Scientific Report 914
This report is also available as a PDF document .
Abstract
This loan comprised 40 SEIS-UK seismometers operating alongside 10 Cambridgeowned seismometers to provide coverage of 150 km of the Northern Volcanic Zone (NVZ) active spreading centre in Iceland, with denser local arrays around Askja and Krafla volcanoes, during July 2010 September 2011 (Fig. 1). The loan of 30 SEIS-UK seismometers was extended to September 2012 to enable monitoring of increased seismic activity under Vatnaj kull at the south end of the NVZ. We purchased a further 15 seismometers using NERC funding, thus maintaining a network of 55 instruments during the extension period. The work was supported by a NERC research grant and NERC post-doc and Ph.D studentships as well as several M.Sci. research projects. Preliminary results include: mapping of several thousand earthquakes along the rift zone in both the lower and upper crust; tomographic evidence for the magma chamber under Askja; detailed analysis of how strike-slip faulting accommodates shear in the transfer zone between rift segments; evidence for strongly non-double-couple source mechanisms at Krafla; demonstration of faulting in the brittle layer at extremely low stress drops of 13 bars, probably caused by migration of carbon dioxide from an underlying magmatic intrusion.
Background
The Northern Volcanic Zone (NVZ) of Iceland is persistently seismically active, ultimately due to spreading of ~2 cm/yr being accommodated across its rift segments. However, the permanent monitoring network of the Icelandic Meteorological Office (IMO) is relatively sparse in the NVZ. Our network has provided far denser coverage around Askja (last eruption 1961) and Krafla (last eruption episode 1975-1984) volcanoes, with additional stations along the length of the NVZ beyond and between these two most recently active volcanic centres (Fig. 1). Increased activity under Vatnajökull, at the southern end of the network, is of particular interest because several sub-glacial volcanoes there have the potential to form both hazardous jökulhlaups (glacial floods) and ash-rich plumes as a result of ice-magma interaction, exemplified by Eyjafjallajökull volcano in 2010. Monitoring and understanding the magmatic plumbing systems of Askja (which erupted explosively in 1875 with catastrophic consequences for much of east Iceland) and Krafla (which has a significant commercial geothermal power plant operating in its caldera) remains very important.
Survey procedure
Loan 914 aggregated instruments that were already deployed in Iceland under Loans 857 (15x 16 Gb Güralp 6TDs around Askja) and 891 (15x 8 Gb 6TDs around Krafla caldera). Additional to these 30 instruments, ten more 16 Gb 6TDs were deployed in July 2010 under Loan 914. Ten Cambridge-owned 4 Gb instruments were also already in the field, giving a total network of 50 instruments in July 2010. In July 2011, we deployed 10 new Cambridge-owned 16 Gb 6TDs and five 16 Gb ESPDs (network total = 65 instruments). Subsequently, in September 2011, we returned 10 SEIS-UK 6TDs (8 Gb), whilst 30 SEIS-UK 6TDs and all Cambridge-owned instruments remained deployed under the Loan 914 extension (network total = 55 instruments). The Loan 914 period ended in September 2012. The network was serviced in July/August 2010, March (Krafla only)/July/September 2011, and April (Askja only)/July 2012, with some redistribution of stations in July/September 2011 and July 2012. Station locations are given in the Appendix and Figure 1.
Seismometers were either buried directly in soil or on cement plinths inside barrel vaults. We recorded at 100 or 200 samples per second (sps) during the shorter summer service interval, and 50 sps over winter, which allowed us to record continuously all year without exceeding the 16 Gb sensors’ memory capacity. Stations were powered by a combination of solar panels (≥120 W for sites expected to run all winter), batteries (≥230 Ah, often ≥345 Ah) and, in some cases, wind turbines. Use of wind turbines was only partly successful due to ice accumulation and very strong winds during the Icelandic winter. Nevertheless, power loss over winter was limited by the generous use of batteries and solar panels, which were mounted on wooden A-frames at least 50 cm above the surface. Equipment failure was minimised compared to previous campaigns by mounting breakout boxes on stakes above ground such that they did not become waterlogged and frozen in ice.
Data quality
The majority of local events recorded in the NVZ are of only small magnitudes (ML < 2). Nonetheless, many local events have clear P- and S-wave arrivals (Fig. 2), and regional events of ML≥2 can be observed across the entire network along the length of the NVZ.
Processing and modelling
Data were quality controlled in the field and converted to miniseed format using SEIS-UK standard procedures. An initial catalogue of automatic earthquake locations has been obtained using Coalescence Microseismic Mapping (CMM) software (Drew et al., 2013). We have refined earthquake locations using manual P- and S-wave arrival picks and investigated source mechanisms using P-wave first-motion polarities and P/S amplitude ratios for simple double-couple source inversions as well as full moment tensor solutions (Green et al., in press; Watson et al., in prep.). Manual Pand S-wave arrival time picks have been used to carry out tomography around Askja (Mitchell et al., 2013).
Preliminary findings
During the loan period, the network has recorded several thousand events at Krafla and Askja, including deep crustal events (Fig. 3). At Krafla, all well-constrained earthquakes are shallow (< 4 km) and coincide spatially with the location of either geothermal fields associated with the power station in the caldera, or are located above, or close to, the putative magma chamber. At Askja, most activity is also at relatively shallow levels, above the brittle-ductile transition (~6-7 km), but we also observe much deeper activity in the normally ductile mid- and lower crust which we attribute to melt movement.
Interpretation to date
We interpret all of the seismicity deeper than approximately 8 km as being associated with magma migration and intrusion in the crust (Key et al., 2011; White et al., 2011; White et al., 2012), where the high strain rates cause brittle failure in otherwise ductile crust. Shallower earthquakes are likely due to hydrothermal fluid circulation driven by heat from magma chambers, geothermal exploitation and tectonic stresses across the rift zone. Preliminary tomographic results suggest that there is a magma chamber at ~6-8 km depth beneath Askja (Fig. 4; Mitchell et al., 2013). We have found evidence (in the earthquake distribution and fault-plane solutions) for spreading being accommodated by bookshelf faulting in the transfer zone between the Askja and Kverkfjöll rift segments (Fig. 5; Green et al., in press). At Krafla, we observe well-constrained, strongly non-double-couple earthquakes that we interpret as being associated with the evacuation and collapse of fluid-filled cavities (Watson et al., in prep.). Stress modelling and the timing of earthquakes around a mid-crustal dyke intrusion at Upptyppingar (Fig. 4) suggest that subsequent, shallower earthquakes are triggered by volatiles that have exsolved from the magma in the mid-crustal dyke (Martens & White, 2013).
Conclusions and Recommendations
The network deployed along the NVZ with a focus on Askja and Krafla volcanoes has enabled a significantly lower detection threshold than the permanent national monitoring network operated by the IMO. The dense array has facilitated detailed analysis of earthquake locations, tomography and source mechanisms, thus allowing us to image magmatic intrusions and speculate on source processes and the relationship between magmatism and the seismicity we have recorded. The rich dataset acquired under Loan 914 is the subject of ongoing study, including denser travel time tomography around Askja and Krafla, investigation of anisotropy caused by cracks in the upper crust and in the Krafla geothermal area, regional crustal velocity structure control by ambient noise analysis, receiver function analyses of Moho and mantle discontinuities, ongoing temporal changes in seismicity caused by the 2007 Upptyppingar intrusion, and the use of regional earthquakes for crustal structure studies along the NVZ.
Publications
Green. R. G., White. R. S., Greenfield. T. (2013), Motion in the north Iceland rift zone accommodated by bookshelf faulting. Nature Geoscience, in press. (Loan 914).
Key, J., White, R. S., Soosalu, H. & Jakobsdóttir, S. S. (2011). Multiple melt injection along a spreading segment at Askja, Iceland. Geophysical Research Letters, 38, L05301, doi:10.1029/2010GL046264 (Loans 842, 857 & 914)
Key, J., White, R. S., Soosalu, H. & Jakobsdóttir, S. S. (2011). Correction to “Multiple melt injection along a spreading segment at Askja, Iceland”, Geophysical Research Letters, 38, L10308, doi:10.1029/2011GL047491 (Loans 842, 857 & 914)
Martens, H. R. & White, R. S. (2013). Triggering of microearthquakes in Iceland by volatiles released from a dyke intrusion, Geophysical Journal International, 194 (3), 17381754, doi: 10.1093/gji/ggt184 (Loan 914)
Mitchell, M., White, R. S., Roecker, S. & Greenfield, T. Tomographic image of melt storage beneath Askja volcano, Iceland using local microseismicity, Geophysical Research Letters, in press. (Loans 857 & 914)
Soosalu, Heidi, Key, Janet & White, Robert S. (2011), Laatikollisesta varaseismometrejä rift-alueen alakuoren synnyn jäljille, [From a box of spares to tracking down lower crust generation at a rift],_ Geologi_, 63, 3643. (Loans 842, 857 & 914)
Watson, Z., Tarasewicz, J., Pugh, D., White, R. S. and Brandsdóttir, B., (in prep.), Non-double-couple earthquakes at Krafla volcano, Iceland, Geophysical J. Int., (Loans 891 & 914)
White, Robert S., Redfern, Simon A. T. & Chien, Su-Ying (2012). Episodicity of seismicity accompanying melt intrusion into the crust, Geophysical Research Letters, 39, L08306, doi:10.1029/2012GL051392 (Loan 914)
Dissertations
Green, Robert (2012) Swarm microseismicity and upper crustal faulting between the Askja and Kverkfjöll volcanic systems, Iceland M.Sci. dissertation, University of Cambridge, U.K. (Loan 914).
Greenfield, T. (2011), Microseismicity of the Krafla Volcanic System, Iceland, M.Sci. dissertation, University of Cambridge, U.K. (Loan 914).
Key, Janet (2011) Tracking melt with lower crustal earthquakes at Askja, Iceland, PhD Dissertation, University of Cambridge, U.K. (Loans 842, 857 & 914)
Mitchell, Michael (2011) 3-D Tomographic Inversion of Local Microseismic Events to Image the Askja Magma Chamber, M.Phil. dissertation, University of Cambridge, U.K. (Loans 842, 857 & 914)
Watson, Z. (2013), Microseismicity within the Krafla Caldera, NE Iceland, M.Sci. dissertation, University of Cambridge, U.K. (Loans 891 & 914)
Conferences (25 abstracts)
Work based on Loan 914 has also been presented at:
American Geophysical Union Conference, San Francisco, USA, 2012; 2013;
Magmatic Rifting and Active Volcanism Conference, Addis Ababa, January 2012;
Volcanic and Magmatic Studies Group, Cambridge, 2011; Bristol 2013;
British Geophysical Association Conference, Cambridge, U. K., Sept 2013;
AGU Chapman Conference, Hawaii, 2012;
European Seismological Commission, Salina 2011; El Hierro 2012; Sulawesi 2013;
EGU General Assembly, Vienna, 2011.
Appendix
