Constraining the wavefield of volcano-seismic events on Mt. Etna, Italy through a rotational sensor and seismic array observations

Abstract. Long-period (LP) events and tremor are characteristic seismic signals of active volcanoes, offering insight into underlying fluid-driven processes. However, their emergent wavefield is complex and challenging to characterise. Here we decipher the LP event and tremor wavefield composition using a seismic array combined with a rotational sensor co-located with a seismometer (6C station). This study analyses and compares directional and phase velocity estimates by processing a 25 d long dataset from a rotational sensor and an array of seven broadband stations deployed at Mt. Etna, Italy, in August–September 2019. We derive the back azimuths (BAz) of LP events and tremor from both the seismometer array and the 6C station, and we compare these estimates with a reference BAz obtained from the network locations from the Istituto Nazionale di Geofisica e Vulcanologia-Osservatorio Etneo (INGV-OE) on Mt. Etna. Volcanic tremor occurs in distinct phases with varying seismic and surface activity. Depending on the phase, either the array or the 6C method provides reliable tremor BAz estimates, agreeing well with the INGV-OE reference. We find that BAz estimates of both methods are shifted southward relative to the reference location for the LP events. We attribute the larger southward deviation observed in the 6C results to local heterogeneities which exert a stronger influence on the 6C station than on the array. Based on the array derived slownesses we infer that the tremor and LP events mainly consist of surface waves. Further, the rotational sensor recordings suggest a wavefield dominated by SH-type waves. In combination with the observed temporal evolution of the 6C phase velocity in narrow frequency bands, we infer Love-wave dominance. This study highlights the value of a rotational sensor to constrain the wavefield in a deterministic way in a complex volcanic environment.