SO2 and BrO emissions of Masaya volcano from 2014 to 2020

Masaya (Nicaragua, 12.0∘ N, 86.2∘ W; 635 m a.s.l.) is one of the few volcanoes hosting a lava lake, today. This study has two foci: (1) discussing the state of the art of long-term SO2 emission flux monitoring with the example of Masaya and (2) the provision and discussion of a continuous data set on volcanic gas data with a large temporal coverage, which is a major extension of the empirical database for studies in volcanology as well as atmospheric bromine chemistry. We present time series of SO2 emission fluxes and BrO/SO2 molar ratios in the gas plume of Masaya from March 2014 to March 2020 – covering the three time periods (1) before the lava lake appearance, (2) a period of high lava lake activity (November 2015 to May 2018), and (3) after the period of high lava lake activity. For these three time periods, we report average SO2 emission fluxes of (1000±200), (1000±300), and (700±200) t d-1 and average BrO/SO2 molar ratios of (2.9±1.5)×10-5, (4.8±1.9)×10-5, and (5.5±2.6)×10-5.Our SO2 emission flux retrieval is based on a comprehensive investigation of various aspects of spectroscopic retrievals, the wind conditions, and the plume height. We observed a correlation between the SO2 emission fluxes and the wind speed in the raw data. We present a partial correction of this artefact by applying dynamic estimates for the plume height as a function of the wind speed. Our retrieved SO2 emission fluxes are on average a factor of 1.4 larger than former estimates based on the same data.Further, we observed different patterns in the BrO/SO2 time series: (1) an annual cyclicity with amplitudes between 1.4 and 2.5×10-5 and a weak semi-annual modulation, (2) a step increase by 0.7×10-5 in late 2015, (3) a linear trend of 1.4×10-5 per year from November 2015 to March 2018, and (4) a linear trend of -0.8×10-5 per year from June 2018 to March 2020. The step increase in 2015 coincided with the lava lake appearance and was thus most likely caused by a change in the magmatic system. We suggest that the cyclicity might be a manifestation of meteorological cycles. We found an anti-correlation between the BrO/SO2 molar ratios and the atmospheric water concentration (correlation coefficient of -0.47) but, in contrast to that, neither a correlation with the ozone mixing ratio (+0.21) nor systematic dependencies between the BrO/SO2 molar ratios and the atmospheric plume age for an age range of 2–20 min after the release from the volcanic edifice. The two latter observations in