Segmented lateral dyke growth in a rifting event at Bárðarbunga volcanic system, Iceland

Journal name:
Nature
Volume:
517,
Pages:
191–195
Date published:
DOI:
doi:10.1038/nature14111
Received
Accepted
Published online

Crust at many divergent plate boundaries forms primarily by the injection of vertical sheet-like dykes, some tens of kilometres long1. Previous models of rifting events indicate either lateral dyke growth away from a feeding source, with propagation rates decreasing as the dyke lengthens2, 3, 4, or magma flowing vertically into dykes from an underlying source5, 6, with the role of topography on the evolution of lateral dykes not clear. Here we show how a recent segmented dyke intrusion in the Bárðarbunga volcanic system grew laterally for more than 45 kilometres at a variable rate, with topography influencing the direction of propagation. Barriers at the ends of each segment were overcome by the build-up of pressure in the dyke end; then a new segment formed and dyke lengthening temporarily peaked. The dyke evolution, which occurred primarily over 14 days, was revealed by propagating seismicity, ground deformation mapped by Global Positioning System (GPS), interferometric analysis of satellite radar images (InSAR), and graben formation. The strike of the dyke segments varies from an initially radial direction away from the Bárðarbunga caldera, towards alignment with that expected from regional stress at the distal end. A model minimizing the combined strain and gravitational potential energy explains the propagation path. Dyke opening and seismicity focused at the most distal segment at any given time, and were simultaneous with magma source deflation and slow collapse at the Bárðarbunga caldera, accompanied by a series of magnitude M > 5 earthquakes. Dyke growth was slowed down by an effusive fissure eruption near the end of the dyke. Lateral dyke growth with segment barrier breaking by pressure build-up in the dyke distal end explains how focused upwelling of magma under central volcanoes is effectively redistributed over long distances to create new upper crust at divergent plate boundaries.

At a glance

Figures

  1. Overview of data.
    Figure 1: Overview of data.

    a, Earthquakes 16 August to 6 September 2014 (dots) and horizontal ground displacements measured by GPS (arrows) on a map with central volcanoes (oval outlines), calderas (hatched), and northern Vatnajökull. Relatively relocated epicentres and displacements are colour coded according to time of occurrence (key at top left), other single earthquake locations are in grey. Rectangles show areas displayed in Fig. 2; thin lines within them show inferred dyke segments. The red shading at Bárðarbunga caldera shows subsidence up to 16 m inferred from radar profiling on 5 September. The star marks the location of the magma source inferred from modelling. Also shown are ice cauldrons formed (circles), outline of lava flow mapped from radar image on 6 September, and eruptive fissures (white). b, Wrapped RADARSAT-2 interferogram spanning 8 August to 1 September 2014. Shading at Bárðarbunga caldera shows unwrapped one day (27–28 August) COSMO-SkyMed interferogram with maximum LOS (line of sight) increase of 57 cm. Also shown are earthquakes (grey dots), boundaries of graben activated in the dyke distal area (hatched lines), and location of interferograms in c and d (boxes). c, Unwrapped Cosmo-SkyMed interferogram spanning 13–29 August. d, Unwrapped TerraSAR-X interferogram spanning 26 July 2012 to 4 September 2014. Satellite flight and viewing direction are shown with black and white arrows, respectively. LOS displacement is positive away from the satellite for all interferograms shown.

  2. Seismicity and magma volume along the dyke, 16 August to 6 September 2014.
    Figure 2: Seismicity and magma volume along the dyke, 16 August to 6 September 2014.

    Relocated earthquakes shown in Fig. 1 are indicated, with same colour coding. a, Daily cumulative seismic moment at 0.5 km intervals along the dyke. b, Plan-view of four rotated areas along the dyke. Arrows indicate geographic north. Dots denote epicentres, black lines dyke segments, and open circles ice cauldrons. Fault-plane solutions for selected earthquakes are shown. c, Earthquake depths referenced to sea level. d, Daily vertically integrated volume of magma along the dyke inferred from geodetic modelling.

  3. Dyke model.
    Figure 3: Dyke model.

    a, Median of the posterior probability of opening for dyke patches inferred from modelling (key at right), and relocated earthquake hypocentres (black and grey dots, in front of and behind the dyke plane, respectively) relative to sea level. Red stars mark the eruption sites. b, Preferred direction of dyking for different segments based on a model of combined strain and gravitational potential energy release. Blue lines represent dyke segments and grey dots earthquake epicentres. Black dots indicate the beginning of each segment and surrounding arc of coloured points represent possible end points for different strikes of propagation. Their colour (key at right) indicates (E − Emin)/(Emax − Emin), where E is the energy state for a particular strike, and Emin and Emax are the respective minimum and maximum energy state for that segment. Background shows bedrock topography with calderas outlined (in black). Grey thick line shows the edge of the Vatnajökull ice cap.

  4. Seismicity and volume change.
    Figure 4: Seismicity and volume change.

    Dots indicate earthquakes along the length of the dyke (left-hand y axis) as a function of time; stars, volume change (right-hand y axis) of the dyke (blue) and of the magma source (red). Earthquakes are colour coded by time (key along bottom), with dot radius proportional to earthquake magnitude. The volumes (error bars show 95% confidence intervals) are estimated from available geodetic data for each day using a model of a point pressure source and two dip-slip faults beneath the caldera. The magma source volumes are scaled by a factor of two, such that the value estimated for 5 September from GPS and InSAR data alone becomes equal to that estimated when the caldera subsidence is added to the inversion. Shading indicates the Holuhraun eruptions (see main text).

  5. Tectonic map showing seismic and geodetic stations.
    Extended Data Fig. 1: Tectonic map showing seismic and geodetic stations.

    Filled uptriangles correspond to continuous GPS stations, open triangles to campaign GPS sites, filled downtriangles correspond to seismic stations, and black ‘stars’ correspond to co-located GPS and seismic stations. Station names for GPS are indicated with four upper-case letters and for seismic stations with three lower-case letters. The different tracks for SAR satellite data are plotted with solid lines for RADARSAT-2, dashed lines for COSMO-SkyMed (CSK) and dotted lines for TerraSAR-X (TSX). Orbit numbers are indicated for CSK and TSX. The red stars correspond to the 2014 eruptive fissures at Holuhraun. Background map shows ice caps (white), central volcanoes (dotted lines), calderas (hatched lines) and fissure swarms (grey shading)46. Names of selected volcanoes are shown, and T indicates Tungnafellsjökull central volcano. Inset, locations of the Eastern Volcanic Zone (EVZ), the Western Volcanic Zone (WVZ), the Northern Volcanic Zone (NVZ) and the South Iceland Seismic Zone (SISZ), with their fissure swarms and central volcanoes; box shows area of the main figure.

  6. Map and table of dyke segments defined by seismicity.
    Extended Data Fig. 2: Map and table of dyke segments defined by seismicity.

    a, Location of dyke segments (red lines labelled with numbers) delineated by relatively relocated earthquakes. The triangles show locations of the nearest seismic stations used to locate the events. Green, stations operated by the Icelandic Meteorological Office (IMO); blue, station operated by the University of Cambridge; and red, station operated by University College Dublin. The two stations on the ice are joint IMO and University of Cambridge stations. All stations telemetered data to IMO. Also shown are ice cauldrons formed (circles), outline of lava flow mapped from radar image on 6 September, and boundaries of graben activated in the dyke distal area (hatched lines). b, The dyke segments. Columns show segment number (Nr), latitude and longitude of beginning (Lat1, Lon1) and end points (Lat2, Long2) of each segment, segment length (L), depth range (D), strike, dip, the RMS value of the deviation (in metres) of the earthquakes from the plane they define, and the number of earthquakes used to define each dyke segment plane (#Eq).

  7. Subsidence at Bar[eth]arbunga volcano revealed by aeroplane radar profiling.
    Extended Data Fig. 3: Subsidence at Bárðarbunga volcano revealed by aeroplane radar profiling.

    Blue contours show the elevation of the ice surface in the Bárðarbunga caldera on 5 September 2014, about three weeks after the onset of unrest. The data were obtained using aircraft flown 100–150 m above glacier surface, using radar altimetry and submetre differential GPS Omnistar. The system provides 2 m absolute accuracy of surface elevation along the survey profiles47, shown as black dotted lines. The subsidence relative to the pre-unrest ice surface is indicated with coloured shading (key at right). It is greatest in the central part of the caldera, where it had a maximum of 16 m.

  8. Geodetic model with a contracting point pressure source, a two-segment-dyke and no a priori constraints.
    Extended Data Fig. 4: Geodetic model with a contracting point pressure source, a two-segment-dyke and no a priori constraints.

    The dyke segments are rectangular with uniform opening37 and the model shown is that with maximum probability. Free parameters are: position and volume change of the point pressure source39, and position, strike, dip and opening of the dyke. The panels show from left to right: data (a, d, g), model (b, e, h) and residuals (c, f, i). GPS data in all panels span 15 August to 4 September 2014. The top row shows an interferogram spanning 6 July 2012 to 4 September 2014; the middle row shows an interferogram spanning 2 August to 3 September 2014; the bottom row shows the data from Extended Data Fig. 3 from aeroplane radar profiling, but these data were not used to constrain the model.

  9. Geodetic model with a contracting point pressure source, two caldera faults, and a four-segment dyke.
    Extended Data Fig. 5: Geodetic model with a contracting point pressure source, two caldera faults, and a four-segment dyke.

    The dyke is divided into multiple rectangular patches37 and the model shown is that with maximum probability. Lateral dyke position is fixed from relocated seismicity. Position of the point source and faults, volume change of the point source and opening and strike-slip of the dyke are free parameters. Left column (a, d, g), data; middle column (b, e, h), model; right column (c, f, i), residuals. The data are detailed in Extended Data Fig. 4 and all of the data were used to constrain the model.

  10. Geodetic model with a contracting flat-topped chamber, two caldera faults, and a four-segment dyke.
    Extended Data Fig. 6: Geodetic model with a contracting flat-topped chamber, two caldera faults, and a four-segment dyke.

    A deflating penny-shaped crack38 is used to represent the top of a flat-topped chamber48, and the dyke is divided into multiple rectangular patches37. The model shown is that with maximum probability. Lateral dyke position is fixed from relocated seismicity. Position of the crack and faults, volume change and radius of the crack and opening and strike-slip of the dyke are free parameters. Left column (a, d, g), data; middle column (b, e, h), model; right column (c, f, i), residuals. The data are detailed in Extended Data Fig. 4 and all of the data were used to constrain the model.

  11. Path of dyke propagation from energy considerations.
    Extended Data Fig. 7: Path of dyke propagation from energy considerations.

    ah, Energy profiles for dyke segments 1, 2, 3, 4, 5, 6, 7a and 8a, as described in Fig. 3b. Blue lines indicate the strain energy potential change as a function of the strike, and the red lines the gravitational potential change. Green is the total potential energy change. Energy is shown in terajoules (1012 J). The lowest point on each energy curve is defined as 0 TJ. Error bars, 1 s.d. of the error in the numerical integration.

  12. Topography, earthquake depths, and lithostatic pressure along the dyke path.
    Extended Data Fig. 8: Topography, earthquake depths, and lithostatic pressure along the dyke path.

    a, Bedrock (brown line) and ice topography (light blue line) along the dyke path. b, Depth of earthquake hypocentres (grey dots) below sea level projected on the dyke segments and lines (red) of constant lithostatic pressure, assuming constant crustal density of 2,800 kg m−3 and ice density of 920 kg m−3. Line spacing corresponds to 25 MPa. c, Lithostatic pressure at sea level calculated along dyke segments 1, 2, 3, 4, 5, 6, 7a, 8a and 8b (blue line). The calculations take into account both the subglacial bedrock topography and ice thickness. Light blue triangles indicate the beginning of a segment and red triangles the end of a segment. It is assumed that between segments the dyke propagates along a straight path. Dyke propagation was halted for the longest time at the end of segment 4 (see Figs 2 and 4), before an increase in lithostatic pressure.

References

  1. Wright, T. J. et al. Geophysical constraints on the dynamics of spreading centres from rifting episodes on land. Nature Geosci. 5, 242250 (2012)
  2. Einarsson, P. & Brandsdottir, B. Seismological evidence for lateral magma intrusion during the July 1978 deflation of the Krafla volcano in NE-Iceland. J. Geophys. 47, 160165 (1980)
  3. Buck, W. R., Einarsson, P. & Brandsdottir, B. Tectonic stress and magma chamber size as controls on dike propagation: constraints from the 1975–1984 Krafla rifting episode. J. Geophys. Res. 111, B12404, http://dx.doi.org/10.1029/2005JB003879 (2006)
  4. Belachew, M. et al. Comparison of dike intrusions in an incipient seafloor-spreading segment in Afar, Ethiopia: seismicity perspectives. J. Geophys. Res. 116 (B6), B06405, http://dx.doi.org/10.1029/2010JB007908 (2011)
  5. Gudmundsson, A. Infrastructure and mechanics of volcanic systems in Iceland. J. Volcanol. Geotherm. Res. 64, 122 (1995)
  6. Hartley, M. E. & Thordarson, T. The 1874–1876 volcano-tectonic episode at Askja, North Iceland: lateral flow revisited. Geochem. Geophys. Geosyst. 14, 22862309 (2013)
  7. Segall, P. Earthquake and Volcano Deformation (Princeton Univ. Press, 2009)
  8. Segall, P., Llenos, A. L., Yun, S.-H., Bradley, A. M. & Syracuse, E. M. Time-dependent dike propagation from joint inversion of seismicity and deformation data. J. Geophys. Res. Solid Earth 118, 57855804 (2013)
  9. Björnsson, H. Hydrology of Ice Caps in Volcanic Regions Vol. 45 (Societas Scientarium Islandica, 1988)
  10. Gudmundsson, M. T. & Högnadóttir, T. Volcanic systems and calderas in the Vatnajökull region, central Iceland, constraints on crustal structure from gravity data. J. Geodyn. 43, 153169 (2007)
  11. Björnsson, H. & Einarsson, P. Volcanoes beneath Vatnajökull, Iceland: evidence from radio-echo sounding, earthquakes and jökulhlaups. Jökull 40, 147168 (1990)
  12. Thordarson, T. & Larsen, G. Volcanism in Iceland in historical time: volcano types, eruption styles and eruptive history. J. Geodyn. 43, 118152 (2007)
  13. Einarsson, P. et al. Center of the Iceland hotspot experiences volcanic unrest. Eos 78, 369375 (1997)
  14. Pagli, C. et al. Crustal deformation associated with the 1996 Gjálp subglacial eruption, Iceland: InSAR studies in affected areas adjacent to the Vatnajökull ice cap. Earth Planet. Sci. Lett. 259, 2433 (2007)
  15. Reddy, J. N. An Introduction to Continuum Mechanics (Cambridge Univ. Press, 2013)
  16. Dahm, T. Numerical simulations of the propagation path and arrest of fluid-filled fractures in the Earth. Geophys. J. Int. 141, 623638 (2000)
  17. Maccaferri, F., Bonafede, M. & Rivalta, E. A quantitative study of the mechanisms governing dike propagation, dike arrest and sill formation. J. Volcanol. Geotherm. Res. 208, 3950 (2011)
  18. Fialko, Y. A. & Rubin, A. What controls the along-strike slopes of volcanic rift zones? J. Geophys. Res. 104, 2000720020 (1999)
  19. Toda, S., Stein, R. S. & Sagiya, T. Evidence from the AD 2000 Izu islands earthquake swarm that stressing rate governs seismicity. Nature 419, 5861 (2002)
  20. Staudacher, T. et al. The April 2007 eruption and the Dolomieu crater collapse, two major events at Piton de la Fournaise (La Réunion Island, Indian Ocean). J. Volcanol. Geotherm. Res. 184, 126137 (2009)
  21. Böðvarsson, R., Rögnvaldsson, S. T., Slunga, R. & Kjartansson, E. The SIL Data Acquisition System — At Present and Beyond Year 2000 (Report VÍ-R98005-JA04, Icelandic Meteorological Office, 1998)
  22. Slunga, R., Rögnvaldsson, S. Th. & Bödvarsson, R. Absolute and relative locations of similar events with application to microearthquakes in southern Iceland. Geophys. J. Int. 123, 409419 (1995)
  23. Stefánsson, R. et al. Earthquake prediction research in the South Iceland seismic zone and the SIL project. Bull. Seismol. Soc. Am. 83, 696716 (1993)
  24. Darbyshire, F. A., Bjarnason, I. Th., White, R. S. & Flóvenz, Ó. G. Crustal structure above the Iceland mantle plume imaged by the ICEMELT refraction profile. Geophys. J. Int. 135, 11311149 (1998)
  25. Lomax, A., Virieux, J. J., Volant, P. & Berge, C. in Advances in Seismic Event Location (eds Thurber, C. H. & Rabinowitz, N.) 101134 (Kluwer, Amsterdam, 2000)
  26. Rögnvaldsson, S. T. & Slunga, R. R. Routine fault plane solutions for local networks: a test with synthetic data. Bull. Seismol. Soc. Am. 83, 12321247 (1993)
  27. Herring, T., King, R. W. & McCluskey, S. M. Introduction to GAMIT/GLOBK Release 10.4 (Massachusetts Institute of Technology Technical Report, Cambridge, 2010)
  28. Rosen, P. A., Henley, S., Peltzer, G. & Simons, M. Updated repeat orbit interferometry package released. Eos Trans. AGU 85, 47 (2004)
  29. Kampes, B. Delft Object-oriented Radar Interferometric Software: User’s Manual and Technical Documentation v4.02 (Delft Technical University, 1999)
  30. Werner, C. et al. Gamma SAR and interferometric processing software. Proc. ERS-ENVISAT Symp Vol. 461 (European Space Agency Special Publication, 2000)
  31. Jóhannesson, T. et al. Ice-volume changes, bias-estimation of mass-balance measurements and changes in subglacial water bodies derived by LiDAR-mapping of the surface of Icelandic glaciers. Ann. Glaciol. 54, 6374 (2013)
  32. Magnússon, E., Björnsson, H., Dall, J. & Pálsson, F. Volume changes of Vatnajökull ice cap, Iceland, due to surface mass balance, ice flow, and subglacial melting at geothermal areas. Geophys. Res. Lett. 32, L05504, http://dx.doi.org/10.1029/2004GL021615 (2005)
  33. Goldstein, R. R. & Werner, C. Radar interferogram filtering for geophysical applications. Geophys. Res. Lett. 25, 40354038 (1998)
  34. Goldstein, R., Zebker, H. & Werner, C. Satellite radar interferometry-two-dimensional phase unwrapping. Radio Sci. 23, 713720 (1988)
  35. Chen, C. W. & Zebker, H. A. Two-dimensional phase unwrapping with use of statistical models for cost functions in nonlinear optimization. J. Opt. Soc. Am. 18, 338351 (2001)
  36. Decriem, J. et al. The 2008 May 29 earthquake doublet in SW Iceland. Geophys. J. Int. 181, 11281146 (2010)
  37. Okada, Y. Internal deformation due to shear and tensile faults in a half-space. Bull. Seismol. Soc. Am. 82, 10181040 (1992)
  38. Fialko, Y., Khazan, Y. & Simons, M. Deformation due to a pressurized horizontal circular crack in an elastic half-space, with applications to volcano geodesy. Geophys. J. Int. 146, 181190 (2001)
  39. Mogi, K. Relations between the eruptions of various volcanoes and the deformations of the ground surfaces around them. Bull. Earthq. Res. Inst. Univ. Tokyo 36, 99134 (1958)
  40. Hooper, A. et al. Increased capture of magma in the crust promoted by ice-cap retreat in Iceland. Nature Geosci. 4, 783786 (2011)
  41. Mosegaard, K. & Tarantola, A. Monte Carlo sampling of solutions to inverse problems. J. Geophys. Res. 100, 1243112447 (1995)
  42. Hooper, A. et al. Importance of horizontal seafloor motion on tsunami height for the 2011 Mw = 9.0 Tohoku-Oki earthquake. Earth Planet. Sci. Lett. 361, 469479 (2013)
  43. Árnadóttir, T., Jiang, W., Feigl, K. L., Geirsson, H. & Sturkell, E. Kinematic models of plate boundary deformation in southwest Iceland derived from GPS observations. J. Geophys. Res. 111, B07402, http://dx.doi.org/10.1029/2005JB003907 (2006)
  44. Sturkell, E. & Sigmundsson, F. Continuous deflation of the Askja caldera Iceland, during the 1983-1998 non-eruptive period. J. Geophys. Res. 105, 2567125684 (2000)
  45. Delaney, P. T. & McTigue, D. F. Volume of magma accumulation or withdrawal estimated from surface uplift or subsidence, with application to the 1960 collapse of Kilauea volcano. Bull. Volcanol. 56, 417424 (1994)
  46. Jóhannesson, H. & Sæmundsson, K. Geological Map of Iceland, Tectonics 1:600.000 (Icelandic Institute of Natural History, 2009)
  47. Gudmundsson, M. T., Högnadóttir, Þ., Kristinsson, A. B. & Gudbjörnsson, S. Geothermal activity in the subglacial Katla caldera, Iceland, 1999-2005, studied with radar altimetry. Ann. Glaciol. 45, 6672 (2007)
  48. Yun, S., Segall, P. & Zebker, H. Constraints on magma chamber geometry at Sierra Negra Volcano, Galápagos Islands, based on InSAR observations. J. Volcanol. Geotherm. Res. 150, 232243 (2006)

Download references

Author information

Affiliations

  1. Nordic Volcanological Center, Institute of Earth Sciences, University of Iceland, IS-101 Reykjavík, Iceland

    • Freysteinn Sigmundsson,
    • Elías Rafn Heimisson,
    • Stéphanie Dumont,
    • Michelle Parks,
    • Vincent Drouin,
    • Thóra Árnadóttir,
    • Magnús T. Gudmundsson,
    • Thórdís Högnadóttir,
    • Hildur María Fridriksdóttir,
    • Páll Einarsson,
    • Eyjólfur Magnússon,
    • Bryndís Brandsdóttir,
    • Ásta Rut Hjartardóttir,
    • Rikke Pedersen,
    • Helgi Björnsson &
    • Finnur Pálsson
  2. Centre for the Observation and Modelling of Earthquakes and Tectonics (COMET), School of Earth and Environment, University of Leeds, Leeds LS2 9JT, UK

    • Andrew Hooper &
    • Karsten Spaans
  3. GNS Science, PO Box 30368, Lower Hutt 5040, New Zealand

    • Sigrún Hreinsdóttir
  4. Icelandic Meteorological Office, IS-150 Reykjavík, Iceland

    • Kristín S. Vogfjörd,
    • Benedikt G. Ófeigsson,
    • Gunnar B. Gudmundsson,
    • Kristín Jónsdóttir,
    • Hildur María Fridriksdóttir &
    • Martin Hensch
  5. Canada Centre for Mapping and Earth Observation, Natural Resources Canada, 560 Rochester Street, Ottawa, Ontario K1A 0E4, Canada

    • Sergey Samsonov
  6. Department of Earth Sciences, University of Cambridge, Madingley Road, Cambridge CB3 0EZ, UK

    • Robert S. White,
    • Thorbjörg Ágústsdóttir,
    • Tim Greenfield &
    • Robert G. Green
  7. Department of Geosciences, University of Arizona, Tucson, Arizona 85721, USA

    • Richard A. Bennett
  8. Department of Geosciences, The Pennsylvania State University, University Park, Pennsylvania 16802, USA

    • Halldór Geirsson &
    • Peter C. La Femina
  9. Department of Earth Sciences, University of Gothenburg, SE-405 30 Gothenburg, Sweden

    • Erik Sturkell
  10. Seismology Laboratory, School of Geological Sciences, University College Dublin, Belfield, Dublin 4, Ireland

    • Christopher J. Bean,
    • Martin Möllhoff,
    • Aoife K. Braiden &
    • Eva P. S. Eibl

Contributions

The writing of the paper, and the research it is based on, was coordinated by F.S., A.H., S.H., K.S.V., B.G.Ó. and other members of Icelandic Meteorological Office (IMO) seismic monitoring team. All authors contributed ideas and input to the research and writing of the paper. Modelling of geodetic data was done by A.H., E.R.H. and Th.A. Analysis and operation of continuous GPS sites were carried out by S.H., B.G.Ó., H.M.F, R.A.B., V.D., H.G. and P.C.L. Relative earthquake locations were calculated by K.S.V., seismic data presentation and relative locations were by G.B.G., focal mechanisms were determined by M.H. and single earthquake location determination was led by K.J. and SIL seismic monitoring group. Interferometric analysis was carried out by S.D., K.S., M.P., V.D. and S.S., with a composite digital elevation model prepared by E.M. The combined strain and gravity potential in relation to the dyke propagation was modelled by E.R.H. in consultation with A.H. and F.S. Campaign GPS measurements were carried out by S.D., V. D., M.P., Á.R.H., E.S., F.S. and others. M.T.G. and Th.H. mapped the collapse of the Bárðarbunga caldera and the formation of ice cauldrons over the path of the dyke with aircraft radar profiling. Á.R.H., E.M. and P.E. led mapping of graben formed and of the eruptive fissure as shown here. P.E., B.B. and R.P. contributed to the interpretation of events, and H.B. and F.P. provided bedrock topography and ice thickness for Vatnajökull ice cap. R.S.W., Th.Á., T.G., R.G.G., C.J.B., M.M., A.K.B. and E.P.S.E. contributed and analysed seismic data.

Competing financial interests

The authors declare no competing financial interests.

Corresponding author

Correspondence to:

Author details

Extended data figures and tables

Extended Data Figures

  1. Extended Data Figure 1: Tectonic map showing seismic and geodetic stations. (423 KB)

    Filled uptriangles correspond to continuous GPS stations, open triangles to campaign GPS sites, filled downtriangles correspond to seismic stations, and black ‘stars’ correspond to co-located GPS and seismic stations. Station names for GPS are indicated with four upper-case letters and for seismic stations with three lower-case letters. The different tracks for SAR satellite data are plotted with solid lines for RADARSAT-2, dashed lines for COSMO-SkyMed (CSK) and dotted lines for TerraSAR-X (TSX). Orbit numbers are indicated for CSK and TSX. The red stars correspond to the 2014 eruptive fissures at Holuhraun. Background map shows ice caps (white), central volcanoes (dotted lines), calderas (hatched lines) and fissure swarms (grey shading)46. Names of selected volcanoes are shown, and T indicates Tungnafellsjökull central volcano. Inset, locations of the Eastern Volcanic Zone (EVZ), the Western Volcanic Zone (WVZ), the Northern Volcanic Zone (NVZ) and the South Iceland Seismic Zone (SISZ), with their fissure swarms and central volcanoes; box shows area of the main figure.

  2. Extended Data Figure 2: Map and table of dyke segments defined by seismicity. (599 KB)

    a, Location of dyke segments (red lines labelled with numbers) delineated by relatively relocated earthquakes. The triangles show locations of the nearest seismic stations used to locate the events. Green, stations operated by the Icelandic Meteorological Office (IMO); blue, station operated by the University of Cambridge; and red, station operated by University College Dublin. The two stations on the ice are joint IMO and University of Cambridge stations. All stations telemetered data to IMO. Also shown are ice cauldrons formed (circles), outline of lava flow mapped from radar image on 6 September, and boundaries of graben activated in the dyke distal area (hatched lines). b, The dyke segments. Columns show segment number (Nr), latitude and longitude of beginning (Lat1, Lon1) and end points (Lat2, Long2) of each segment, segment length (L), depth range (D), strike, dip, the RMS value of the deviation (in metres) of the earthquakes from the plane they define, and the number of earthquakes used to define each dyke segment plane (#Eq).

  3. Extended Data Figure 3: Subsidence at Bárðarbunga volcano revealed by aeroplane radar profiling. (607 KB)

    Blue contours show the elevation of the ice surface in the Bárðarbunga caldera on 5 September 2014, about three weeks after the onset of unrest. The data were obtained using aircraft flown 100–150 m above glacier surface, using radar altimetry and submetre differential GPS Omnistar. The system provides 2 m absolute accuracy of surface elevation along the survey profiles47, shown as black dotted lines. The subsidence relative to the pre-unrest ice surface is indicated with coloured shading (key at right). It is greatest in the central part of the caldera, where it had a maximum of 16 m.

  4. Extended Data Figure 4: Geodetic model with a contracting point pressure source, a two-segment-dyke and no a priori constraints. (452 KB)

    The dyke segments are rectangular with uniform opening37 and the model shown is that with maximum probability. Free parameters are: position and volume change of the point pressure source39, and position, strike, dip and opening of the dyke. The panels show from left to right: data (a, d, g), model (b, e, h) and residuals (c, f, i). GPS data in all panels span 15 August to 4 September 2014. The top row shows an interferogram spanning 6 July 2012 to 4 September 2014; the middle row shows an interferogram spanning 2 August to 3 September 2014; the bottom row shows the data from Extended Data Fig. 3 from aeroplane radar profiling, but these data were not used to constrain the model.

  5. Extended Data Figure 5: Geodetic model with a contracting point pressure source, two caldera faults, and a four-segment dyke. (450 KB)

    The dyke is divided into multiple rectangular patches37 and the model shown is that with maximum probability. Lateral dyke position is fixed from relocated seismicity. Position of the point source and faults, volume change of the point source and opening and strike-slip of the dyke are free parameters. Left column (a, d, g), data; middle column (b, e, h), model; right column (c, f, i), residuals. The data are detailed in Extended Data Fig. 4 and all of the data were used to constrain the model.

  6. Extended Data Figure 6: Geodetic model with a contracting flat-topped chamber, two caldera faults, and a four-segment dyke. (451 KB)

    A deflating penny-shaped crack38 is used to represent the top of a flat-topped chamber48, and the dyke is divided into multiple rectangular patches37. The model shown is that with maximum probability. Lateral dyke position is fixed from relocated seismicity. Position of the crack and faults, volume change and radius of the crack and opening and strike-slip of the dyke are free parameters. Left column (a, d, g), data; middle column (b, e, h), model; right column (c, f, i), residuals. The data are detailed in Extended Data Fig. 4 and all of the data were used to constrain the model.

  7. Extended Data Figure 7: Path of dyke propagation from energy considerations. (520 KB)

    ah, Energy profiles for dyke segments 1, 2, 3, 4, 5, 6, 7a and 8a, as described in Fig. 3b. Blue lines indicate the strain energy potential change as a function of the strike, and the red lines the gravitational potential change. Green is the total potential energy change. Energy is shown in terajoules (1012 J). The lowest point on each energy curve is defined as 0 TJ. Error bars, 1 s.d. of the error in the numerical integration.

  8. Extended Data Figure 8: Topography, earthquake depths, and lithostatic pressure along the dyke path. (333 KB)

    a, Bedrock (brown line) and ice topography (light blue line) along the dyke path. b, Depth of earthquake hypocentres (grey dots) below sea level projected on the dyke segments and lines (red) of constant lithostatic pressure, assuming constant crustal density of 2,800 kg m−3 and ice density of 920 kg m−3. Line spacing corresponds to 25 MPa. c, Lithostatic pressure at sea level calculated along dyke segments 1, 2, 3, 4, 5, 6, 7a, 8a and 8b (blue line). The calculations take into account both the subglacial bedrock topography and ice thickness. Light blue triangles indicate the beginning of a segment and red triangles the end of a segment. It is assumed that between segments the dyke propagates along a straight path. Dyke propagation was halted for the longest time at the end of segment 4 (see Figs 2 and 4), before an increase in lithostatic pressure.

Supplementary information

PDF files

  1. Supplementary Information (15.1 MB)

    This file contains Supplementary Figures 1-8 and Supplementary Tables 1-3.

Additional data