Multi-scale observations of the variability of magmatic CO2 emissions, Mammoth Mountain, CA, USA

Abstract

One of the primary indicators of volcanic unrest at Mammoth Mountain is diffuse emission of magmatic CO2, which can effectively track this unrest if its variability in space and time and relationship to near-surface meteorological and hydrologic phenomena versus those occurring at depth beneath the mountain are understood. In June–October 2013, we conducted accumulation chamber soil CO2 flux surveys and made half-hourly CO2 flux measurements with automated eddy covariance and accumulation chamber (auto-chamber) instrumentation at the largest area of diffuse CO2 degassing on Mammoth Mountain (Horseshoe Lake tree kill; HLTK). Estimated CO2 emission rates for HLTK based on 20 June, 30 July, and 24–25 October soil CO2 flux surveys were 165, 172, and 231td−1, respectively. The average (June–October) CO2 emission rate estimated for this area was 123td−1 based on an inversion of 4527 eddy covariance CO2 flux measurements and corresponding modeled source weight functions. Average daily eddy covariance and auto-chamber CO2 fluxes consistently declined over the four-month observation time. Wavelet analysis of auto-chamber CO2 flux and environmental parameter time series was used to evaluate the periodicity of, and local correlation between these variables in time–frequency space. Overall, CO2 emissions at HLTK were highly dynamic, displaying short-term (hourly to weekly) temporal variability related to meteorological and hydrologic changes, as well as long-term (monthly to multi-year) variations related to migration of CO2-rich magmatic fluids beneath the volcano. Accumulation chamber soil CO2 flux surveys were also conducted in the four additional areas of diffuse CO2 degassing on Mammoth Mountain in July–August 2013. Summing CO2 emission rates for all five areas yielded a total for the mountain of 311td−1, which may suggest that emissions returned to 1998–2009 levels, following an increase from 2009 to 2011.