Long-term year-round observations of magmatic CO2 emissions on Mammoth Mountain, California, USA

Abstract Diffuse emission of magmatic CO2 is one of the main indicators of volcanic unrest at Mammoth Mountain, but the presence of deep seasonal snowpack at the site has hindered year-round CO2 flux observations. A permanent eddy covariance station was established at the largest area of diffuse CO2 degassing on Mammoth Mountain (Horseshoe Lake tree kill) that measured CO2 fluxes (Fc) and meteorological parameters on a half-hourly basis. From July 22, 2014 to May 24, 2020, Fc ranged from −35 to 10,546 g m−2 d−1. Fc decreased on average by 53% over the study period, tracking the long-term decline in CO2 emissions following the last major increase that occurred at the Horseshoe Lake tree kill area from 2009 to 2011. Statistical and spectral analyses were applied to the Fc and ancillary meteorological parameter time series to understand (1) relationships between these parameters, (2) their dominant periodicities, and (3) changes in Fc that may be unexplained by meteorological forcing. Variations in detrended Fc (Fcdt) were most strongly correlated with wind direction and atmospheric temperature, followed by atmospheric pressure on diurnal to annual time scales, but wind direction likely exerted the most direct control on Fcdt. Comparison of the smoothed (180-d span) Fcdt time series to the time series of average-daily snow water equivalent measured ~1 km away suggested that snowpack may have suppressed CO2 emissions. No evidence of a change in CO2 emissions related to the last major seismic swarm beneath Mammoth Mountain on February 2–18, 2014 was observed.