Monitoring volcanic systems through cross-correlation of coincident A-Train satellite data.

Thursday, 18 December 2014
Verity J. B. Flower1, Simon A Carn1 and Robert Wright2, (1)Michigan Technological University, Houghton, MI, United States, (2)Univ. Hawaii/HIGP, Honolulu, HI, United States
The remote location and inaccessibility of many active volcanic systems around the world hinders detailed investigation of their eruptive dynamics. One methodology for monitoring such locations is through the utilisation of multiple satellite datasets to elucidate underlying eruption dynamics and aid volcanic hazard mitigation. Whilst satellite datasets are often analysed individually, here we exploit the multi-platform NASA A-Train satellite constellation, including the Ozone Monitoring Instrument (OMI) on Aura and Moderate Resolution Imaging Spectroradiometer (MODIS) on Aqua. OMI measures volcanic emissions (e.g. sulphur dioxide, ash) whilst MODIS enables monitoring of thermal anomalies (e.g. lava flows, lava lakes, pyroclastic deposits), allowing analysis of a more diverse range of volcanic unrest than is possible using a single measurement technique alone, and permitting cross-correlation between datasets for specific locations to assess cyclic activity. A Multi-taper (MTM) Fast Fourier Transform (FFT) analysis was implemented at an initial sample site (Soufriere Hills volcano [SHV], Montserrat) facilitating cycle identification and subsequent comparison with existing ground-based data. Corresponding cycles at intervals of 8, 12 and ~50 days were identified in both the satellite-based SO2 and thermal infrared signals and ground-based SO2 measurements (Nicholson et al. 2013), validating the methodology. Our analysis confirms the potential for identification of cyclical volcanic activity through synergistic analysis of satellite data, which would be of particular value at poorly monitored volcanic systems. Following our initial test at SHV, further sample sites have been selected in locations with varied eruption dynamics and monitoring capabilities including Ambrym (Vanuatu), Kilauea (Hawaii), Nyiragongo (DR Congo) and Etna (Italy) with the intention of identifying not only cyclic signals that can be attributed to volcanic systems but also those which are artefacts of the measurement technique employed.