Now showing 1 - 6 of 6
  • Publication
    Multiple Coincident Eruptive Seismic Tremor Sources During the 2014-2015 Eruption at Holuhraun, Iceland
    We analyze eruptive tremor during one of the largest effusive eruptions in historical times in Iceland (2014/2015 Holuhraun eruption). Seismic array recordings are compared with effusion rates deduced from Moderate Resolution Imaging Spectroradiometer recordings and ground video monitoring data and lead to the identification of three coexisting eruptive tremor sources. This contrasts other tremor studies that generally link eruptive tremor to only one source usually associated with the vent. The three sources are (i) a source that is stable in back azimuth and shows bursts with ramp-like decrease in amplitude at the beginning of the eruption: we link it to a process below the open vents where the bursts correlate with the opening of new vents and temporary increases in the lava fountaining height; (ii) a source moving by a few degrees per month while the tremor amplitude suddenly increases and decreases: back azimuth and slowness correlate with the growing margins of the lava flow field, whilst new contact with a river led to fast increases of the tremor amplitude; and (iii) a source moving by up to 25∘ southward in 4 days that cannot be related to any observed surface activity and might be linked to intrusions. We therefore suggest that eruptive tremor amplitudes/energies are used with caution when estimating eruptive volumes, effusion rates, or the eruption explosivity as multiple sources can coexist during the eruption phase. Our results suggest that arrays can monitor both the growth of a lava flow field and the activity in the vents.
    Scopus© Citations 27  348
  • Publication
    Helicopter vs. volcanic tremor: Characteristic features of seismic harmonic tremor on volcanoes
    We recorded high-frequency (> 10 Hz) harmonic tremor with spectral gliding at Hekla Volcano in Iceland. Particle motion plots indicated a shallow tremor source. We observed up to two overtones beneath our Nyquist frequency of 50 Hz and could resolve a source of closely spaced pulses of very short duration (0.03-0.1 s) on zoomed seismograms. Volcanic tremor with fundamental frequencies above 5 Hz, frequency gliding and/or repetitive sources similar to our observations were observed on different volcanoes around the world. However, this frequency content, duration and occurrence of volcano-related tremor was not observed in the last 35 years of seismic observations at Hekla. Detailed analysis reveals that this tremor was related to helicopters passing the volcano. This study relates the GPS track of a helicopter with seismic recordings of the helicopter at various distances. We show the effect the distance, number of rotor blades and velocity of the helicopter has on the observed up and down glidings at up to 40 km distance. We highlight similarities and differences between volcano-related and helicopter tremor in order to help avoid misinterpretations.
    Scopus© Citations 29  507
  • Publication
    Tremor-rich shallow dyke formation followed by silent magma flow at Bárðarbunga in Iceland
    The Bárðarbunga eruption in Iceland in 2014 and 2015 produced about 1.6 km3 of lava. Magma propagated away from Bárðarbunga to a distance of 48 km in the subsurface beneath Vatnajökull glacier, emerging a few kilometres beyond the glacier's northern rim. A puzzling observation is the lack of shallow (<3 km deep), high-frequency earthquakes associated Q.1 with shallow dyke formation near the subaerial and subglacial eruptive sites, suggesting that near-surface dyke formation is seismically quiet. However, seismic array observations and seismic full wavefield simulations reveal the presence and nature of shallow, pre-eruptive, long-duration seismic tremor activity. Here we use analyses of seismic data to constrain therelationships between seismicity, tremor, dyke propagation and magma flow during the Bárðarbunga eruption. We show that although tremor is usually associated with magma flow in volcanic settings, pre-eruptive tremor at Bárðarbunga was probably caused by swarms of microseismic events during dyke formation, and hence is directly associated with fracturing of the upper 2-3 km of the crust. Subsequent magma flow in the newly formed shallow dyke was seismically silent, with almost a complete absence of seismicity or tremor. Hence, we suggest that the transition from temporarily isolated, large, deep earthquakes to many smaller, shallower, temporally overlapping earthquakes (< magnitude 2) that appear as continuous tremor announces the arrival of a dyke opening in the shallow crust, forming a pathway for silent magma flow to the Earth's surface. 
      778Scopus© Citations 37
  • Publication
    Helicopter location and tracking using seismometer recordings
    We use frequency domain methods usually applied to volcanic tremor to analyse ground based seismic recordings of a helicopter. We preclude misinterpretations of tremor sources and show alternative applications of our seismological methods. On a volcano, the seismic source can consist of repeating, closely spaced, small earthquakes. Interestingly, similar signals are generated by helicopters, due to repeating pressure pulses from the rotor blades. In both cases the seismic signals are continuous and referred to as tremor. As frequency gliding is in this case merely caused by the Doppler effect, not a change in the source, we can use its shape to deduce properties of the helicopter and its flight path. We show in this analysis that the number of rotor blades, rotor revolutions per minute (RPM), helicopter speed, flight direction, altitude and location can be deduced from seismometer recordings. Access to GPS determined flight path data from the helicopter offers us a robust way to test our location method.
    Scopus© Citations 12  494
  • Publication
    Micrometre-scale deformation observations reveal fundamental controls on geological rifting
    Many of the world's largest volcanic eruptions are associated with geological rifting where major fractures open at the Earth's surface, yet fundamental controls on the near-surface response to the rifting process are lacking. New high resolution observations gleaned from seismometer data during the 2014 Bároarbunga basaltic dyke intrusion in Iceland allow us unprecedented access to the associated graben formation process on both sub-second and micrometre scales. We find that what appears as quasi steady-state near-surface rifting on lower resolution GPS observation comprises discrete staccato-like deformation steps as the upper crust unzips through repetitive low magnitude (M < 0) failures on fracture patches estimated between 300 m and 1200 m in size. Stress drops for these events are one to two orders of magnitude smaller than expected for tectonic earthquakes, demonstrating that the uppermost crust in the rift zone is exceptionally weak.
      97Scopus© Citations 12
  • Publication
    Persistent Shallow Background Microseismicity on Hekla Volcano, Iceland: A Potential Monitoring Tool
    Hekla is one of Iceland's most active volcanoes. Since 1970 it has erupted four times with a period of quiescence of 14 years since the last eruption. We detected persistent levels of background microseismicity with a temporary seismic network in autumn 2012. An amplitude based as well as an arrival-time based location method was applied to two populations of events and located them at shallow depths on the northern flank, close to the summit. This seismicity has not been identified previously by the permanent seismic network in Iceland as it is below its detection threshold. The detected events were either short, higher frequency events with distinct arrivals located beneath the summit on the northern flank of Hekla or longer, emergent, lower frequency events about 4 km northeast of the summit at 200¿400 m depth below the surface. Estimated moment magnitudes were MW = -1.1 to -0.1 and MW = -0.9 to -0.0 and local magnitudes ML = -0.5 to +0.3 and ML = -0.3 to +0.3, respectively. This seismicity does not show any correlation with gas output but is located at the steepest slopes of the edifice. Hence we suggest that the current shallow microseismicity at Hekla is structurally controlled. This offers a possible opportunity of using near summit microseismicity as a tool for monitoring emerging unrest at Hekla. Microseismicity rates will be very sensitive to small stress perturbations due to magma migration at depth. Currently in the absence of microseismicity monitoring, Hekla switches from apparently quiescent to fully eruptive on the order of only 1 h.
      413Scopus© Citations 6