Infrasound data analysis of signals produced by Greenlandic glaciers

Abstract

Monitoring geo-physical activities in remote areas on a daily basis can be challenging and resource intensive. Passive remote sensing by means of infrasound waves allows resource efficient and continuous observation from distance. Infrasound waves are used, for example, to detect, monitor and analyse natural and man-made events, like avalanches, volcanoes, nuclear missile tests or - as in this paper - run-off and calving events from glaciers. In this respect, this paper will deploy and analyse data produced by the infrasound array I18DK in Qaanaaq (Greenland). Firstly, the most active infrasound sources are analysed in time/frequency domain for the year 2016. After determination of the back azimuth of the main sources, beamforming was used to analyse the spectral content. The closest, land-terminating glacier in north-eastern direction shows a constant rumbling at a frequency range of 1-2 Hz. In comparison, the three analysed sea-terminating glaciers in southern direction have short-lived, impulsive events with a broader frequency content. Secondly, near-field sources of infrasound - with a measured spherical wave - are recognised and an attempt is made to localise these sources. Measured spherical waves may contain information of the source location. The Qtau parameter is presented as a discrimination tool between far-field plane waves and near-field spherical waves. This parameter describes the misfit of time difference of arrivals (TDOA) compared to a plane wave. The retrieved sources are then localised with a linear inversion based on the TDOA and a forward approach. Accurate source locations are for both algorithms found using synthetic data. The measured Qtau values indicate a near-field source to the glacier located in north-eastern area of the infrasound station. The inverse source localisation shows a clear source direction, however, an exact localisation remains a challenge. By comparison, the forward problem shows similar directions. The analysed Qtau values are presented as meaningful method to recognise near-field infrasound sources. A source localisation based on the TDOA and the used forward approach is only possible in the near surrounding of an infrasound station. A higher sampling rate, larger array configuration and a proper adjustment of the array orientation have large impact on the identification of spherical waves and hence, the source localisation.