Multi-component gas emission measurements of the active lava lake of Nyiragongo, DR Congo

Artikel i vetenskaplig tidskrift, 2017

Between 2007 and 2011 four measurement campaigns (June 2007, July 2010, June 2011, and December 2011) were carried out at the crater rim of Nyiragongo volcano, DR Congo. Nyiragongo is one of the most active volcanoes in Africa. The ground-based remote sensing technique Multi-Axis Differential Optical Absorption Spectroscopy (MAX-DOAS), which uses scattered sunlight, the in-situ Multi-Component Gas Analyzer System (Multi-GAS) and alkaline impregnated filter were simultaneously applied during all field trips. The bromine monoxide to sulfur dioxide (BrO/SO2) and carbon dioxide to sulfur dioxide (CO2/SO2) molar ratios were determined, among other ratios. During the different field trips variations of the level of the lava lake up to several tens of meters were observed during intervals of the order of minutes up to days and also between the years. The measured gas ratios presented covariations with the lava lake level changes. BrO/SO2 ratios and CO2/SO2 ratios showed similar behavior. Annual CO2/SO2 and BrO/SO2 average values are generally positively correlated. In June 2011 increased BrO/SO2 as well as increased CO2/SO2 ratios have been observed before a sudden decrease of the lava lake. Overall the Cl/S ratio, determined by filter-pack sampling, shows an increasing trend with time, which is accompanied by a decreasing sulfur dioxide flux, the later measured nearly continuously by automated MAX-DOAS instruments since 2004. Mean gas emission fluxes of CO2, Cl and ‘minimum-BrO’ fluxes are calculated using their ratio to SO2. The first two show an increase with time, in contrast to the SO2 fluxes. A simple conceptual model is proposed which can explain in particular the June 2011 data, but as well our entire data set. The proposed model takes up the idea of convective magma cells inside the conduit and the possible temporary interruption of part of the cycling. We propose than two alternatives to explain the observed gas emission variation: 1. It is assumed that the diffuse and fumarolic degassing could have significant influence on measured gas composition. The measured gas composition might rather represent a gas mixture of plume, diffuse and fumarolic degassing than only representing the volcanic plume. 2. It is proposed that the interruption of the convection has taken place in the upper part of the conduit and deep degassing of CO2 and bromine initially continues while mixing already with gas emissions from an ageing source, which is characterized by an already diminishing sulfur content. These complex process but as well as the gas mixing of different sources, could explain general features of our dataset, but can unfortunately neither be confirmed nor disproven by the data available today.