Permanent, seasonal, and episodic seismic sources around Vatnajökull, Iceland from the analyses of correlograms

This study aims at characterizing different seismic sources in the region of the Vatnajökull glacier using seismic interferometry. Vatnajökull is the largest Icelandic icecap, covering 4active volcanic systems. The seismic context is therefore very complex with glacial and volcanic events occurring simultaneously and a classification between the two can become cumbersome. 

We used seismic interferometry or cross-correlation of seismic noise on seismic data from 2011 to 2019). Being based on continuous records, this passive monitoring method is not relying on earthquakes to locate seismic sources. We computed the cross-correlation functions between every pair of seismic stations using MSNoise for different frequency bands, from 0.5 to 8 Hz. The first step towards the location of seismic sources was to calculate the propagation velocities for each frequency range. The total range of velocities is between 1.39 km/s and 3.92 km/s. Then, we used two different location methods based on the calculated propagation velocities. The first method is based on hyperbole’s geometry and provides the location of seismic sources as the intersection between several hyperboles, while the second one, the Ballmer’s method (Ballmer et al. 2013), is based on the calculation of theoretical differential times and provides location probabilities for the seismic sources. We located and characterized persistent oceanic seismic noise located along the southern shoreline of Iceland potentially associated with waves activity and geometry of the shore, as well as a seasonal glacial tremor around outlet glaciers in the west part of the Vatnajökull icecap, potentially linked to glacial processes inside the glacier or in the glacial rivers. The uncertainty of a few kilometers is observed. Some limitations exist for these methods. For example, The Ballmer’s method (Ballmer et al. 2013) is reliable for seismic sources inside the seismic network but can only give an azimuthal direction for seismic sources located outside of it. When using hyperboles, slightly different propagation velocities between pairs of stations can affect the precision of the intersection. Therefore, the association of the two methods is important to diminish the impact of these limitations. 

These results provide a better understanding of the seismic background of this region and will be compared and validated with other localization methods in the future.