Abstract. Using data from a network of 58 seismic stations, we characterise a large landslide that occurred at the southeastern corner of the Askja caldera, Iceland, on 21 July 2014, including its precursory tremor and mass wasting aftermath. Our study is motivated by the need for deeper generic understanding of the processes operating not only at the time of catastrophic slope failure, but also in the preparatory phase and during the transient into the subsequent stable state. In addition, it is prompted by the high hazard potential of the steep caldera lake walls at Askja as tsunami waves created by the landslide reached famous tourist spots 60 m above the lake level. Since direct observations of the event are lacking, the seismic data give valuable details on the dynamics of this landslide episode. The excellent seismic data quality and coverage of the stations of the Askja network made it possible to jointly analyse the long- and short-period signals of the landslide to obtain information about the triggering, initiation, timing, and propagation of the slide. The seismic signal analysis and a landslide force history inversion of the long-period seismic signals showed that the Askja landslide was a single, large event starting at the SE corner of the caldera lake at 23:24:05 UTC and propagating to the NW in the following 2 min. The bulk sliding mass was 7–16 × 1010 kg, equivalent to a collapsed volume of 35–80 × 106 m3, and the centre of mass was displaced horizontally downslope by 1260 ± 250 m during landsliding. The seismic records of stations up to 30 km away from the landslide source area show a tremor signal that started 30 min before the main landslide failure. It is harmonic, with a fundamental frequency of 2.5 Hz and shows time-dependent changes of its frequency content. We attribute the complex tremor signal to accelerating and decelerating stick-slip motion on failure planes at the base and the sides of the landslide body. The accelerating motion culminated in aseismic slip of the landslide visible as a drop in the seismic amplitudes down to the background noise level 2 min before the landslide high-energy signal begins. We propose that the seismic signal of the precursory tremor may be developed as an indicator for landslide early-warning systems. The 8 hours after the main landslide failure are characterised by smaller slope failures originating from the destabilised caldera wall decaying in frequency and magnitude. We introduce the term afterslides for this subsequent, declining slope activity after a large landslide.
GPS Derived Seismic Signals for Far Field Earthquake Epicenter Location Estimation
