, Chile, in order to quantify SO2 loss rates. These were chosen mainly because their plumes are emitted to different parts of the troposphere (planetary boundary layer and free troposphere), and they represent humid and dry atmospheres, respectively. At both volcanoes we conducted measurements for three to four days. Average SO2 loss rates of ∼10 -4 s-1 (e-folding time of ∼2.78 hours) were calculated at both volcanoes. However, those at SHV are more accurate and representative of the general case in Montserrat. Those at Lascar are subject to large errors due to uncertainties in plume geometry and effects of UV scattering; therefore the loss rates are probably very slow to negligible. An underestimate of 70--146% of the at source SO2 emission rate was calculated at SHV. This suggests that the global volcanic SO2 emission rate may be underestimated, as it is based on measurements typically taken downwind of volcanoes, by which time significant loss of SO2 may have taken place.; Finally, to complement the gas measurements at Lascar we also measured aerosol properties with a Microtops II sun photometer. The plume optical depths were extremely low (generally below 0.1), and the plumes were very optically thin. The plumes showed bimodal and trimodal distributions of particles, with a predominance of smaller particles (sulfates). We also observed that the particles in a plume measured downwind have larger effective radii, and their distributions showed an increase in the large particle mode (large water droplets and/or ice). We conclude that the size evolution in the plumes at Lascar is controlled mainly by adsorption/absorption processes close to the vent.