Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Reviewed research article Seismicity in Iceland: 1994–2007 Steinunn S. Jakobsdóttir Physics Department, Icelandic Meteorological Office, Bústaðavegur 9, Reykjavík, IS-150, Iceland [email protected] Abstract — Since the end of 1993, the digital, automatic seismic system, known as the SIL-system, has acted as the national seismic network in Iceland. The number of stations in the network has increased gradually over the 14 years from 1994 to 2007. The detection capability of the system has increased at the same pace, especially along the volcanic zones. The general seismicity of Iceland, as monitored by the growing network, is presented in this paper. The main activity is, as expected, along the plate boundary, but seismicity has also been observed at intraplate locations. Instead of setting a lower limit to the size of earthquakes to be considered in this paper, changes in the sensitivity are discussed along with the seismic activity. Several major events have been observed during the observation period. In June 2000, two magnitude 6.6 earthquakes occurred in the South Iceland Seismic Zone, followed by a few earthquakes larger than magnitude 5. A few episodes caused by magma injections were recorded and four volcanic eruptions monitored. In addition to the large earthquakes in June 2000, seven earthquakes larger than magnitude 5 occurred in this period, two in Bárðarbunga in the years before the 1996 Gjálp eruption, two near the triple junction between the Reykjanes Peninsula, the Western Volcanic Zone and the South Iceland Seismic Zone, two in the Tjörnes Fracture Zone and one on the Reykjanes Peninsula. Intraplate earthquakes were recorded in the northwestern part of Iceland, Vestfirðir, in 1994 and 2006, and in Guðlaugstungur, between Langjökull and Hofsjökull, in 2004. Additionally, earthquakes are located near Surtsey every year. Two episodes of activity have been recorded in the Esjufjöll central volcano in the southeastern part of Vatnajökull, and several swarms have been detected under Öræfajökull. All the main ice caps cover seismically active central volcanoes, explaining the high seismicity beneath them. INTRODUCTION In 1991, a new era of seismic monitoring started in Iceland, when a new digital seismic system, the SIL system, was upgraded to fully automatic operation. The SIL seismic system is a highly automated threecomponent data acquisition and monitoring system, consisting of the SIL network and automatic processing software. It was designed and initiated within the Nordic SIL project in 1988 - 1994 (Stefánsson et al., 1993, Böðvarsson et al., 1996, Böðvarsson et al., 1999). The motivation for the project was the imminent occurrence of large earthquakes in the South Iceland Lowland (SIL). The design goal for the system was to automatically detect and locate earthquakes as small as magnitude Ml 0 or less for monitoring purposes, and in order to get high quality data for earthquake prediction research. The network was originally installed and operated in SW-Iceland in the South Iceland Seismic Zone (Figure 1). It was expanded to the Tjörnes Fracture Zone in 1993, whereby it became a regional network for all Iceland, covering the main seismic zones. In the following years, the network was expanded gradually into the highland along the rift zone. In 1994 the number of stations was 18, but at the end of 2007 the seismic network consisted of 51 stations (Figure 1). JÖKULL No. 58, 2008 75 S. S. Jakobsdóttir Figure 1. The SIL network. Stations are marked with triangles and color coded according to age. Inverted triangles represent stations that are no longer operated. Boxes mark the areas plotted in Figures 4 and 8. Abbreviations are as follows: A: Askja, AF: Arnarfjörður, BB: Bárðarbunga, DE: Dyrhólaey EE: Eldey, EF: Esjufjöll, EJ: Eyjafjallajökull, EVZ: Eastern Volcanic Zone, GE: Grímsey, GJ: Geitlandsjökull, GR: Grímsvötn/Grímsfjall, GS: Geirfuglasker, GT: Guðlaugstungur, H: Hekla, HT: Herðubreiðartögl, HV: Hveravellir, KF: Kistufell, KR: Kolbeinsey Ridge, KvF: Kverkfjöll, LH: Lokahryggur, MJ: Mýrdalsjökull, NVZ: Northern Volcanic Zone, RF: Reykjarfjörður, RP: Reykjanes Peninsula, RR: Reykjanes Ridge, SISZ: South Iceland Seismic Zone, SE: Surtsey, TFZ: Tjörnes Fracture Zone, TJ: Torfajökull, UT: Upptyppingar, WVZ: Western Volcanic Zone ÞJ: Þórisjökull, ÖJ: Öræfajökull. – SIL-netið. Jarðskjálftastöðvarnar eru sýndar með þríhyrningum og aldur þeirra er táknaður með litum. Þríhyrningar með eitt horn niður tákna stöðvar sem ekki eru lengur í rekstri. Opnir ferningar skilgreina svæðin sem myndir 4 og 8 byggja á. Skammstafanir eru útskýrðar í enska textanum. This paper is based on data recorded by the SIL system during the years 1994 to 2007. Its aim is to give an overview of the seismic activity during this period. The focus is kept on the seismicity, both the everyday activity and on anomalous swarms and sequences. Less attention is paid to structures derived from earthquake studies or other studies of the seis- 76 JÖKULL No. 58, 2008 mic zones. The depth distribution of earthquakes is not discussed in detail, as more work is needed to obtain well-constrained depth determinations. In general, the focal depth of earthquakes in Iceland is less than 8–12 km, indicating that this is the thickness of the brittle crust. Focal depths down to 25–30 km are observed, but they are not always well-constrained as Seismicity in Iceland: 1994–2007 the closest stations in many cases are too far from the epicenter, and the one-dimensional velocity models do not show enough detail. The occurrence of fairly well-constrained deep earthquakes is discussed in this paper and some results are given from other studies, including depth determinations obtained by mapping of active faults by relative location of swarms. During the study period there have been four volcanic eruptions in Iceland and a few magmatic episodes that did not reach the surface. The seismic manifestation of the eruptions and magma movements is also described here. Over ∼250,000 earthquakes in the magnitude range -2 to 6.6 were recorded and located during this period. TECTONIC SETTING In summary, the tectonic setting of Iceland is characterized by its position on the Mid-Atlantic Ridge (MAR) and by the Icelandic Hotspot (Tryggvason et al.1983; Wolfe et al., 1997; Allen et al., 2002; Bjarnason 2008). The MAR is shifted to the east in the vicinity of the mantle plume, which is centered under Vatnajökull ice cap (Figure 1). The largest earthquakes, of magnitudes up to 7.2 (Stefánsson and Halldórsson, 1988), occur in the South Iceland Seismic Zone (SISZ), between the Western Volcanic Zone (WVZ) and the Eastern Volcanic Zone (EVZ), and in the Tjörnes Fracture Zone (TFZ), connecting the Northern Volcanic zone (NVZ) to the Kolbeinsey Ridge (KR). The earthquake faults in the SISZ strike N-S, lining up parallel, almost perpendicular to the underlying plate boundary. The TFZ is more complex, composed of three WNW-ESE trending segments (Sæmundsson 1974; Einarsson 1991; Rögnvaldsson et al., 1998a). Volcanic eruptions are frequent in the NVZ and the EVZ. Eruptions are generally preceded by seismic activity that differs from one volcano to another. Since 1970, Hekla volcano is known to be seismically quiescent until about an hour before an eruption while most other volcanoes have a longer preparation period. For a more detailed study of the tectonics of Iceland, see Einarsson (2008). THE SIL SYSTEM At each station, the incoming data-stream is preprocessed by a computer at the site. Transients in the data flow are detected and defined by onset time, amplitude and duration. Different parameters are calculated, such as apparent velocity, azimuth and spectral parameters. This information is packed into short messages and sent to the central computer in Reykjavík, where an automatic phase association process defines events and sends requests for wave-form data to the stations (Böðvarsson et al., 1996). Automatic locations and magnitudes of earthquakes are usually available within 1-3 minutes of their occurrence and focal mechanisms are determined for every earthquake using spectral amplitudes (Slunga, 1981; Rögnvaldsson and Slunga, 1993, 1994). All automatic locations are manually checked, and both locations and focal mechanisms are re-run with corrected phase picks. A one minute average in three frequency intervals is calculated automatically for all three components of the seismic data and plotted continuously on-line in so called tremor plots (Böðvarsson et al., 1999). These plots are used to monitor changes in background noise and seismic activity. A relative location method is available in the system. It is used to obtain precise locations, within tens of meters in relative accuracy, thereby making it possible to map active faults (Rögnvaldsson and Slunga, 1994; Slunga et al., 1995). There are two types of automatic magnitude calculations used in the SIL-system. The local magnitude, Ml , is empirical, based on the maximum peakto-peak amplitude in a 10 seconds interval around the S-wave at all stations: Ml = log10 A + 2.1 * log10 D - 4.8 where A is the maximum velocity amplitude of highpass filtered waveforms with a cut off frequency at 2 Hz and scaled to the response of Lennartz 1 Hz sensor and Nanometrics RD3 digitizer, D is the distance between the earthquake and the station in km. It is calibrated to give similar magnitude values as the magnitudes calculated from peak-to-peak readings of the former analog regional network. The other magnitude is a local moment magnitude, called Mlw or Ml (Mo ): JÖKULL No. 58, 2008 77 S. S. Jakobsdóttir Ml (Mo ) = log10 (Mo ) – 10 where Mo is the seismic moment in Nm. This formula is used for Mlw <2 and is slightly modified for larger earthquakes (Stefánsson et al., 2005). In this work the empirical local magnitude Ml is mostly used. In 1994 the SIL system was complete in detection of earthquakes in Iceland down to magnitude ∼Ml 2. In the SISZ earthquakes down to Ml -1.5 were detected and located, but the magnitude of completeness in the area was ∼Ml 0 (Wyss and Stefánsson 2006). Earthquakes down to magnitude ∼Ml 0.5 – Ml 1 were detected and located in TFZ and down to ∼Ml 1,5 in the highland. Since 2006 most of the highland is well within the Ml 1 detection limit and only in the NW part of the country, at Vestfirðir, is the detection limit higher than Ml 1.5 (Ágústsson, 2006). had a magnitude larger than Ml 2, which is similar to the magnitude threshold used in former bulletins, based on analog recordings. Figure 3 shows the epicenter distribution of earthquakes larger than Ml 1. The seismic activity is mainly concentrated in the seismic zones. More than 80% of the epicentres are located in the TFZ and in the area from SISZ through the Hengill triple junction and to the Reykjanes Peninsula (RP). Earthquakes, detected by the network, occur every week in these zones and every year in the connecting volcanic zones, mainly at volcanic centers. In addition, activity is recorded every year near to Surtsey, south of Iceland. THE DATASET Seismicity as presented here is based on automatic detections and locations of earthquakes that have been manually corrected for errors. In some cases, earthquakes that were not detected automatically have been manually added. These are usually either earthquakes that are close in time, or undetected earthquakes that can be seen on the continuous tremor plots. The earthquakes that escape automatic detection are mostly so called LP (long period) earthquakes, characterized by emergent onsets and dominated by low frequency and can even be difficult to locate manually. They occur mainly in volcanic areas, e.g. under the western part of Mýrdalsjökull-Katla. Explosions due to construction work are frequently detected, but care has been taken to mark them as such and to exclude them in this study. It is still possible that a few explosions have passed through as earthquakes, mainly in areas that are known to be seismically active. Results from relative locations are not used directly in this work. GENERAL SEISMICITY IN ICELAND 1994–2007 Several seismic swarms and volcanic events have been observed during this period. Over 250,000 earthquakes were located within Iceland and the surrounding offshore shelf (Figure 2) and about 10,000 of these 78 JÖKULL No. 58, 2008 Figure 2. Number of earthquakes recorded by the SIL system during 1994–2007, within the co-ordinates 63◦ N–67◦N and 13◦ W–25◦ W. Number of earthquakes larger than Ml 0, Ml 1 and Ml 2 are shown with colors. – Árlegur fjöldi jarðskjálfta, sem skrásettir voru í SIL-kerfinu frá 1994 til 2007 innan svæðis sem afmarkast af hnitunum 63◦ N–67◦ N og 13◦ V– 25◦ V. Fjöldi skjálfta af stærð yfir Ml 0, Ml 1 og Ml 2 er sýndur með mismunandi litum. In the TFZ the Grímsey lineament and the Húsavík-Flatey fault are most active (Stefánsson et al., 2008). Two earthquakes of magnitude ∼ M5–5.5 occurred in the TFZ during the study period. A Mw 5.5 earthquake (Table 1) occurred on 8 February 1994 in the western part of the zone and a Mw 5.8 (Table 1) in the northern part on 16 September 2002. Seismicity in Iceland: 1994–2007 Magnitude ∼M 5 earthquakes were recorded yearly by the SIL system prior to the 1996 eruption in Vatnajökull, a magnitude b 5.7 earthquake on 5 May 1994, a magnitude mb 4.9 on 11 December 1995 and a magnitude Mw 5.6 on 29 September 1996 (Table 1), marking the onset of the eruption. In the Hengill area around the triple junction between RP, WVZ and SISZ, intense activity was observed in 1994 to 1998. This activity culminated with two Ml 5 earthquakes in 1998. Two earthquakes of magnitudes Ms 6.6 (Table 1) occurred in the SISZ in 2000, with thousands of aftershocks in the following days and months. On 23 August 2003 a magnitude Mw 5.2 was located at Sveifluháls, near Kleifarvatn on the Reykjanes Peninsula. Figure 3 shows many events located in Skeiðarárjökull, an outlet glacier south of Vatnajökull. These events are classified as icequakes, as they are shallow and are associated with increased water flow through the glacier (Roberts, 2005). The icequakes will not be discussed in this paper; for further details, see Roberts (2005). Comparing the map in Figure 3 with the seismicity map of Einarsson and Sæmundsson (1987) shows that all areas that were active in the former period (1982–1985) are also active in the study period, ex- Figure 3. Map of earthquakes larger than Ml 1 recorded by the SIL system in 1994–2007. Earthquakes larger than M 5 are marked with stars. – Kortið sýnir jarðskjálfta stærri en Ml 1, sem skráðir voru í SIL-kerfinu á árunum 1994 til 2007. Stjörnurnar tákna skjálfta stærri en M 5. JÖKULL No. 58, 2008 79 S. S. Jakobsdóttir Table 1. Earthquakes with magnitude larger than 5 from the NEIC website. Locations and magnitudes from the SIL system are in the last four columns. – Skjálftar sem mælst hafa yfir 5 stig samkvæmt alþjóðlegu neti skjálftamæla (NEIC). Staðsetningar og reiknaðar stærðir samkvæmt SIL-netinu eru í fjórum seinustu dálkunum. YEAR MO DA TIME LAT LONG km Mb Me Ms Mw SIL-LAT SIL-LONG Mlw Ml 1994 1994 1995 02 05 12 08 05 11 032754.90 66.51 051449.77 64.58 052246.63 64.55 -19.22 -17.48 -17.79 10 5.3 8 5.7 10 4.9 5.3 5.2 4.5 5.50 HRV 66.451 64.634 64.671 -19.249 -17.430 -17.504 4.7 4.6 4.6 4.8 4.2 3.9 1996 1998 1998 2000 09 06 11 06 29 04 13 17 104818.97 213654.19 103834.50 154041.73 64.78 63.993 63.882 63.966 -17.57 -21.290 -21.595 -20.487 10 10 10 10 5.3 5.1 4.9 5.7 5.4 5.1Z 4.4Z 6.6Z 5.60 HRV 5.40 HRV 5.10 HRV 6.80 CSEM 6.50 HRV 64.687 64.036 63.963 63.973 -17.495 -21.290 -21.352 -20.367 4.8 5.0 5.1 5.5 4.4 4.7 4.9 5.7 2000 2000 2002 2003 06 06 09 08 17 21 16 23 154252.63 005146.88 184826.72 020011.89 63.965 63.980 66.94 63.92 -20.633 -20.758 -18.46 -22.26 10 10 10 10 5.7 6.1 6.8 6.6Z 5.5 5.7 5.80 GS 4.8 4.5 5.20 HRV 63.943 63.972 66.975 63.905 -20.460 -20.711 -18.437 -22.085 5.8 5.8 5.3 5.0 5.0 5.4 5.0 4.3 cept for a unique swarm occurring about 20 km south of Dyrhólaey in July 1985. Several new active areas are seen in 1994–2007. A few earthquakes were recorded at Vestfirðir in 1994 and in 2006. A few sequences have been located north of Langjökull and earthquake swarms were recorded at Guðlaugstungur, NE of Langjökull, in 2004. These earthquakes are all outside the present rift zones and transform faults. Hofsjökull has three distinct seismic areas that are active at different times and Eyjafjallajökull has had some earthquake episodes. Activity on the Reykjanes Ridge near Geirfuglasker and Eldeyjarboði has increased since 2000 with several Ml 4 earthquakes (Höskuldsson et al., 2007). The last big earthquake swarms occurred on the Reykjanes Ridge in February and April 2007. The station density has increased considerably during the period discussed in this paper. An overview of the evolution of the network with time is seen in Figure 1, where the time of installation is color coded. SEISMIC ACTIVITY IN THE HIGHLANDS Most of the observed activity in the highlands is under or in the vicinity of ice caps. Each ice cap covers 80 JÖKULL No. 58, 2008 one or more central volcanos. The Askja central volcano and its surroundings is the only seismically active area that is not close to an ice cap. Earthquakes in the highland are of magnitudes up to Ml 3–4, except under Bárðarbunga in Vatnajökull where earthquakes of magnitudes up to Ml 4.4 (Mw 5.6, Table 1) were observed before the Gjálp eruption in 1996 (Figure 4h). Langjökull The main seismic activity in the western part of the highlands is the area of the Langjökull ice cap in the WVZ. In the first half of the period, before 2000, most of the activity was beneath the southern part of the ice cap, with swarms occurring in Geitlandsjökull and in Þórisjökull (Figures 3 and 4). A swarm was recorded at Kaldidalur, west of Geitlandsjökull, in 1994. The most active period started with a swarm in Geitlandsjökull in March 1998 and proceeded with swarms in Þórisjökull in September 1998 and July 1999. The Mw 6.6 earthquake in Holt in June 2000 was followed by seismic activity in other areas near Langjökull. Earthquakes were recorded near Geysir, due north of the mainshock, and the activity continued north of Sandfell in a ∼15 km long zone south of the ice cap (Hjaltadóttir and Vogfjörð, 2005). Fur- Seismicity in Iceland: 1994–2007 Figure 4. Magnitudes greater than Ml 1 as a function of time for areas shown by boxes in Figures 1, 5 and 6. Note that scales are different on different plots. a: Langjökull, b: Hofsjökull, c: Torfajökull, d: Goðabunga, e: Katla caldera, f: Lokahryggur, g: Grímsvötn, h: Bárðarbunga, i: Kistufell, j: Kverkfjöll, k: Herðubreiðartögl, l: Askja. For Goðabunga and Katla caldera magnitudes larger than 1.8 are plotted. The gray line shows the accumulated number of earthquakes during the period. – Stærð jarðskjálfta stærri en Ml 1 sem fall af tíma. Svæðin eru merkt inn á myndir 1, 5 og 6 og eru eftirfarandi: a: Langjökli, b: Hofsjökli, c: Torfajökli, d: Goðabungu, e: Kötluöskjunni, f: Lokahrygg, g: Grímsvötnum, h: Bárðarbungu, i: Kistufelli, j: Kverkfjöllum, k: Herðubreiðartöglum, l: Öskju. Athugið að ekki er sami skali á öllum myndunum. Einnig eru einungis sýndir skjálftar stærri en Ml 1,8 í Goðabungu og Kötluöskjunni. Gráa línan sýnir uppsafnaðan fjölda skjálfta yfir tímabilið. JÖKULL No. 58, 2008 81 S. S. Jakobsdóttir 82 JÖKULL No. 58, 2008 Seismicity in Iceland: 1994–2007 JÖKULL No. 58, 2008 83 S. S. Jakobsdóttir 84 JÖKULL No. 58, 2008 Seismicity in Iceland: 1994–2007 ther north, at the NE corner of Langjökull and NW of Hveravellir, another area was activated. These areas have all been active since 2000. There is decreased activity in Geitlandsjökull and Þórisjökull after the 17 June 2000 earthquake, but in 2007 the activity at the NE flank of Langjökull has increased. In addition, small swarms have occasionally occurred both north, west and south of Langjökull during the study period. Hofsjökull Beneath the Hofsjökull ice cap three separate areas were seismically active during the study period (Figures 3 and 4b). Small swarms occur regularly beneath the western part of the ice-filled caldera in the western part of Hofsjökull and this is the most persistent activity beneath the ice cap. In the northern part, the activity is more sporadic. A dense swarm of 30 earthquakes was located in the NE part of the ice cap in November 1999. Other activity in the NE part has been observed as single earthquakes, some events might be icequakes. Torfajökull There is persistent seismic activity within the Torfajökull volcanic center with magnitudes up to Ml ∼3 (Figures 3 and 4c). Earthquakes are recorded regularly with occasional occurrences of small swarms. Soosalu et al. (2006b) and Lippitsch et al. (2005) studied the seismicity within Torfajökull area with data from a temporary local network they operated in 2002, in addition to SIL data. They found that high frequency earthquakes occur at shallow depths (less than 6 km) in the western part of the Torfajökull caldera while low frequency earthquakes with emergent onsets cluster in the southern part, near the glaciers (Figures 1 and 3), at depths reaching from close to the surface down to 14 km. Mýrdalsjökull-Katla The seismic activity in Mýrdalsjökull is mostly at two sites, within the Katla caldera and at Goðabunga on the western flank of the ice cap (Figure 5). The number of earthquakes located beneath Goðabunga is about 5 times the activity within the caldera (Figures 4d and e) (Jakobsdóttir et al. 2002; Þorbjarnardóttir et al., 2003). The seismicity in Goðabunga shows seasonal variation and is most intense in the second half of the year (Einarsson and Brandsdóttir, 2000; Jónsdóttir et al., 2007). The earthquakes are shallow with emergent onsets and can be difficult to locate with precision. Einarsson and Brandsdóttir (2000) argue that the seasonal variation is mainly due to change in hydrostatic pressure associated with the seasonal melting of the ice cap. An episode of unrest was observed beneath Goðabunga from the end of 2001 to the end of 2004. The seasonal variation in seismicity in this period was overlain by constant activity resulting in four times higher seismic activity than the years before and after (Þorbjarnardóttir and Guðmundsson, 2003; Guðmundsson et al., 2004; Hjaltadóttir et al., 2005a). Earthquakes within the caldera are of tectonic origin and are more uniform in time. JÖKULL No. 58, 2008 85 S. S. Jakobsdóttir Figure 5. Earthquakes in Eyjafjallajökull and Mýrdalsjökull larger than Ml 1 in 1994–2007. The boxes represent (from left to right) Eyjafjallajökull, Goðabunga and Katla caldera. – Kortið sýnir jarðskjálfta stærri en Ml 1 undir Eyjafjallajökli og Mýrdalsjökli á árunum 1994 til 2007. Ferningarnir afmarka svæðin sem skilgreind eru sem Eyjafjallajökull, Goðabunga og Kötluaskja á myndum 4 og 8. Uplift of the Katla caldera was observed by geodetic surveying during the period of unrest, but uplift at Goðabunga is less prominent (Sturkell et al., 2006a; Sturkell et al., 2008). Soosalu et al. (2006a) suggested a cryptodome under Goðabunga as the cause of the activity. Since 2005 the number of earthquakes per year is again similar to the activity in 1998–1999. Jónsdóttir et al. (2007) observe that the characteristics of earthquakes in the caldera are similar to volcanic areas, while the activity under Goðabunga is highly anomalous. They find annual variation in both areas, but it is more prominent under Goðabunga where the maximum has 2–3 months time delay compared to the time of maximum melting of the ice cap. Jónsdóttir et al. (2008) reach the conclusion that the LP events at Goðabunga are icequakes. Pinel et al. (2007) verified that the horizontal velocities measured by GPS in 1999–2004 cannot be explained only by the unloading from the melting of the ice cap and argue that magma inflow is needed to explain the observed inflation of the Katla volcano. A few deep-seated earthquakes (∼23 km) occurred near Hjörleifshöfði, southeast of Mýrdalsjökull in December 2007 in the magni- 86 JÖKULL No. 58, 2008 tude range Ml 0.7–1.6 (Icelandic Meteorological Office web site: hraun.vedur.is/ja/viku/2007/vika− 49). NW-Vatnajökull The NW part of Vatnajökull, above the hotspot, is one of the most active areas in Iceland, both seismically and volcanically. Other parts of Vatnajökull are not as active. The seismic activity can be divided into five main sites, each with different characteristics. These are Lokahryggur, Grímsvötn, Bárðarbunga, Kistufell and Kverkfjöll. Lokahryggur is located to the west of Gjálp/Grímsvötn (Figure 6). Two cauldrons, known as Skaftárkatlar, exist in the ice cover above Lokahryggur. Seismic activity on Lokahryggur is rather uniform in time with 40–50 earthquakes of magnitude Ml >1 recorded every year except before and after the Gjálp eruption in 1996– 1997, when the activity was much greater (Figures 4f and 6). The geothermal areas under the Skaftárkatlar cauldrons do not show more seismic activity than elsewhere on Lokahryggur. Seismicity in the Grímsvötn volcanic center is closely related to the eruptive activity (Figures 4g Seismicity in Iceland: 1994–2007 and 6), increasing as the pressure increases under the caldera, it is most intensive just before the onset of an eruption and almost stops after an eruption (see also in the section: Seismic signature of eruptions). In the three years before the Gjálp eruption in 1996 the Bárðarbunga volcanic center was fairly active with a ∼Ml 5 earthquake recorded every year. The activity was concentrated in and around the Bárðarbunga caldera (Figure 6), but until August 1996 the closest SIL-station was about 120 km away (Figure 1) and locations were not so precise (see subsequent section on the seismic signature of eruptions). After the Gjálp eruption the activity decreased and the Grímsvötn eruption in 1998 is not seen clearly in the Bárðarbunga data (Figure 4h). In November 2002 a swarm occurred in the NE part of Bárðarbunga with a Ml 4 earthquake and since then the area has again been more active, with earthquakes of magnitudes up to Ml 3–4. Earthquakes at Bárðarbunga are usually difficult to detect at Grímsfjall, the site of the closest station, but stations that were installed around Hálslón reservoir in 2004–2006 (Figure 1) enhance the detectability considerably. North of the Bárðarbunga volcanic center and just south of Kistufell is another concentration of earthquakes. The seismic activity in this area increased prior to the Gjálp eruption, especially in the last 6 months preceding and during the eruption (Figures 4i and 6). The largest earthquakes observed were ∼Ml 3. Only a few earthquakes were located in this area after the Gjálp eruption until April 2003. Since then, continuous activity has been recorded with a few swarms Figure 6. Earthquakes in NW-Vatnajökull larger than Ml 1 in 1994–2007. Location of the Gjálp 1996 volcanic fissure is shown in green and the two Grímsvötn eruptions in 1998 and 2004 are shown with yellow and stars. The boxes show areas plotted in Figure 4. – Kortið sýnir jarðskjálfta stærri en Ml ∼1 í norðvestanverðum Vatnajökli á árunum 1994–2007. Gjálp gossprungan er sýnd með grænum lit og Grímsvatnagosstöðvarnar frá 1998 og 2004 eru sýndar með gulri og blárri stjörnu. Ferningarnir afmarka svæði sem notuð eru við gerð 4. myndar. JÖKULL No. 58, 2008 87 S. S. Jakobsdóttir and earthquakes up to magnitude Ml 3–4. The observed activity has been increasing during the last few years, but that may be biased by the fact that several stations have been added to the network (Figure 1). Observed seismic activity at Kverkfjöll has increased very much in pace with the increasing number of stations in the area (Figures 4j and 6). The largest event recorded there was magnitude Ml 3.5 earthquake in November 1994. A few swarms have been detected since 2005 the largest one, comprising 20 earthquakes in the magnitude range Ml 1–3, occurred in August 2005. Some 50–70 earthquakes larger than Ml 1 are now located in the area every year (Figure 4j). Askja/Herðubreiðartögl The most active area in the highland outside the ice caps is Askja volcano and Herðubreiðartögl. Swarms have been recorded at Herðubreiðartögl since 1996 (Figures 3 and 4k). The largest recorded swarm occurred in June-July 1998 with earthquakes up to magnitude Ml 4. This swarm occurred just before two new stations were added to the network at about 50 km distance (Figure 1). The activity has been gradually increasing since late 2003. To some extent, this coincides with the increased number of stations, but not all the increased activity can be thus accounted for as no stations were added near Herðubreiðatögl from September 1998 until December 2004 when the stations around Hálslón reservoir were deployed. The earthquakes at the Askja volcanic center are in general small (Figures 3 and 4l). The largest earthquake recorded in this period was of magnitude Ml 3.5 and occurred in January 2004. The activity is mostly concentrated in the SE part of the volcanic center, but small swarms are repeatedly recorded north and northwest of Askja. The activity is steady and in recent years, with a denser network (Figure 1), earthquakes are recorded every year. The Askja caldera has been subsiding, at least since 1983 (Sturkell et al., 2006b). This is interpreted as a deflation of two magma chambers at different depths (Sturkell et al., 2006b; Pagli et al., 2006). Soosalu et al. (2008) studied the seismicity near Askja and Herðubreiðartögl with data from temporary local networks operated in 2006 and 2007, in addition to data from the SIL system. They located most of the earthquakes at depths 88 JÖKULL No. 58, 2008 down to 7–8 km (Soosalu et al., 2008). In addition they detected small, low frequency earthquakes occurring at depths between 15–34 km under most of the area. Some of these can be found in the SIL database. EARTHQUAKE SEQUENCES IN LESS ACTIVE AREAS In 1994–2007 earthquake swarms have occasionally occurred at a few intraplate locations and in central volcanoes and fissure swarms that generally show little activity. The magnitude of earthquakes in these swarms have not exceeded Ml 3.5 and they are mostly very small (Figures 7 and 8). Vestfirðir In February 1994 an earthquake of magnitude Ml 2.3 occurred in the northern part of Vestfirðir (Figure 7). It was felt in the town of Ísafjörður, some 35 km away. Only one earthquake was recorded, at a similar location as in 1964, when a single earthquake was located in a swarm of felt earthquakes. In October 1994 six earthquakes were located south of Arnarfjörður, some 85 km to the southwest from the February earthquake. The swarm lasted nine days and the earthquakes were in the magnitude range Ml 2–3. No more activity was recorded in the NW part of Vestfirðir, but in September 2006 a swarm of earthquakes occurred close to an extinct central volcano at Reykjarfjörður in the eastern part of Vestfirðir. Altogether 22 earthquakes in the magnitude range Ml 1.5–3.5 were located in a period of five days. (Þorbjarnardóttir et al., 2007). In 1996 and 1997 earthquakes were recorded offshore, about 25 km northeast of the Reykjarfjörður swarm. Guðlaugstungur Two seismic swarms were observed at Guðlaugstungur, about 20 km NNE of Langjökull, in 2004 (Figures 1 and 3). The first one lasted from the end of February to April with most earthquakes occurring in March (Figure 8a). Over 50 earthquakes of magnitudes Ml 1–2 were recorded in this period. In September the activity resumed and four sequences were recorded until the beginning of October. A couple of earthquakes reached Ml 3.5, both occurring towards the end of the swarm on 29 September and 1 Seismicity in Iceland: 1994–2007 October, respectively. Hjaltadóttir et al. (2005a) studied the swarms and found the first one grouped along a N-S lineament while the latter was concentrated in a NNW elongated dense cluster. Most of the earthquakes in both clusters were located at depths of 2-6 km, but in both sequences a few earthquakes reached a focal depth of 13 km. Eyjafjallajökull Before the SIL system started, no earthquakes were located beneath Eyjafjallajökull. In the first years, 1991–1993, some 70 earthquakes of magnitudes Ml 0–Ml 2.5 were located, mainly at the northern edge and near the summit crater (Figure 5). Only three earthquakes were detected in 1991, but at the end of April 1992 the activity increased. At the end of May 1994 an intense swarm started at the northern flank with 120 earthquakes recorded in one month (Figure 8b). The following September a swarm of 20 earthquakes occurred about 5 km north of the ice cap. In the next months earthquakes were measured in both areas and near the summit crater, but only a few earth- quakes were detected in the southern part of the ice cap. At the end of January 1995 the activity decreased again. In February–April 1996 activity was again observed, starting with a swarm in the area north of the ice cap, then a few small swarms near the crater and ending with a small swarm south of the crater. The next two and a half years were rather quiet. In December 1998 one more episode started with activity concentrated at the northern flank of the ice cap. On 5 July 1999 a swarm was observed in the SE part of the ice cap for the first time. From 10 August to 12 September and from 19 October to 29 November all earthquakes were located in the southern part. Earthquakes occurred both in the south and the north for the rest of the year. The last sequence in this episode, with activity mainly at the northern margin of the ice cap, lasted from 11 March to 30 May 2000. Since then a few small swarms have occurred, mainly in the northern part and north of Eyjafjallajökull. The largest earthquake in Eyjafjallajökull, of magnitude Ml 3.6, occurred in a swarm on 1 March 1999. Figure 7. Intra plate earthquakes located in Vestfirðir (Þorbjarnardóttir et al., 2008). – Jarðskjálftar, sem staðsettir hafa verið á Vestfjörðum. JÖKULL No. 58, 2008 89 S. S. Jakobsdóttir Figure 8. Magnitudes greater than Ml 1 as a function of time for areas shown by boxes in Figures 1 and 5. Note that scales are different on different plots. a: Guðlaugstungur, b: Eyjafjallajökull, c: Öræfajökull, d: Esjufjöll, e: Upptyppingar, f: Hengill. The gray line shows the accumulated number of earthquakes during the period. – Stærð jarðskjálfta stærri en Ml 1 sem fall af tíma. Svæðin eru merkt inn á myndir 1 and 5 og eru eftirfarandi: a: Guðlaugstungur, b: Eyjafjallajökull, c: Öræfajökull, d: Esjufjöll, e: Upptyppingar, f: Hengill. Athugið að ekki er sami skali á öllum myndunum. Gráa línan sýnir uppsafnaðan fjölda skjálfta yfir tímabilið. Sturkell et al. (2003a) and Pedersen and Sigmundsson (2004, 2006) have demonstrated two periods of uplift under Eyjafjallajökull. From tilt-, GPS- and InSAR data they find an uplift of the order of 15 cm centered at the southern margin of the ice cap in the time interval from 29 September 1993 to 19 September 1994 90 JÖKULL No. 58, 2008 and a second uplift starting between 19 July and 20 August 1999 and ending between February and May 2000. No uplift was found in relation to the swarms in 1996. Pedersen and Sigmundsson modeled the uplifts as sill formations at 4.5 and 6.3 km depths respectively with a variable amount of opening. This is in Seismicity in Iceland: 1994–2007 JÖKULL No. 58, 2008 91 S. S. Jakobsdóttir good agreement with Dahm and Brandsdóttir (1997) who found the best located earthquakes during the swarm in May and June 1994 at depths of 4.7–6.3 km. Hjaltadóttir et al. (2007) relocated earthquakes in the SIL database and the relocated hypocentres, both in 1994 and 1999, formed horseshoe-shaped clusters under the northern margin at depths between 9 and 11 km. This is well below the proposed sill intrusion and probably represents a feeding channel. They located the 1996 activity at 20–23 km depth and suggested that the deep origin of the earthquakes might explain the lack of uplift associated with that activity. They also mapped two smaller clusters in 1999, one situated at 6 km depth near the summit crater about 2.5 km west of the main activity and the other at 10 km depth, 2 km to the south. A small cluster in 2006 was mapped to a depth of 16 km beneath the easternmost of the two northward going outlet glaciers. Öræfajökull Some 50 earthquakes in the magnitude range Ml 0.5– 2.7 have been recorded under Öræfajökull since 1995, mostly in two swarms occurring in October 2002 and December 2005 and also during March to September 2007 (Figure 8c). In the 2002 swarm 8 earthquakes of magnitudes up to Ml 2 were detected. The swarm in 2005 included 10 earthquakes, the largest one again of magnitude Ml 2. A Ml 2.7 earthquake occurred in March 2005. In 2007, 11 earthquakes were recorded, all smaller than Ml 2. 92 JÖKULL No. 58, 2008 Esjufjöll Two swarms have been observed in the southern part of the Esjufjöll volcanic center during the study period. The first swarm started in late November 1996 with two Ml 2 earthquakes (Figure 8d). No seismicity was detected in the following weeks or until the swarm resumed on 21 December. During the next 5 days, 15 earthquakes were recorded in the magnitude range Ml 1–2.5 and on January 3 the swarm was completed with a magnitude Ml 2 earthquake. The closest stations at that time were at ∼60 and ∼95 km distance, both installed in October 1996. The second swarm occurred in 2002 (Figure 8d). It started in a similar way with a single ∼Ml 2 earthquake detected on 25 September. More than three weeks later the activity resumed and over 90 earthquakes in the magnitude range Ml 0.8–3.5 were recorded in the second half of October. The two largest earthquakes, Ml 3–3.5, occurred on 21 October. A few earthquakes were recorded a week later and finally about a month later, on 16 December, six more earthquakes with magnitudes up to Ml 2.5 occurred within a couple of hours. At this time the two closest stations were at about 40 km distance from Esjufjöll. Since 2005, when stations were added north of Vatnajökull, earthquakes have been detected every year in the Esjufjöll volcanic centre. Upptyppingar In February 2007 seismic activity began near Upptyppingar in the Kverkfjöll fissure swarm (Figures 3 Seismicity in Iceland: 1994–2007 and 8e). The seismic activity consisted of many small swarms and was increasing in the first months with a few periods of low activity in between. It culminated in July–August and then calmed down again. A new burst of swarms started on 16 October and lasted for a month. The most intensive swarm by end of 2007 began on 7 December with over 1000 earthquakes recorded in five days. It is still ongoing although a little calmer at the time of writing (Jakobsdóttir et al., 2008). Since February 2007, almost 5000 earthquakes have been detected in the magnitude range Ml -0.1– 2.2, but only a couple of earthquakes have reached magnitude Ml 2. The earthquakes are deep, 12–20 km, and epicenter locations are concentrated to small areas that migrate in time. The hypocenters do not define a single plane, but form an elongated cluster that dips ∼45◦ toward the south. Since August the activity has migrated to the northeast to the shallower end of the cluster (depth range 12–17 km). Such a long lasting swarm of deep earthquakes has never been observed before in Iceland, but with the station spacing prior to 1998 it would probably not have been detected. It is suggested here that this episode might be a formation of a magma chamber or a pluton with repeated injections of sills (Kavanagh et al., 2006). LARGER EVENTS The June 2000 earthquake sequence On 17 June 2000 at 15:40 h an earthquake that was widely felt in SW-Iceland struck in the middle of the South Iceland Lowland (SIL). The magnitude of this earthquake is given by NEIC as Ms 6.6 (Table 1). The earthquake was expected, as long term statistical forecast stated a ∼90% probability of an earthquake larger than magnitude 6 to occur in the area within 20 years and for various reasons it was expected to occur in the seismic gap observed at the epicenter area of the 17 June earthquake (Stefánsson and Halldórsson, 1988). It had also been pointed out that the gap coincided with a long-term concentration of microearthquake activity in the seismic zone (Stefánsson et al., 1993) and although it was assumed more likely that the microearthquakes indicated high stress in preparation of a large earthquake it could not be excluded that they reflected aseismic strain release. At least three earthquakes were triggered by the S wave of the main shock in SISZ and on the Reykjanes Peninsula (RP) during the first 30 seconds (Figure 3) (Árnadóttir et al., 2003; Vogfjörð, 2003; Vogfjörð et al., 2005a; Anatoli et al., 2006). The magnitudes were in the range 3.5–5.5. Some additional M5 earthquakes occurred along the zone in the next minutes (Table 1) and thousands of aftershocks were measured along the 100 km long belt in the following days. The area around Geysir, some 35 km north of the main shock, was activated and showed increased seismicity for about a year. One of the aftershock sites in the SISZ was south of the lake Hestvatn. As the activity continued there and based on the knowledge of former strain release in SISZ, a forecast was given to the Civil Protection Department that an earthquake of similar size, centered near Hestvatn, might strike at any time. The warning was issued 25 hours prior to the Ms 6.6 earthquake which struck early on 21 June, within the box given by the forecast as the most likely location (Stefánsson et al., 2000, 2003). The June 2000 earthquakes have been studied by several authors and they were the backbone of the EU project PREPARED (Stefánsson 2006; Stefánsson et al., 2006a,b). Hjaltadóttir and Vogfjörð (2004, 2005) mapped the main faults with relative locations of the aftershocks. They found them composed of several segments, reaching to 10 km depth. The faults have the overall strike of 7◦ and 179◦ , respectively. They are both near vertical except that the northern part of the Hestvatn fault dips 77◦ east. This is in good agreement with fault plane solutions obtained by Harvard, that give strike 4◦ , dip 87◦ and rake -164◦ for the 17 June earthquake and strike 2◦ , dip 85◦ and rake -167◦ for the 21 June event. The CSEM solutions give strike 16◦ , dip 68◦ , rake -158◦ and strike 358◦, dip 85◦ , rake -175◦ , respectively. The faults have also been modeled to fit various types of data with similar results. Pagli et al. (2003) and Pedersen et al. (2001, 2003) modeled the faults on the Reykjanes Peninsula and in SISZ using InSAR data and GPS data whereas Suhadolc used strong motion data to constrain his model (Hjaltadóttir et al., 2005b; Stefánsson et al., 2006b). Árnadóttir et JÖKULL No. 58, 2008 93 S. S. Jakobsdóttir al. (2001, 2003, 2004, 2005) have studied the crustal deformation before, during and after the earthquakes using GPS data and Jónsson et al. (2003) compared post-earthquake ground movement with water level in boreholes. Lindman et al. (2006) studied the physics of the aftershock sequences. No casualties were caused by the earthquakes, but some buildings had to be abandoned due to damages. The Hengill-Ölfus episode in 1994–1998 In August 1994 a swarm of more than 5000 earthquakes occurred near Hrómundartindur north of Hveragerði. A slight increase in seismicity had been observed already a month earlier. The swarm started on 13 August with a magnitude Ml 3.5 earthquake and 6 more earthquakes of magnitude >Ml 3.5 occurred during the next 8 days. The swarm lasted 10 days. The largest earthquake in the swarm reached magnitude Ml 4 and occurred on 20 August. This swarm marked the beginning of an unusually active period at the triple junction near Hengill (Figure 8f) (Rögnvaldsson et al., 1998b,c). It lasted for about five years and culminated in two ∼Ml 5 earthquakes in June and November 1998. Over 90,000 earthquakes were located altogether in this episode, which makes 36% of the earthquakes in the SIL data base for the whole period studied in this paper. The activity was mainly concentrated north of Hveragerði, but increased seismicity was simultaneously observed on a WSW-ENE trending belt southwest of Hveragerði and west of SISZ. The Ml 4.7 earthquake (NEIC 5.40 Mw , Table 1), that occurred on 4 June 1998, ruptured the area between these two concentrations and the Ml 4.9 (NEIC 5.10 Mw ) earthquake on 13 November 1998 originated in the southern belt. Vogfjörð et al. (2005b) mapped in detail the faults that were active in 1996–1998. They find near vertical NE and E striking faults most common with right lateral slip on NE striking faults and left lateral slip on E striking faults. Normal faulting is also seen. An uplift of ∼7.5 cm centered south of Hrómundartindur volcanic center was observed from interferograms covering the period from July 1993 to September 1998, using InSAR (Feigl, 2000). Uplift was also seen in leveling data from 1992–1995 (Sigmundsson et al., 1997). The uplift was interpreted as magma injection and modeled 94 JÖKULL No. 58, 2008 with a Mogi source at a depth of 6.5–7 km. Neither of these uplift studies included the November 1998 rupture. Other earthquakes greater than M 5. The NEIC bulletin reports 10 earthquakes with magnitude greater than 5 in the time interval 1994–2007 (Figure 3, Table 1). Two of these are the 1998 earthquakes in the Hengill episode and three more belong to the June 2000 episode in SISZ. The first ∼M 5 earthquake occurred on 8 February 1994 some 35 km off the north coast in the western part of the TFZ. The local magnitude Ml is 4.8 while the NEIC bulletin gives Mw 5.5 (Table 1). This was the first earthquake in a swarm that lasted for 3 days with more than 800 recorded earthquakes on a NW trending extension of the Húsavík-Flatey fault. The next big earthquake occurred in Bárðarbunga on 5 May 1994. The local magnitude was Ml 4.2 while the NEIC magnitude was mb 5.7. Only 2 aftershocks were detected. The following year, on 11 December 1995, a Ml 3.9 earthquake (NEIC mb 4.9) struck Bárðarbunga and again on 29 September 1996 a Ml 4.4 (NEIC Mw 5.6) earthquake occurred (Table 1). This event marked the onset of seismic unrest under Bárðarbunga that culminated in an eruption in Gjálp, SE of Bárðarbunga, 1 1/2 days later (see also the section: Seismic signature of eruptions). No earthquakes M>4.5 have been reported by NEIC in Bárðarbunga since the Gjálp eruption. On 16 September 2002 a Ml 5.0 (NEIC Mw 5.8) was recorded in the northern part of the TFZ, 53 km NNV of Grímsey (Þorbjarnardóttir and Guðmundsson 2003). More than 450 aftershocks were observed in the following 5 days. The location of the swarm is near the submarine mountain Stóragrunn that was mapped by Brandsdóttir et al. (2002). The last earthquake larger than 5 reported by NEIC was a Mw 5.2 earthquake with Ml 4.3 that occurred on 23 August 2003 (Table 1). The epicenter location was at Sveifluháls on the Reykjanes Peninsula, 3 km north of Krísuvík. The aftershock distribution was along a 6.5 km long belt striking NS and along a 4 km belt striking ∼60◦ to ENE from the main shock. A north-south oriented strike slip fault is in agreement with the optimum fault plane solution (Guðmundsson Seismicity in Iceland: 1994–2007 et al., 2004). Vogfjörð et al. (2004, 2005a) mapped the fault using aftershocks and GPS data. The result reveals a vertical strike-slip event striking N-S. SEISMIC SIGNATURE OF ERUPTIONS The buildup of pressure under active volcanoes is often manifest as an increase in seismic activity. The earthquakes associated with this increased pressure and eventual magma movement are often very small. With a more sensitive seismic network in the vicinity of volcanoes, the possibility of detecting the preparation phase of an eruption is enhanced. Gjálp 1996 When Gjálp, in northwestern Vatnajökull (Figure 6), erupted in October 1996 the SIL network consisted mostly of two network clusters, one at the SISZ and another at the TFZ. Two highland stations, some 100 and 50 km west of Bárðarbunga, were implemented a month before and the day after the onset of the eruption. Nevertheless some increase in seismic activity was observed at Bárðarbunga from the end of August 1995 and the most intensive swarm recorded by the SIL system at Lokahryggur occurred in February 1996. No earthquakes were recorded at the location of the Gjálp fissure in the months prior to the eruptions. The volcanic episode started with a Mw 5.6 earthquake in Bárðarbunga at 10:48 on 29 September (Table 1). Nettles and Ekström (1998) find the fault mechanism consistent with reverse faulting propagating, as a series of subevents, around the surface of a cone. The intensive aftershock activity shifted to the Gjálp location some hours before the onset of the subglacial eruption in the evening of 30 September (Einarsson et al., 1997; Vogfjörð 1999). The eruption lasted 2 weeks and has been described by several authors (Guðmundsson et al., 1997; Einarsson et al., 1997). Seismic activity continued at Gjálp and Lokahryggur until June 1997. Grímsvötn 1998 and 2004 The seismic behavior of the two Grímsvötn eruptions in 1998 and 2004 was in many ways similar (Figures 4g and 6). Increased seismic activity was detected some months before the eruptions and the area became quiet for a while after the eruptions. The first earthquakes prior to the eruption in December 1998 were recorded in March the same year and then one event was recorded every month until November when three earthquakes were detected. Early morning on 18 December an earthquake swarm was observed and the eruption started at about 9:30. No earthquake was located in Grímsvötn for months after the eruption. The two closest stations at that time were about 55 km away. Sturkell et al. (2003b) observed a deflation of a magma chamber associated with the eruption. Before the eruption they had measured uplift of the caldera rim, Grímsfjall. The uplift started again after the eruption implying a new inflation period. In 2001 a seismic station was installed on the caldera rim on Grímsfjall and five more stations were within a distance of 110 km around the Vatnajökull ice cap. As a result, earthquakes were now measured every month in Grímsvötn. In July 2003, 15 months prior to the second eruption, an increase was observed in the seismicity and on 17 October 2004 the seismic activity was again seen to increase. Monitoring of icequakes and microseismicity revealed a glacial flood starting from Grímsvötn caldera on 29 October and bursting out through an outlet glacier the following day. Two days later, in the early morning of 1 November, a magnitude Ml 3 earthquake occurred in a swarm that lasted for a few hours. At 19:30 the swarm resumed as the magma started to move towards the surface marking the onset of the eruption (Vogfjörð et al., 2005c). Very few earthquakes were recorded in the area during the next year (Figure 4g). Hekla 2000 Hekla volcano is known to be seismically quiescent between eruptions. Only a few earthquakes are detected and they generally have magnitudes Ml <1. The first earthquake recorded by the SIL system before the Hekla eruption on 26 February 2000 occurred at 17:05 h with magnitude Ml 0.3. The biggest earthquake in the swarm that followed had magnitude Ml 2.2 and occurred at 18:17 h, which is two minutes before the onset of the eruption. After the first hours of the eruption no locatable earthquakes were observed, but continuous tremor was recorded. This is typical behavior, similar to what was observed prior to and during the eruption in Hekla in 1991. Further studies of Hekla, JÖKULL No. 58, 2008 95 S. S. Jakobsdóttir based on SIL data and data from a permanent analog network (Einarsson and Björnsson, 1976, 1987) are found in Soosalu and Einarsson (2004, 2006) and Soosalu et al. (2005). DISCUSSION AND CONCLUSIONS The seismicity in Iceland during 1994–2007 has been described above. The detection threshold for different areas has been changing in this period, but instead of limiting the study to earthquakes of magnitudes higher than the magnitude of completeness in 1994, an effort has been made to compare differences in observed seismic activity at different times. As the network becomes denser, new areas of activity are detected and the information obtained becomes more detailed. One of the details that are emerging in the last years is the occurrence of deep earthquakes. It is now confirmed that deep earthquakes of a tectonic type occur near Eyjafjallajökull (Hjaltadóttir et al., 2007) and at Upptyppingar, while deep earthquakes observed near Askja are LP earthquakes with emergent onsets (Soosalu, 2008). It has also been possible to detect an imminent volcanic eruption at an earlier stage with lower detection thresholds. Intraplate earthquakes are occasionally observed, both as single earthquakes and as swarms. There are 250,000 recorded earthquakes in the SIL database for the period 1994-2007. Of these over 200,000 were recorded in the seismic zones: TFZ (48,800), SISZ (35,200), Hengill triple junction (99,200) and RP (18,900), but these areas have had the lowest detection threshold during the whole period. On Reykjanes Ridge, around Eldey and Geirfuglasker, some 2000 earthquakes were detected. 16,600 earthquakes were observed in Mýrdalsjökull and 3,700 in NW-Vatnajökull. About 8,000 earthquakes have been recorded in the highland north of Vatnajökull, where 5,000 of them occurred at Upptyppingar in 2007. Bárðarbunga has resumed its previous activity after a four-year-period of quiet following the Gjálp eruption and seismicity near Kistufell seems to be increasing. Seismic activity in Herðubreiðartögl has been increasing in the last four years. 96 JÖKULL No. 58, 2008 Acknowledgements This paper is based on the work of the monitoring team at the Physics Department of the Icelandic Meteorological Office over the 14 year period from 1994 to 2007. Reynir Böðvarsson and Ragnar Slunga are the main designers of the SIL-system. Gunnar B. Guðmundsson and Matthew J. Roberts contributed most of the figures. Gunnar, Matthew and Bergþóra Þorbjarnardóttir critically read and improved the manuscript. All maps were made with GMT software (Wessel and Smith, 1998). The graphs were made using R (R Development Core Team, 2008). Insightful comments by the reviewers, Björn Lund and Sveinbjörn Björnsson, helped to improve the paper. ÁGRIP Jarðskjálftavirkni á Íslandi á árunum 1994 til 2007 Seinni hluta árs 1993 voru settar upp fimm stafrænar jarðskjálftamælistöðvar í kringum Tjörnesbrotabeltið til viðbótar 11 stöðvum á Suðurlandi og stöð á Akureyri. Þar með varð hið sjálfvirka jarðskjálftamælikerfi, SIL-kerfið, að landsneti jarðskjálftamæla. Sumarið 1994 var sett upp stöð í Borgarfirði. Á næstu árum óx kerfið úr 18 stöðvum, árið 1994, í 51 stöð í árslok 2007. Stöðvarnar eru að mestu leyti staðsettar í kringum flekaskilin (1. mynd). Eftir því sem stöðvum fjölgaði jókst næmni kerfisins, sérstaklega á hálendinu. Hér er fjallað um jarðskjálftavirkni á Íslandi frá ársbyrjun 1994. Á tímabilinu hafa flestir skjálftar á landinu stærri en 2 verið staðsettir, en á sumum svæðum hafa mun smærri skjálftar verið skráðir. Þannig hafa nánast allir skjálftar niður í stærð ∼0 verið skráðir á Suðurlandsundirlendi og skjálftar stærri en ∼1 í Tjörnesbrotabeltinu. Í lok tímabilsins voru stöðvar orðnar það margar að flestir skjálftar á landinu stærri en ∼1–1,5 voru skráðir, nema skjálftar á Vestfjörðum. Í stað þess að takmarka umfjöllunina í þessari grein við skjálfta stærri en þá sem mælanlegir voru árið 1994, er gerð grein fyrir aukinni næmni á mismunandi svæðum samhliða umræðunni um virknina. Töluvert er fjallað um smáskjálftavirkni, bæði á vel þekktum jarðskjálftasvæðum sem og á svæðum þar sem jarðskjálftar eru óalgengir. Seismicity in Iceland: 1994–2007 Um 250.000 jarðskjálftar voru skráðir í SILkerfinu á þessu tímabili (2. og 3. mynd). Þar af mældust um 48.800 skjálftar í Tjörnesbrotabeltinu, um 35.200 í Suðurlandsbrotabeltinu og um 18.900 á Reykjanesskaga. Mynd 4. sýnir skjálftavirkni sem fall af tíma fyrir helstu jarðskjálftasvæðin á hálendinu og undir jöklunum. Svæðin sem um er að ræða eru merkt á 1., 5. og 6. mynd. Tvær skjálftahrinur hafa mælst á Vestfjörðum, annars vegar við Arnarfjörð árið 1994 og hins vegar við Reykjarfjörð árið 2006 (7. mynd). Á 8. mynd má sjá skjálftavirkni sem fall af tíma fyrir nokkur svæði þar sem jarðskjálftavirkni er stopulli og fyrir Hengilssvæðið, sem er eitt virkasta svæði landsins. Á Hengilssvæðinu - á mótum vestra gosbeltisins, Suðurlandsbrotabeltisins og Reykjanesskaga mældust ∼99.200 skjálftar á tímabilinu. Á þessu svæði var sérstaklega mikil skjálftavirkni á árunum 1994 til 1998 og náði hún hámarki með tveimur skjálftum af stærð ∼5 í júní og nóvember 1998. Á sama tíma lyftist land á svæðinu og hefur atburðurinn verið túlkaður sem uppsöfnun kviku á um 7 km dýpi. Á Hengilssvæðinu hefur næmnin verið mikil á öllu tímabilinu og mælast þar skjálftar allt niður í stærð -2. Kortið á 5. mynd sýnir skjálfta stærri en 1 undir Eyjafjallajökli og Mýrdalsjökli. Virkni undir Eyjafjallajökli mældist mest í hrinum á árunum 1994, 1996 og 1999 (mynd 7b) og er talin tengjast innskotavirkni. Undir Mýrdalsjökli er virknin mest við Goðabungu (mynd 4d). Flestir skjálftar mælast þar síðsumars og fram á vetur. Á árunum 2001 til 2005 var virknin mikil allt árið, en þó með árstíðabundinni aukningu. Tvær kenningar eru um orsakir virkninnar, annars vegar að hraungúll mjakist í átt að yfirborði undir Goðabungu og hins vegar að virknin sé í raun ísskjálftar tengdir jöklinum. Töluvert minni jarðskjálftavirkni er undir Kötluöskjunni og hún er ekki eins greinilega tengd árstíðum (mynd 4e). Landlyfting mældist undir öskjunni á árunum 2001 til 2004, en við Goðabungu hefur bæði mælst lyfting og sig. Tveir stórir skjálftar urðu á Suðurlandi í júní 2000. Bylgjan frá fyrri skjálftanum, þann 17. júní, setti af stað röð skjálfta af stærð 3.5–5.5 vestan við stóra skjálftann, allt vestur að Kleifarvatni. Þúsundir skjálfta mældust á Suðurlandi á næstu vikum og mán- uðum á eftir. Yfirvofandi eldgos má yfirleitt sjá út frá skjálftavirkni. Mismunandi eldfjöll sýna mismunandi aðdraganda. Þannig hefst skjálftavirkni undir Heklu rúmlega klukkutíma fyrir gos, en almennt má sjá aukna jarðskjáftavirkni dögum og jafnvel árum fyrir eldgos. Þannig mátti sjá lengri aðdraganda að Gjálpar- og Grímsvatnagosunum í Vatnajökli. Fyrir gosið í Grímsvötnum 2004 mátti sjá aukna virkni meira en ári fyrir gos (mynd 4g) og skjálftavirknin jókst enn meira um tveimur vikum fyrir gos. Í báðum gosunum í Grímsvötnum á tímabilinu mátti greina tilsvarandi aukningu fyrir gos og eftir goslok datt skjálftavirknin alveg niður í marga mánuði á eftir í báðum tilfellum (mynd 4g). Mun fleiri mælistöðvar voru í nágrenni gosstöðvanna árið 2004 en í gosinu 1998. Í Gjálpargosinu hófust forskjálftar við Bárðarbungu og skjálftavirknin færðist síðan að gosstöðvunum rétt fyrir gos. Eftir gosið hélt skjálftavirkni áfram í nokkra mánuði í Bárðarbungu (mynd 4h) og á Lokahrygg (mynd 4f). Mesta jarðskjálftavirknin á Lokahrygg á tímabilinu mældist í kjölfar gossins. REFERENCES Allen, R., G. Nolet, W. J. Morgan, K. S. Vogfjörð, M. Netles, G. Ekström, B. H. Bergsson, P. Erlendsson, G. Foulger, S. S. Jakobsdóttir, B. Julian, M. Pritchard, S. Ragnarsson and R. Stefánsson 2002. Plume driven plumbing and crustal formation in Iceland. J. Geophys. Res. 107(B8), doi:10.1029/2001JB000584. Antonioli, A., M. E. Belardinelli, A. Bizzarri and K. S. Vogfjörð 2006. Evidence of instantaneous dynamic triggering during the seismic sequence of year 2000 in south Iceland. J. Geophys. Res. 111(B03302), doi:10.1029/2005JB003935. Ágústsson, K. 2006. Mat á næmni SIL jarðskjálftamælinetsins: Hugmyndir um framtíðaruppbyggingu. Icelandic Meteorological Office-Report 06014, 22 pp. Árnadóttir, Þ., S. Hreinsdóttir, G. B. Guðmundsson, P. Einarsson, M. Heinert and C. Völksen 2001. Crustal deformation measured by GPS in the South Iceland Seismic Zone due to two large earthquakes in June 2000. Geophys. Res. Lett. 28(21), 4031–4033. Árnadóttir, Þ., S. Jónsson, R. Pedersen and G. B. Guðmundsson 2003. Coulomb stress changes in the South Iceland Seismic Zone due to two large earthquakes in June 2000. Geophys. Res. Lett. 30(5), doi:10.1029/2002GL016495. Árnadóttir, Þ., H. Geirsson and P. Einarsson 2004. Coseismic stress changes and crustal deformation on the Reykjanes Peninsula due to triggered earthquakes on 17 June 2000. J. Geophys. Res. 109(B09307), doi:10.1029/2004JB003130. JÖKULL No. 58, 2008 97 S. S. Jakobsdóttir Árnadóttir, Þ., S. Jónsson, F. F. Pollitz, W. Jiang and K. L. Feigl 2005. Postseismic deformation following the June 2000 earthquake sequence in the South Iceland Seismic Zone. J. Geophys. Res. 110(B12308), doi:10.1029/2005JB003701. Bjarnason, I. Th. 2008. An Iceland hotspot saga. Jökull 58, this issue. Brandsdóttir, B., R. Detrick, G. Helgadóttir, E. Kjartansson, B. Richter, K. Gunnarsson, S. Þ. Guðlaugsson, N. Driscoll and G. Kent 2002. A new perspective of the tectonics of the Tjörnes Fracture Zone, offshore northern Iceland, from EM300 Multibeam Bathymetry, High Resolution MCS and CHIRP Sonar Profiles. Eos Trans. AGU 83, F731. Böðvarsson, R., S. Th. Rögnvaldsson, S. S. Jakobsdóttir, R. Slunga and R. Stefánsson 1996. The SIL data acquisition and monitoring system. Seism. Res. Lett. 67, 35–46. Böðvarsson, R., S. Th. Rögnvaldsson, R. Slunga and E. Kjartansson 1999. The SIL data acquisition system-at present and beyond year 2000. Phys. Earth Planet. Inter. 113, 89–101. Dahm, T. and B. Brandsdóttir 1997. Moment tensors of microearthquakes from the Eyjafjallajökull volcano in South Iceland. Geophys. J. Int. 130, 183–192. Einarsson, P. 1991. Earthquakes and present-day tectonism in Iceland. Tectonophysics 189, 261–279. Einarsson, P. 2008. Plate boundaries, rifts and transforms in Iceland. Jökull 58, this issue. Einarsson, P. and S. Björnsson 1976. Seismic activity associated with the 1970 eruption of Volcano Hekla in Iceland. Jökull 26, 8–19. Einarsson, P. and S. Björnsson 1987. Jarðskjálftamælingar á Raunvísindastofnun Háskólans. In: Þ. I. Sigfússon (editor), Í hlutarins eðli. Festschrift for Þorbjörn Sigurgeirsson, Menningarsjóður, Reykjavík, 251–278. Einarsson, P. and K. Sæmundsson 1987. Earthquake epicenters 1982–1985 and volcanic systems in Iceland (map). In: Þ. I. Sigfússon (editor), Í hlutarins eðli. Festschrift for Þorbjörn Sigurgeirsson, Menningarsjóður, Reykjavík. Einarsson, P., B. Brandsdóttir, M. T. Guðmundsson, H. Björnsson, K. Grönvold and F. Sigmundsson 1997. Unrest near the center of the Iceland hotspot: The October 1996 subglacial eruption in Vatnajökull. Eos Trans. AGU 78, 369, 373–375. Einarsson P. and B. Brandsdóttir 2000. Earthquakes in the Mýrdalsjökull area, Iceland, 1978–1985: Seasonal correlation and connection with volcanoes. Jökull 49, 59–73. Feigl, K., J. Gasperi, F. Sigmundsson and A. Rigo 2000. Crustal deformation near Hengill volcano, Iceland 1993–1998: Coupling between magmatic activity and faulting inferred from elastic modelling of satellite radar interferograms. J. Geophys. Res. 105, 25655–25670. Guðmundsson, G. B., S. S. Jakobsdóttir and B. S. Þorbjarnardóttir 2004. Seismicity in Iceland 2003. Jökull 54, 67–74. Gudmundsson, M. T., F. Sigmundsson and H. Björnsson 1997. Ice-volcano interaction of the 1996 Gjálp subglacial eruption, Vatnajökull, Iceland. Nature 389, 954–957. Hjaltadóttir, S. and K. S. Vogfjörð 2004. WP5.1. Mapping subsurface faults in southwestern Iceland with the microearthquakes induced by the June 17th and June 21st 98 JÖKULL No. 58, 2008 earthquakes. In: PREPARED-first periodic report. February 1, 2003-January 31, 2004. Icelandic Meteorological OfficeReport 04014, 80–84. Hjaltadóttir S. and K. S. Vogfjörð 2005. Subsurface fault mapping in Southwest Iceland by relative location of aftershocks of the June 2000 earthquakes. Icelandic Meteorological OfficeResearch Report 21, 18 pp. Hjaltadóttir, S., H. Geirsson and Þ. Skaftadóttir 2005a. Seismicity in Iceland 2004. Jökull 55, 107–119. Hjaltadóttir, S., K. S. Vogfjörð, Þ. Árnadóttir, P. Einarsson and P. Suhadolc 2005b. A model of the release of the two June 2000 earthquakes based on all available observations. Icelandic Meteorological Office-Report 05020, 9 pp. Hjaltadóttir, S., K. S. Vogfjörð and R. Slunga 2007. Relocated earthquakes in Eyjafjallajökull between 1991 and 2006: Does deep seismicity indicate intruding material from below? In: ESC Working Group “Earthquakes and Volcanoes" Meeting: "Hot and Cold: Seismicity associated with geothermal areas and ice-covered volcanoes”, abstract, Nesjavellir, Iceland. Höskuldsson, Á., R. Hey, E. Kjartansson and G. B. Guðmundsson 2007. The Reykjanes Ridge between 63◦ 10’N and Iceland. J. Geodynamics 43, 73–86. Jakobsdóttir, S. S., G. B. Guðmundsson and R. Stefánsson 2002. Seismicity in Iceland 1991–2000 monitored by the SIL seismic system. Jökull 51, 87–94. Jakobsdóttir, S. S., M. J. Roberts, G. B. Guðmundsson, H. Geirsson and R. Slunga 2008. Earthquake swarms at Upptyppingar, North-east Iceland: a sign of magma intrusion? Studia Geophysica et Geodaetica, accepted. Jónsdóttir, K., A. Tryggvason, R. Roberts, B. Lund, H. Soosalu and R. Böðvarsson 2007. Habits of a glacier covered volcano: Seismicity and structure study of the Katla volcano, South Iceland. Annals of Glaciology 45, 169–177. Jónsdóttir, K., R. Roberts, A. Tryggvason, B. Lund, V. Pohjola, S. S. Jakobsdóttir and R. Böðvarsson 2008. Local lp-events study in a glaciated volcanic environment in south Iceland. In: Abstract of the EGU General Ass., Vienna, Austria. Jónsson, S., P. Segall, R. Pedersen and G. Björnsson 2003. Postearthquake ground movements correlated to pore-pressure transients. Nature 424, 179–183. Kavanagh, J. L., T. Menand and R. S. J. Sparks 2006. An experimental investigation of sill formation and propagation in layered elastic media. Earth Planet. Sci. Lett. 245, 799–813. doi:10.1016/j.epsl.2006.03.025. Lindman, M., B. Lund, R. Roberts and K. Jónsdóttir 2006. Physics of the Omori law: Inferences from interevent time distributions and pore pressure diffusion modeling. Tectonophysics 424, 209–222. Lippitsch, R., R. S. White and H. Soosalu 2005. Precise hypocenter relocation of microearthquakes in a high-temperature geothermal field: the Torfajökull central volcano, Iceland. Geophys. J. Int. 160, 371–388. Nettles, M. and G. Ekström 1998. Faulting mechanism of anomalous earthquakes near Bárdarbunga Volcano, Iceland. J. Geophys. Res. 103, 17,973–17,983. Seismicity in Iceland: 1994–2007 Pagli, C., R. Pedersen, F. Sigmundsson and K. L. Feigl 2003. Triggered fault slip on June 17, 2000 on the Reykjanes Peninsula, SW Iceland captured by radar interferometry. Geophys. Res. Lett. 30(6), 1273, doi:10.1029/2002GL015310. Pagli, C., F. Sigmundsson, Þ. Árnadóttir, P. Einarsson and E. Sturkell 2006. Deflation of the Askja volcanic system: contraints on the deformation source from combined inversion of satellite radar interferograms and GPS measurements. Journ. Volc. Geotherm. Res. 152, 97–108. Pedersen, R., F. Sigmundsson, K. L. Feigl and Þ. Árnadóttir 2001. Co-seismic interferograms of two Ms = 6.6 earthquakes in the South Iceland Seismic Zone, June 2000. Geophys. Res. Lett. 28, 3341–3344. Pedersen, R., S. Jónsson, Þ. Árnadóttir, F. Sigmundsson and K. L. Feigl 2003. Fault slip distribution of two Mw = 6.5 earthquakes in South Iceland estimated from joint inversion of InSAR and GPS measurements. Earth Plan. Sci. Lett. 213, 487–502. Pedersen, R. and F. Sigmundsson 2004. InSAR based sill model links spatially offset areas of deformation and seismicity for the 1994 unrest episode at Eyjafjallajökull volcano. Iceland. Geophys. Res. Lett. 31, L14610. Pedersen, R. and F. Sigmundsson 2006. Temporal development of the 1999 intrusive episode in the Eyjafjallajökull volcano, Iceland, derived from InSAR images. Bull. Volc. 68, 377– 393. Pinel, V., F. Sigmundsson, E. Sturkell, H. Geirsson, P. Einarsson, M. T. Guðmundsson and Þ. Högnadóttir 2007. Discriminating volcano deformation due to magma movements and variable surface loads: application to Katla subglacial volcano, Iceland. Geophys. J. Int. 169(1), 325–338. R Development Core Team 2008. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. ISBN 3-900051-07-0. URL http://www.R-project.org. Roberts, M. J. 2005. Aerial and seismic observations of the August 2005 jökulhlaup from Grænalón. Icelandic Meteorological Office-Report 05022, 18 pp. Rögnvaldsson, S. Th. and R. Slunga 1993. Routine fault plane solutions for local and regional networks: a test with synthetic data. Bull. Seism. Soc. Am. 11, 1247–1250. Rögnvaldsson, S. Th. and R. Slunga 1994. Single and joint fault plane solutions for microearthquakes in South Iceland. Tectonophysics 237, 73–86. Rögnvaldsson, S. Th., Á. Guðmundsson and R. Slunga 1998a. Seismotectonic analysis of the Tjörnes Fracture Zone, an active transform fault in north Iceland. J. Geophys. Res. 103(B12), 30,117Ů30,129. Rögnvaldsson, S. Th., G. B. Guðmundsson, K. Ágústsson, S. S. Jakobsdóttir, R. Slunga and R. Stefánsson 1998b. Overview of the 1993–1996 seismicity near Hengill. Icelandic Meteorological Office-Research Report VÍ-R98006, 16 pp. Rögnvaldsson, S. Th. Þ. Árnadóttir, K. Ágústsson, Þ. Skaftadóttir, G. B. Guðmundsson, G. Björnsson, K. S. Vogfjörð, R. Stefánsson, R. Böðvarsson, R. Slunga, S. S. Jakobsdóttir, E. Kjartansson, B. S. Þorbjarnardóttir, P. Erlendsson, B. H. Bergsson, S. Ragnarsson, P. Halldórsson, B. Þorkelsson and M. Ásgeirsdóttir 1998c. Skjálftahrina í Ölfusi í nóvember 1998. Icelandic Meteorological Office-Report VÍ-G98046, 19 pp. Sigmundsson, F., P. Einarsson, S. Th. Rögnvaldsson, G. R. Foulger, K. M. Hodgkinson and G. Þorbergsson 1997. The 1994– 1995 seismicity and deformation at the Hengill triple junction, Iceland: Triggering of earthquakes by minor magma injection in zone of horizontal shear stress. J. Geophys. Res. 102, 15,151Ů15,161. Slunga, R. 1981. Earthquake source mechanism determination by use of body-wave amplitudes-an application to Swedish earthquakes. Bull. Seism. Soc. Am. 71, 25–35. Slunga, R., S. Th. Rögnvaldsson and R. Böðvarsson 1995. Absolute and relative locations of similar events with application to microearthquakes in southern Iceland. Geophys. J. Int. 123, 409–419. Soosalu H. and P. Einarsson 2004. Seismic constraints on magma chambers at Hekla and Torfajökull volcanoes, Iceland. Bull. Volc. 66, 276–286. Soosalu, H., P. Einarsson and B. S. Þorbjarnardóttir 2005. Seismic activity related to the 2000 Hekla eruption, Iceland. Bull. Volc. 68, 21–36. Soosalu H. and P. Einarsson 2006. Seismic characteristics of the Hekla volcano, Iceland. Jökull 55, 87–106. Soosalu, H., K. Jónsdóttir and P. Einarsson 2006a. Seismicity crisis at the Katla volcano, Iceland-signs of a cryptodome? J. Volc. Geotherm. Res. 153, 177–186. Soosalu, H., R. Lippitsch and P. Einarsson 2006b. Low-frequency earthquakes at the Torfajökull volcano, South Iceland. J. Volc. Geotherm. Res. 153, 187–199. Soosalu, H., C. Knox, R. S. White, P. Einarsson, S. S. Jakobsdóttir and E. Sturkell 2007. Seismicity at the Askja volcano and its surroundings, North Iceland. In: Abstracts of the Volcanic and Magmatic Studies Group, Winter Meeting, Oxford, United Kingdom. Soosalu, H., R. S. White, A. J. Key, C. Knox, P. Einarsson and S. S. Jakobsdóttir 2008. Lower-crustal earthquakes reflect magma movements beneath the north Iceland rift near Askja. In: Abstracts of the EGU General Assembly, Vienna, Austria. Stefánsson, R. and P. Halldórsson 1988. Strain build-up and strain release in the South Iceland seismic zone. Tectonophysics 152, 267–276. Stefánsson, R., R. Böðvarsson, R. Slunga, P. Einarsson, S. S. Jakobsdóttir, H. Bungum, S. Gregersen, J. Havskov, J. Hjelme and H. Korhonen 1993. Earthquake prediction research in the South Iceland seismic zone and the SIL project. Bull. Seism. Soc. Am. 83, 696–716. Stefánsson, R., G. B. Guðmundsson and P. Halldórsson 2000. The two large earthquakes in the South Iceland seismic zone on June 17 and 21, 2000. Icelandic Meteorological OfficeReport 00010, 9 pp. Stefánsson, R., G. B. Guðmundsson and P. Halldórsson 2003. The South Iceland earthquakes 2000 - a challenge for earthquake prediction research. Icelandic Meteorological Office-Report 03017, 21 pp. JÖKULL No. 58, 2008 99 S. S. Jakobsdóttir Stefánsson, R., F. Bergerat, M. Bonafede, R. Böðvarsson, S. Crampin, P. Einarsson, K. L. Feigl, C. Goltz, Á. Guðmundsson, F. Roth, R. Sigbjörnsson, F. Sigmundsson, P. Suhadolc, M. Wyss, J. Angelier, Þ. Árnadóttir, M. E. Belardinelli, A. Clifton, L. Dubois, G. B. Guðmundsson, P. Halldórsson, S. Hjaltadóttir, B. Lund, S. Ólafsson, S. Richwalski, C. SensSchönfelder, R. Slunga, A. Tryggvason, K. S. Vogfjörð and B. Þorkelsson 2005. PREPARED - second periodic report. February 1, 2004 - January 31, 2005. Icelandic Meteorological Office-Report 05007, 107 pp. Stefánsson, R. 2006. PREPARED - final report. Icelandic Meteorological Office-Report 06009, 65 pp. Stefánsson, R., M. Bonafede, F. Roth, P. Einarsson, Þ. Árnadóttir and G. B. Guðmundsson 2006a. Modelling and parameterizing the Southwest Iceland earthquake release and deformation process. Icelandic Meteorological Office-Report 06005, 49 pp. Stefánsson, R., F. Bergerat, M. Bonafede, R. Böðvarsson, S. Crampin, P. Einarsson, K. L. Feigl, C. Goltz, Á. Guðmundsson, F. Roth, R. Sigbjörnsson, F. Sigmundsson, P. Suhadolc, M. Wyss, J. Angelier, Þ. Árnadóttir, M. E. Belardinelli, G. Björnsson, A. Clifton, L. Dubois, G. B. Guðmundsson, P. Halldórsson, S. Hjaltadóttir, Á. R. Hjartardóttir, G. Jónsson, M. Khodayar, B. Lund, B. Ófeigsson, S. Ólafsson, S. Richwalski, R. Slunga, P. Theodórsson, K. S. Vogfjörð, B. S. Þorbjarnardóttir and B. Þorkelsson 2006b. PREPARED - third periodic report. February 1, 2005-July 31, 2005. Icelandic Meteorological Office-Report 06008, 131 pp. Stefánsson, R., G. B. Guðmundsson and P. Halldórsson 2008. Tjörnes Fracture Zone. New and old seismic evidence for the link between the North Iceland Rift Zone and the Mid-Atlantic Ridge. Tectonophysics 447, 117–126, doi:10.1016/j.tecto.2006.09.019. Sturkell, E., F. Sigmundsson and P. Einarsson 2003a. Recent unrest and magma movements at Eyjafjallajökull and Katla volcanoes, Iceland. J. Geophys. Res. 108(B8), 2369. Sturkell, E., P. Einarsson, F. Sigmundsson, S. Hreinsdottir and H. Geirsson 2003b. Deformation of Grímsvötn volcano, Iceland:1998 eruption and subsequent inflation. Geophys. Res. Lett. 30, 1182. Sturkell, E., P. Einarsson, F. Sigmundsson, H. Geirsson, H. Ólafsson, R. Pedersen, E. de Zeeuw-van Dalfsen, A. L. Linde, I. S. Sacks and R. Stefánsson 2006a. Volcano geodesy and magma dynamics in Iceland. J. Volc. Geotherm. Res. 150, 14–34. Sturkell, E., F. Sigmundsson and R. Slunga 2006b. 1983–2003 decaying rate of deflation at Askja caldera: Pressure decrease in an extensive magma plumbing system at a spreading plate boundary. Bull. Volc. 68, 727–735. Sturkell, E., P. Einarsson, M. J. Roberts, H. Geirsson, M. T. Guðmundsson, F. Sigmundsson, V. Pinel, G. B. Guðmundsson, H. Ólafsson and R. Stefánsson. Seismic and geodetic insights into magma accumulation at Katla subglacial volcano, Iceland: 1999 to 2005, 2008. J. Geophys. Res. 113(B03212), doi:10.1029/2006JB004851. 100 JÖKULL No. 58, 2008 Sæmundsson, K. 1974. Evolution of the axial rifting zone in northern Iceland and the Tjörnes Fracture Zone. Geol. Soc. Am. Bull. 85, 495–504. Tryggvason, K., E. Huseby and R. Stefánsson 1983. Seismic image of the hypothesized Icelandic hot spot. Tectonophysics 100, 97–118. Vogfjörð, K. S. 2003. Triggered seismicity in SW Iceland after the June 17, Mw = 6.5 earthquake in the South Iceland Seismic Zone: The first five minutes. In: Abstracts from the EGSAGU-EGU Joint Assembly, Nice, France. Vogfjörð, K., S. Th. Rögnvaldsson, R. Slunga, J. Morgan, G. Nolet, R. Allen, P. Erlendsson, S. Ragnarsson, R. Stefánsson, B. Julian and G. R. Foulger 1999. Propagation of seismicity during the September/October 1996 sub-glacial eruption episode near Bárðarbunga Volcano, Iceland. In: Abstracts of the SSA Annual Meeting, Seattle, Washington. Vogfjörð, K. S., H. Geirsson and E. Sturkell 2004. Krísuvíkurhrinan í ágúst 2003: kortlagning brotflata með eftirskjálftum og GPS mælingum. In: Abstracts of the Spring Meeting, Geoscience Society of Iceland 2004. Vogfjörð, K. S., S. Hjaltadóttir and R. Slunga 2005a. The M∼5 triggered events in the South Iceland Seismic Zone on June 17, 2000: Determination of fault plane magnitude and mechanism. In: Abstract of the EGU General Ass., Vienna, Austria. Vogfjörð, K. S., S. Hjaltadóttir and R. Slunga 2005b. Volcanotectonic interaction in the Hengill Region, Iceland during 1993–1998. In: Abstracts, EGU General Ass., Vienna, Austria. Vogfjörð, K. S., S. S. Jakobsdóttir, G. B. Guðmundsson, M. J. Roberts, K. Ágústsson, Þ. Arason, H. Geirsson, S. Karlsdóttir, S. Hjaltadóttir, U. Ólafsdóttir, B. Þorbjarnardóttir, Þ. Skaftadóttir, E. Sturkell, E. B. Jónasdóttir, G. Hafsteinsson, H. Sveinbjörnsson, R. Stefánsson, and T. Jónsson 2005c. Forecasting and Monitoring a Subglacial Eruption in Iceland. Eos Trans. AGU 86, 245 and 248. Wessel, P. and W. H. F. Smith 1998. New, improved version of Generic Mapping Tools Released. Eos Trans. AGU 79, 579. Wolfe, C.-J., I. Th. Bjarnason, J. C. VanDecar and S. C. Solomon 1997. Seismic structure of the Iceland mantle plume. Nature 385, 245–247. Wyss, M. and R. Stefánsson 2006. Nucleation points of recent main shocks in southern Iceland mapped by b-values. Bull. Seism. Soc. Am. 96, 599–608, doi:10.1785/0120040056. Þorbjarnardóttir, B. S., G. B. Guðmundsson and S. S. Jakobsdóttir 2003. Seismicity in Iceland during 2001. Jökull 52, 55–60. Þorbjarnardóttir, B. S. and G. B. Guðmundsson 2003. Seismicity in Iceland 2002. Jökull 53, 49–54. Þorbjarnardóttir, B. S., G. B. Guðmundsson, S. Hjaltadóttir and M. J. Roberts 2007. Seismicity in Iceland during 2006. Jökull 57, 45Ů59.