Intra-arc and back-arc volcano-tectonics: Magma pathways at Holocene Alaska-Aleutian volcanoes

Earth-Science Reviews 167 (2017) 1–26

Contents lists available at ScienceDirect

Earth-Science Reviews journal homepage: www.elsevier.com/locate/earscirev

Invited review

Intra-arc and back-arc volcano-tectonics: Magma pathways at Holocene Alaska-Aleutian volcanoes A. Tibaldi, F.L. Bonali ⁎ Department of Earth and Environmental Sciences, University of Milan-Bicocca, Piazza della Scienza 4, 20126 Milan, Italy

a r t i c l e

i n f o

Article history: Received 20 September 2016 Received in revised form 1 February 2017 Accepted 13 February 2017 Available online 16 February 2017 Keywords: Magma pathway Dyke Stress Earthquake Volcanic hazard

a b s t r a c t The reconstruction of magma pathways at active volcanoes is of paramount importance for the comprehension of their structure and for geohazard assessment. Magma plumbing systems at volcanic arcs may be particularly complicated since the magma rises along fractures that can be consistent with the coeval regional state of stress, the local state of stress, or can form dykes that instead follow pre-existing structures. Magma path orientation can be stable over time or can vary as the consequence of external events like large earthquakes or important modifications in volcano morphology. In order to advance understanding of these issues, we reviewed all available information on the Holocene volcano-tectonics of the Alaska-Aleutian arc and back-arc zones, based on published seismological, interferometric and geological-structural data, geological maps, and official reports. We completed our review with some new measurements of Holocene eruptive fissures, faults, dykes, and morphometric characteristics of pyroclastic cones and volcanic domes aimed at better defining the possible shallow magma paths of the recent-active volcanoes. Finally, we reviewed the possible parameters and models that explain the path configurations. At 32 volcanoes, magma paths strike NW-SE, perpendicular or oblique to the arc but parallel to the regional greatest principal stress. At 20 volcanoes magma paths are parallel to the arc, and 19 volcanoes form rows of coalescent cones that also suggest ascent of magma parallel to the arc. Eight volcanoes display both directions (normal and parallel to the arc), and seismological data indicate that at some volcanoes there has been a rotation of the magma pathway over time. Integration of all data shows that the regional ambient tectonic stress field promotes dyke intrusions normal to the trench. Dykes can also intrude parallel to the trench following stress unclamping from large earthquakes. Trench-parallel dykes and rows of volcanoes can be generated by magma batches that are aligned parallel to the trend of the subduction zone. Once a dyke or a sill is intruded, it locally perturbs the stress field facilitating successive intrusion along a perpendicular direction. © 2017 Elsevier B.V. All rights reserved.

Contents 1. 2. 3.

Introduction . . . . . . . . . . . . . . . Geological and seismo-tectonic overview . . Reconstruction of magma pathway geometry 3.1. Review of the methods . . . . . . . 3.2. Results . . . . . . . . . . . . . . 3.2.1. Adagdak . . . . . . . . . 3.2.2. Akutan . . . . . . . . . . 3.2.3. Amukta . . . . . . . . . . 3.2.4. Aniakchak. . . . . . . . . 3.2.5. Atka and Korovin . . . . . 3.2.6. Bogoslof . . . . . . . . . 3.2.7. Buldir and East Cape . . . . 3.2.8. Chiginagak volcano . . . . 3.2.9. Coats and Yanuska . . . . . 3.2.10. Dutton. . . . . . . . . . 3.2.11. Emmons Lake . . . . . .

⁎ Corresponding author. E-mail address: [email protected] (F.L. Bonali).

http://dx.doi.org/10.1016/j.earscirev.2017.02.004 0012-8252/© 2017 Elsevier B.V. All rights reserved.

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . .

2 3 4 4 5 5 5 5 6 6 7 7 7 7 9 9

2

A. Tibaldi, F.L. Bonali / Earth-Science Reviews 167 (2017) 1–26

3.2.12. Fisher . . . . . . . . . . . . . . . . . . . . . . . . 3.2.13. Fourpeaked. . . . . . . . . . . . . . . . . . . . . . 3.2.14. Frosty . . . . . . . . . . . . . . . . . . . . . . . . 3.2.15. Gareloi . . . . . . . . . . . . . . . . . . . . . . . . 3.2.16. Great Sitkin. . . . . . . . . . . . . . . . . . . . . . 3.2.17. Kagamil . . . . . . . . . . . . . . . . . . . . . . . 3.2.18. Kaguyak . . . . . . . . . . . . . . . . . . . . . . . 3.2.19. Kanaga . . . . . . . . . . . . . . . . . . . . . . . . 3.2.20. Kasatochi. . . . . . . . . . . . . . . . . . . . . . . 3.2.21. Katmai, Novarupta, Trident. . . . . . . . . . . . . . . 3.2.22. Kiska. . . . . . . . . . . . . . . . . . . . . . . . . 3.2.23. Kookooligit . . . . . . . . . . . . . . . . . . . . . . 3.2.24. Iliamna. . . . . . . . . . . . . . . . . . . . . . . . 3.2.25. Imuruk. . . . . . . . . . . . . . . . . . . . . . . . 3.2.26. Ingakslugwat . . . . . . . . . . . . . . . . . . . . . 3.2.27. Little Sitkin . . . . . . . . . . . . . . . . . . . . . . 3.2.28. Makushin . . . . . . . . . . . . . . . . . . . . . . 3.2.29. Mageik . . . . . . . . . . . . . . . . . . . . . . . . 3.2.30. Martin . . . . . . . . . . . . . . . . . . . . . . . . 3.2.31. Moffett. . . . . . . . . . . . . . . . . . . . . . . . 3.2.32. Nunivak . . . . . . . . . . . . . . . . . . . . . . . 3.2.33. Okmok . . . . . . . . . . . . . . . . . . . . . . . . 3.2.34. Pavlof . . . . . . . . . . . . . . . . . . . . . . . . 3.2.35. Redoubt . . . . . . . . . . . . . . . . . . . . . . . 3.2.36. Seguam . . . . . . . . . . . . . . . . . . . . . . . 3.2.37. Segula . . . . . . . . . . . . . . . . . . . . . . . . 3.2.38. Shishaldin . . . . . . . . . . . . . . . . . . . . . . 3.2.39. Semisopochnoi . . . . . . . . . . . . . . . . . . . . 3.2.40. Spurr . . . . . . . . . . . . . . . . . . . . . . . . 3.2.41. St. Paul . . . . . . . . . . . . . . . . . . . . . . . . 3.2.42. Tana . . . . . . . . . . . . . . . . . . . . . . . . . 3.2.43. Tanaga . . . . . . . . . . . . . . . . . . . . . . . . 3.2.44. Ugashik-Mount Peulik . . . . . . . . . . . . . . . . . 3.2.45. Veniaminof . . . . . . . . . . . . . . . . . . . . . . 3.2.46. Vsevidof and Recheshnoy . . . . . . . . . . . . . . . 3.2.47. Westdahl. . . . . . . . . . . . . . . . . . . . . . . 4. Earthquake-induced stress changes . . . . . . . . . . . . . . . . . . . 5. Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.1. Magma pathway geometry . . . . . . . . . . . . . . . . . . . 5.2. Magma paths parallel to the subduction direction . . . . . . . . . 5.3. Magma paths perpendicular to subduction direction and magma flux 5.4. Effect of previous dyke intrusions on magma paths . . . . . . . . 5.5. Effect of major subduction fault slip on magma pathways. . . . . . 5.6. Magma paths at back-arc region . . . . . . . . . . . . . . . . . 6. Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1. Introduction Within the crust, magma rises to the surface through networks of planar structures (intrusive sheets) that form a magma plumbing system. There has been growing consensus on this concept during the last tens of years with increasing field evidence and modelling at various structural levels (see reviews in: Acocella and Neri, 2009; Rivalta et al., 2015; Tibaldi, 2015). Although some debate is still alive for the deepest structural levels, all data indicate that in the shallowest portions of the crust and within volcanic edifices, magma moves along planar fractures: Ground deformation measures, direct observations of eruptions, and the seismicity associated with magma intrusions, provide evidence for the role of tabular sheets (Pollard et al., 1983; Rubin and Pollard, 1987; Peltier et al., 2005; Yamaoka et al., 2005; Aloisi et al., 2006; Mattia et al., 2007). Also field evidence at several eroded volcanoes shows that magma movement takes place along sheets (Tibaldi et al., 2013), both for basic magma and for more felsic magmas (Gudmundsson, 1987, 1988, 1990, 2002; Tibaldi, 2001; Corazzato and Tibaldi, 2006; Pasquarè and Tibaldi, 2007; Tibaldi et al., 2008a,b,

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

10 10 11 11 11 11 11 12 12 12 13 13 13 13 13 13 13 14 14 14 14 14 14 14 15 15 15 15 15 15 15 16 16 16 16 16 16 17 17 18 18 19 21 21 22 23 23

2009a,b). Locally, intrusive sheets may have planar, curved, en-échelon, or more complex geometries, but they invariably have a very high length/thickness ratio. Based on the sheet attitude and the relations with the country rocks, magma can move along vertical to steeplydipping dykes, or inclined sheets and sills, the latter formed when magma is injected between rock layers forming a horizontal or gentlydipping sheet. Volcanic hazard assessment also depends on understanding the structure of magma plumbing systems. Processes occurring at open conduit volcanoes, among which paroxysmal explosions, have been addressed considering the structure of the shallower conduits. For example, Chouet et al. (1997, 2008), at Stromboli Volcano (Italy), studied the wave fields of tremors and explosions showing that the source of these phenomena is located within a planar conduit at a depth b 200 m beneath the summit crater. The modelling of this process along a NEstriking dyke-like conduit is also fully consistent with field observations carried out on outcropping portions of Stromboli's plumbing system (Pasquarè et al., 1993; Tibaldi, 1996, 2001, 2003; Corazzato et al., 2008). Similarly, the assessment of areas most prone to vent and

A. Tibaldi, F.L. Bonali / Earth-Science Reviews 167 (2017) 1–26

eruptive fissure opening depends on correctly understanding the structure of shallow plumbing systems (e.g. Bonali et al., 2011; Tadini et al., 2014). The construction of a plumbing system is a long process that accompanies the growth of a volcano; products from eruptive phases are accompanied and interleaved with sheet intrusions. During the whole history of a volcano, hundreds of sheets may be emplaced, especially for more mafic magmas (Fiske and Jackson, 1972; Walker, 1999). This repetitive process leads intrusions to group, forming sheet swarms that can be vertical (Dieterich, 1988; Carracedo, 1994; Moore et al., 1994; Walter and Schmincke, 2002), or forming ring dykes and cone sheet systems (Tibaldi et al., 2011; Bistacchi et al., 2012) or even more complex arrangements (review in Tibaldi, 2015). These intrusive patterns leave traces at the surface that can be studied by analysing the orientation and location of eruptive fissures, dry fractures and faults, and the morphometric characteristics of pyroclastic cones and vents (e.g. Nakamura, 1977; Tibaldi, 1995; Bonali, 2013). These data can be collected in the field and, especially in remote areas, by way of aerial photos and high-definition satellite images. Ideally, these observations may be integrated with interferometric methods (e.g. Massonnet and Sigmundsson, 2000) and other geophysical studies like seismological ones. Although several attempts have been made to understand magma paths, the results show that magma plumbing systems can be very complicated and can vary at short distances from one volcano to another, and can also vary over time at a given volcano; sometimes they cannot be reconciled with the expected regional stress state (e.g. Wallmann et al., 1990; Roman and Power, 2011; Ruppert et al., 2011; O'Brien et al., 2012; Tibaldi et al., 2017). One of the first studies on magma path orientations at the volcanic arc scale was attempted by Nakamura and Uyeda (1980) and Nakamura et al. (1980) at the Aleutian-Alaska Arc, showing a quite homogeneous pattern. This arc has tens of potentially active volcanoes, and recent eruptions posed serious threats to local communities and to national and international flights, affected by eruptive ash clouds. After 36 years, it is worth reviewing the several advancements that have been made on the volcano-tectonics of this arc and back-arc, in light of the most recent, cutting-edge research on the factors controlling magma paths. In the present work, we review all the published data, from different disciplines, which have looked into the volcano-tectonics of the Holocene Aleutian-Alaska arc and back-arc regions. We sifted through all the available, published seismological, interferometric and geological-structural data, geological maps, as well as official reports. We focused on seismological data such as the distribution of hypocenters and focal mechanism solutions of tectonic and volcano-tectonic earthquakes, and InSAR interferometric data obtained during intrusive events that can provide information on the geometry of the intrusion path. We then reviewed all field data relative to eruptive fissures, faults and dykes associated with Holocene-historic intrusive/eruptive events. Finally, we integrated these data with some measurements of structures and morphometric characteristics of pyroclastic cones and volcanic domes that can be extremely helpful especially in areas of difficult access. The results show a more complex structural pattern than the one proposed in the first studies of Nakamura and Uyeda (1980) and Nakamura et al. (1980). Different trends of magma paths have been recognized, even at a single volcano. In order to shed light on this, we discussed the data also in view of the effect of dyke intrusion on the surrounding local stress field, and the effect of large earthquakes on stress acting on dyke walls. The latter topic is justified by the frequent occurrence of major earthquakes, with Mw N 8, originated at the Aleutian-Alaska subduction zone. This review and its results are useful for a general advancement of the knowledge of the various parameters that may govern magma uprising along a volcanic arc, and also for a more synoptic understanding of the geological-structural characteristics of the Aleutian-Alaska volcanoes, which pose serious geohazards.

3

2. Geological and seismo-tectonic overview The Aleutian-Alaska arc can be divided into two parts: the western insular arc that resulted from the subduction of the Pacific plate beneath the oceanic part of the American plate, and the eastern arc at the continental margin, that resulted from the subduction of the Pacific plate beneath the Mesozoic continental basement of the Alaska Peninsula (Reed and Lanphere, 1973). The arc is 2500 km long from Hayes volcano, west of Anchorage in the Alaska-Aleutian Range, to Buldir Island in the western Aleutian Islands (Fig. 1), hosting over 90 major volcanic centers of Holocene age. At least 29 of these volcanoes had historical eruptions (since 1760 CE) while another 12 volcanic centers are suspected to have had historical eruptions (Miller et al., 1998; Global Volcanism Program, 2013). A total of 265 eruptions were reported since 1760, which should have been originated from these 41 volcanoes. Further 7 volcanic centers in the Aleutian arc have active fumarole fields but no reported historical eruptions (Miller et al., 1998). The Aleutian-Alaska arc is also the most seismically active area in the USA. Since 1900 CE, one major earthquake (Mw ≥ 8) has occurred on average every 13 years (Koehler et al., 2012), one Mw 7–8 earthquake has occurred every two years and six Mw 6–7 earthquakes have occurred per year (Alaska Seismic Hazards Safety Commission, 2012). Tectonic deformation involves also the back-arc region, being distributed over 700 km into the Alaska interior. It is assumed that this widespread deformation zone and the high seismicity are related to the fast plate convergence, in the order of ~5.5 cm/yr, and the gentle subduction angle, especially below the eastern part of the arc (Koehler et al., 2012). As a consequence, there have been multiple, major historic earthquakes: The 1938 Mw 8.2 Shumagin Islands earthquake near the Alaska Peninsula, the 1946 Mw 8.1 Unimak earthquake, the 1957 Mw 8.6 Fox-Andrean of Islands and the 1965 Mw 8.7 Rat Islands earthquakes on the Aleutian Islands, and the 1964 Mw 9.2 Good Friday earthquake on the Alaskan mainland (Plafker, 1969; Beck and Christensen, 1991; Christensen and Beck, 1994; Johnson and Satake, 1994; http://earthquake.usgs.gov/). All these major earthquakes are related to ruptures that were triggered along the shallow section of the gently-dipping thrust interface between the subducting and overriding plates. In particular, west of Kayak Island, the main subduction zone is strongly coupled, as reflected by the occurrence of the 1964 earthquake, which represents the second largest seismic event recorded instrumentally in historic times. The 1964 earthquake (Fig. 1) was accompanied by left-lateral, reverse slip in the order of 20 m, which produced as much as 10 m of uplift of Prince William Sound (Plafker, 1969; Hastie and Savage, 1970). The size of the rupture area was 750 × 200 km (Page et al., 1991). The depth of the hypocenter was restrained to 20 km, although it is sometime regarded as lying between 20 and 50 km (Stauder and Bollinger, 1966). The 1938 event ruptured a 300-km-long segment of the Alaskan arc, which corresponds to the aftershock area (Johnson and Satake, 1994). The 1946 earthquake had a limited aftershock zone, about 100 km in length (Sykes, 1971), but was accompanied by a very strong tsunami (Johnson and Satake, 1997). The 1957 earthquake has the longest aftershock zone of any earthquake ever recorded, in the order of 1200 km, and has been estimated as the third largest earthquake of the 20th Century (Johnson and Satake, 1994). The aftershock area of the 1965 Rat Islands earthquake was 600 km long (Beck and Christensen, 1991). All these earthquakes are compressional, with focal mechanisms of lowangle thrust planes for the 1964 and 1965 earthquakes (Stauder and Bollinger, 1966; Stauder, 1968). The Quaternary crustal faults are known in mainland Alaska and are characterized by the presence of several structures mostly arranged in a curved belt with a northward convex side, north of the volcanic chain (Fig. 1) (Koehler et al., 2012). The longest structures here are the Kaltag and Iditarod-Nixon Fork right-lateral strike-slip faults, striking ENE and NE respectively. More to the east, crossing the Alaska/Canada border, the Denali Fault lies along the southern margin of the Alaskan Range and was the source of the 2002 Mw 7.9 earthquake (Haeussler et al.,

4

A. Tibaldi, F.L. Bonali / Earth-Science Reviews 167 (2017) 1–26

Fig. 1. Active and Holocene volcanoes belonging to the Aleutian-Alaska Arc (red triangles). The black arrow indicates the approximate convergence direction of the plates and estimated velocity; yellow stars represent earthquakes with Mw N 8. Main Quaternary faults are reported (after Koehler et al., 2012). Faults: DF = Denali; KGF = Kaltag; INF = Iditarod-Nixon Fork; AFZ = Amlia Fracture zone (AFZ). Volcano location is from Global Volcanism Program (2013) and Cameron and Nye (2014).

2004, and references therein). This fault has right-lateral strike-slip motion along a plane striking ENE to the west and SE to the east. Closer to the Alaskan volcanic chain, there is a series of Quaternary faults with lengths on the order of tens of km (Fig. 1) (Koehler et al., 2012). These faults strike mostly NNE to NE, parallel to the volcanic alignment, and a few strike N-S. Between the Alaskan volcanic belt and the subduction trench is a swarm of Quaternary faults striking NE in the western part and E-W in the eastern part at the Alaska/Canada border. All such faults are believed to have a reverse kinematics (Koehler et al., 2012). Probably more Quaternary/active faults do exist, also in the intra-arc region, but the remote and rugged terrains of this area have hindered extensive field studies. Previous data on the structures affecting Quaternary volcanoes are derived from studies of the distribution of vents grown on the volcano's slopes (Nakamura, 1977; Nakamura, 1977, Nakamura et al., 1980). These studies indicate that in most of the main Quaternary volcanoes of the Aleutian and Alaskan belt, magma-feeding structures are oriented parallel to the trajectories of the greatest horizontal principal stress (σHmax) that is about NW-SE and correlates with the relative motion between the North American and Pacific plates. Local studies suggest also the presence of NE-striking magma paths and normal faults in the intra-arc zone (e.g. Coats et al., 1961). 3. Reconstruction of magma pathway geometry 3.1. Review of the methods Intrusive sheet swarms are usually expressed at the surface by aligned pyroclastic centers and swarms of eruptive and dry fissures, dykes and faults (Johnson and Harrison, 1990; Strecker and Bosworth, 1991; Annen et al., 2001; Walter and Schmincke, 2002). Dykes form aligned volcanic centers when magma intercepts the surface; several parallel dykes produce corridors containing tens of pyroclastic cones; through the statistical study of the maximum concentration of vents, it is possible to estimate the trend of the underlying sheet swarm (e.g. Nakamura, 1977; Fig. 2). Several statistical methods have been proposed to investigate vent alignment patterns (Lutz, 1986; Wadge and Cross, 1988; Connor, 1990; Lutz and Gutmann, 1995; Germa et al., 2013); however, in order to attribute a series of aligned vents to a single

dyke, it is necessary to combine coeval vents. If the age of vents, or of pyroclastic cones, is not well constrained, errors may affect the accuracy of the results. A more reliable reconstruction of shallow magma paths can be accomplished by studying a number of morphometric parameters, that characterize main craters and parasitic vents (Fig. 3A–B; Tibaldi, 1995; Corazzato and Tibaldi, 2006; Bonali et al., 2011). When a dyke intercepts the volcanic slope, it produces scoria ramparts that are parallel to the dyke, or pyroclastic cones with an elongated base and crater (e.g. Fig. 3A–D), whose azimuth is controlled by magma path strike (Tibaldi, 1995). In order to be able to use the azimuth of the cone base and crater (AbeM and DcM in Fig. 3A-B), the pyroclastic cone must have grown on a substrate with a dip shallower than 7°: In the case of a cone growing on steeper slopes, its elongation is influenced by slope gravity control. Particular attention has been given to the presence of lavas younger than the pyroclastic cone, which can mask the real shape of the cone base. The rim of craters can also be marked by symmetric higher and lower points: the line connecting the two opposite, lower points (Acdp in Fig. 3A-B) is generally parallel to the underlying magma fissure (Tibaldi, 1995), as also shown in Fig. 3D. Fissure eruptions can also produce closely-spaced vents (e.g. Acc in Fig. 3A-C), leading to elongated composite cones with possible different degrees of coalescence (e.g. Fig. 3C and F) (Corazzato and Tibaldi, 2006). All these features are the surface expression of shallow dykes and tend to be perpendicular to the horizontal, least principal stress (σHmin) and parallel to the horizontal greatest principal stress (σHmax). Usually the σHmin coincides with the least principal stress (σ3), whereas the σHmax can be the greatest principal stress (σ1) or the intermediate principal stress (σ2). Fissures and faults can also be directly recognized in the substrate surrounding the volcanoes, providing clues into the style and geometry of the fracture field (e.g. Fig. 3E-F). The morphostructural characteristics of volcanic domes can also provide relevant information (Fig. 4A): Apical depressions (Fig. 4D), the elongation of the base (Fig. 4B) and the alignment of coeval edifices (Fig. 4C) may all reflect the strike of the underlying fractures feeding magma to the surface (Pasquaré and Tibaldi, 2003). Geophysical data recorded during magma injection events may also be helpful: The precise location of volcano-tectonic earthquakes and

A. Tibaldi, F.L. Bonali / Earth-Science Reviews 167 (2017) 1–26

5

If more than one measurement is available for any given magma path (for instance, a number of subparallel dykes), the final orientation is expressed as the mathematical average followed by an error estimate that reflects the data range. 3.2. Results We hereunder describe the data collected for each volcano, in alphabetical order. We examined 91 volcanoes, 5 of which with latest activity in pre-Holocene times, whereas for 42 volcanoes data are not numerous enough to constrain with sufficient precision their potential magma paths. The results are listed in Table 1, together with the location of each volcano and their petrographic types. Examples of the most significant field data are shown in Figs. 5–7, while the whole set of pathways is shown in Fig. 8.

Fig. 2. (A) Sketch 3-D view of radial dykes near the central conduit that, over distance, are reoriented parallel to the horizontal greatest principal stress (σHmax); note vents on opposite volcano flanks (redrawn after Nakamura, 1977). (B) Shaded view of Mount Veniaminof (56°11′53″N, 159°23′27″W) Volcano (Alaska). Black square represents the location of the summit vent of this polygenic volcano, black circles represent vents on volcano flank (after Nakamura et al., 1980). The distribution of flank vents represents the interception of the dyke swarm with the two opposite volcano flanks. Red strip represents the inferred dyke zone (magma path). (C) Theoretical dyke pattern, under a differential horizontal stress, that becomes mostly parallel to the σHmax as at Mount Veniaminof Volcano.

related focal mechanisms can be used to resolve the dominant kinematics/geometry of magma paths. Published interferometric data have also been studied especially in case they were related to dyke injection events. All the available published geological maps and papers dealing with field data have also been reviewed so as to recognize dykes at eroded sectors of volcanoes and eruptive fissures, and to date the various vents/pyroclastic cones. In summary, we based our reconstruction of the azimuth of magma pathways (Table 1) on the following Holocene indicators: i) alignment of closely-spaced vents; ii) alignment of composite craters; iii) preferred distribution of swarms of parasitic and flank cones; iv) trend of elongated vents; v) trend of the base of elongated pyroclastic cones or volcanic domes; vi) trend of crater rim depressed points; vii) trend of dome apical graben; viii) strike of fracture systems; ix) strike of dykes; x) distribution of volcano-tectonic hypocenters; xi) focal mechanism solutions; xii) interferometric fringes. Indicators i) to ix) were determined by integrating available literature and maps (field data), with data from satellite images, morphology from DEM-SRTM90, and 3D view of satellite images on GoogleEarth™. Indicators x) to xii) were derived from the analysis of the literature.

3.2.1. Adagdak Adagdak is a stratovolcano located in the northern sector of Adak Island (51°53′N 176°39′W) (Miller et al., 1998). Although it has long been considered a Pleistocenic edifice (Motyka et al., 1993; Nye et al., 1998), other authors have suggested the most recent eruption to be Holocenic (Marsh pers. comm. in Wood and Kienle, 1990), and Debari et al. (1987) report ultramafic xenoliths found in a post-glacial ejecta deposit emitted by this volcano. There are no data relevant for assessing magma paths at Adagdak volcano; nevertheless, it is possible to reconstruct the local tectonic deformation pattern, and hence state of stress, by analysing the kinematics and geometry of active faults. Several normal faults with downthrow of the hanging-wall blocks dissect Mt. Adagdak; fault planes strike N111° to N137° and dip to the SW (Fig. 6A). The faults show very recent displacements: In fact, they offset water channels and gravity erosive gullies and all surface deposits, producing a series of uphill facing scarps. Offsets of these landforms indicate pure dip-slip motions and thus a N124° ± 13° direction for the active least principal stress σ3. Also Coats (1956a) suggests that these faults are probably recent in age, and have a normal kinematics. However, he comments that they were probably caused by subsidence that occurred when magma was withdrawn from beneath Mount Adagdak. 3.2.2. Akutan The Akutan composite stratovolcano, located 1238 km SW of Anchorage, is one of the most active in the volcanic arc, with numerous eruption during the last two centuries (Global Volcanism Program, 2013). The volcano was affected by a swarm of volcano-tectonic earthquakes in 1996, followed by the opening of N110°-striking ground cracks along its NW flank and by parallel Holocene normal faults that were reactivated in the eastern part (Richter et al., 1998). Interferometric data helped interpret these features as produced by the propagation of a NW-striking, steeply dipping dyke up to within b1 km of the surface of the northwestern flank and the volcano summit (Lu et al., 2000). The intrusion was 2.5 m thick at 0.5 km depth, and its emplacement was accompanied by an earthquake swarm that was teleseismically recorded. During the 1996 crisis, the migration of volcanic-tectonic events suggested a NW-SE trend over the known thermal areas (Miller et al., 1998). These data, together with our observations, enable to hypothesize a vertical dyke plane with a N110° orientation, consistent also with body and surface wave tomography performed by Syracuse et al. (2015). A N120° ± 15° trend was proposed by Nakamura et al. (1980). 3.2.3. Amukta Amukta is stratovolcano located in the central Aleutians, SW of Chaugulak Island; its latest eruption occurred on 3 March 1997. This volcano has had six certain eruptions in historical time from both summit and flank (Global Volcanism Program, 2013) vents. The edifice is affected by a series of pyroclastic cones that grew on its flanks and that bear clear morphometric indicators suggesting N154° trending magma paths. This trend coincides also with the distribution of parasitic cones

6

A. Tibaldi, F.L. Bonali / Earth-Science Reviews 167 (2017) 1–26

Fig. 3. (A) Sketch of morphometric parameters of pyroclastic cones and their relationship with the orientation of the underlying feeding fracture (modified after Tibaldi, 1995). (B) Plan view of a pyroclastic cone with orientation of morphometric parameters: Acc azimuth of the alignment of the center points of coeval centers; Acdp azimuth of the line connecting craterrim depressed points; AceM, AbeM fracture-parallel crater and base elongation; Dbm, DbM minimum and maximum diameter of cone base; Dcm, DcM minimum and maximum diameter of crater (modified after Corazzato and Tibaldi, 2006). (C) Example of two scoria ramparts (NE part of the figure) that can be used as indicators of the strike of the feeding fracture (dashed line), and alignment of two coeval pyroclastic cones (SW part of the figure) lying near the Fisher caldera (54°40′N, 164°23′W). (D) Example of a composite pyroclastic cone with an elongated base and crater, across which a fracture with the same strike can be recognized (white triangles) at the foothills of Pavlof volcano (55°25′10″N, 161°53′42″W). (E) Example of fissures and faults recognized in the substratum surrounding the Kiska volcano (51°57′51″N, 177°27′36″E); note that here fractures strike from N-S to NE-SW and are at high angle, suggesting they can be strike-slip or extensional features. (F) Example of an elongated composite cone and a fracture striking NW-SE at Shishaldin volcano (54°45′19.44″N, 163°58′ 15.96″W).

on the two opposite northwestern and southeastern flanks of the cone, and with a few dykes striking NW-SE and located along the northwestern coastline (Fig. 5A). According to the Global Volcanism Program (2013), six historical eruptions have taken place since 1786 CE.

3.2.4. Aniakchak Aniakchak is a stratovolcano with a summit caldera and intra caldera cones (Lu and Dzurisin, 2014; Bacon et al., 2014), located in the eastern Aleutian arc, 676 km SW of Anchorage (Wallace et al., 2000). At least 40 explosive eruptions have been recognized for this volcano during the past 10 ka (the latest have occurred in 1931 – Global Volcanism Program, 2013; Bacon et al., 2014), whereas the 10-km-wide Aniakchak caldera was formed around 3.4 ka ago (Neal et al., 2000). The caldera contains many pyroclastic cones and lava domes. Vent Mountain and Half Cone are two long-lived vents located inside the caldera floor, whose latest eruption occurred in 1931. Two relatively young domes of coeval age are aligned along a N130° trend, while the northwestern dome's apical structures enable to determine a N126° trend. Pyroclastic cones have morphometric characteristics that provide N110–130° trends, and a 2.3 km long eruptive fissure strikes N109°. The average trend results to be N121° ± 10°.

3.2.5. Atka and Korovin Atka is the largest volcano of the central Aleutians and is made of a central shield volcano and a Pleistocene caldera with several postcaldera centers (Marsh, 1980; Myers et al., 2002): its latest certain eruption occurred in 1812 (Global Volcanism Program, 2013). The most prominent post-caldera stratovolcanoes are Kliuchef and Sarichef, both of which are supposed to have been active in historical time. The summit of Atka is punctuated by a series of craters with a N60° alignment, whereas secondary centers and smaller stratovolcanoes have N147° to N150° elongations. Several dykes are present and strike mostly N155° and secondarily N30° (Myers et al., 2002). Based on all these data we propose two average dominant trends: N151° ± 4° and N45° ± 15°. Korovin stratovolcano is located in the NE part of Atka Island, 1965 km SW of Anchorage; it has erupted eight times since the beginning of the 20th century (Wallace et al., 2000; Global volcanism program, 2013). The volcano is marked by a double summit with two craters located along a NW-SE trending line. A fresh-looking cinder cone lies on the flank of the more eroded Konia volcano that, in turn, is located on the southeastern flank of Korovin. A straight scarp, resembling an east-dipping fault, crosses the summit zones of these cones with a N161° strike. The overall trend of the line connecting all the major and secondary craters is N152°. Miller et al. (1998) show a

A. Tibaldi, F.L. Bonali / Earth-Science Reviews 167 (2017) 1–26

7

Fig. 4. (A) Sketch view of volcanic domes that can show marked characteristics such as apical grabens (crease structures), elongation of the base and alignment of more edifices, which reflect the strike of the underlying fractures feeding magma to the surface. Modified after Pasquaré and Tibaldi (2003). Field examples of: (B) a NW-SE elongated dome (21°50″26.13″ S, 68°9″10.04″W); (C) a NNW-SSE alignment of Holocene domes (21°59″38.25″S, 67°50″0.54″W); (D) a NW-SE elongated dome with the fracture-parallel apical graben (23°5″26.58″S, 67°42″14.60″W). (Examples are from Tibaldi et al. (2009a) and Tibaldi et al. (2017), Central Volcanic Zone, South America).

geological map, based on unpublished data by B.D. Marsh, where a series of dykes affect the area surrounding Korovin volcano with strikes from N150° to N165°. 3.2.6. Bogoslof Bogoslof is stratovolcano located 40 km north of the main Aleutian arc and 1349 km SE of Anchorage (Wallace et al., 2000): It represents the emergent part of a larger submarine volcano that has certainly erupted five times through the last century (Global Volcanism Program, 2013). At this volcano there is a N165° trending alignment of two craters (composing two islands), which correspond to historical eruption sites (Jaggar, 1908). It is also possible to observe a similar elongation of the submarine part of the volcano as evidenced by bathymetric data (Smith, 1937). 3.2.7. Buldir and East Cape Buldir is an eroded stratovolcano located on the southwestern part of the Buldir Island in the western Aleutian arc (Wood and Kienle, 1990). Here, N115–125°-striking normal faults and dykes suggest a possible average N120° ± 5° trend for magma paths. It is worth underscoring that there are no available data suggesting a Holocene activity at this volcano. The younger East Cape volcano, located on the northeastern part of Buldir Island, is a small eroded stratovolcano (Global volcanism program, 2013). East Cape volcano hosts a dome on its southeastern flank that, based on morphological evidence, should be of Holocene age. The dome is elongated in a N100° direction and seems to be affected by a fracture with the same trend. The dome and the summit of the East Cape volcano show a N129° alignment. The whole island is cut by a swarm of normal faults with strikes ranging N115–125° (e.g. Fig. 5C). Based on the average fault strike and on the morphometric parameters of the East Cape volcano and associated dome, we assign a N117° ± 15° trend to the possible magma paths. 3.2.8. Chiginagak volcano Chiginagak is a symmetrical stratovolcano located in the eastern Aleutian arc, approximately 300 km southwest of Anchorage, and is considered hydrothermally active (Wallace et al., 2000; Schaefer et al., 2008). A couple of certain eruptions are reported for this volcano over

the last century (Global Volcanism Program, 2013). It is characterized by a small summit crater, and several lava domes along its NW and SE flanks (Miller et al., 1998). This distribution of flank vents and their morphometric characteristics suggests a N135° trend.

3.2.9. Coats and Yanuska The Coats caldera has a circular shape in plan view, with an average diameter of 3.2 km. It displays a post-glacial morphology and its latest eruption took place on November 2nd, 1937. This caldera seems to be nested within a larger caldera with a diameter of about 10 km (Lamb et al., 1992). The two calderas are located on a quite flat volcano (shield-type) in the eastern part of the Yunaska island (52° 37′ 53″ N, 170° 41′ 49″ W). There are several morphometric indicators inside and outside the smaller caldera. The floor of the smaller caldera is filled with recent lava flows and displays two coeval craters aligned in a N130° direction. On the northern, external caldera flank tens of aligned coeval pyroclastic cones and ramparts can be observed (Fig. 5B). By computing the orientation of all these morphostructures we obtained a N110° trend, that coincides with the strike of fissures suggested by their alignment. A more viscous lava flow on the northern external caldera flank was emitted from a N70° fracture. On the northwestern external caldera flank there are several, partially coalescent pyroclastic cones with morphometric characteristics that enable defining a N103–109° trend. The weighted data suggest a N110° ± 10° trend. Yunaska is a shield volcano with a caldera and associated stratocones, located 1573 km SW of Anchorage, on Yunaska Island (Wallace et al., 2000). It is also defined as a composite edifice made of at least four main volcanic centers (Nicolaysen et al., 1992) that are aligned in NNE-SSW direction. Furthermore, there is a series of aligned craters that punctuate the entire volcanic complex along a N17° direction. Some craters and pyroclastic cones grew on the northern and southern flanks of the major volcanic complex; their morphometric parameters suggest magma paths trending N15° ± 10°.This volcano has certainly erupted 3 times since 1824 CE (Global Volcanism Program, 2013). As a final remark, it is worth noting that the Coats caldera and Yunaska volcano are on the same island and show average magma paths that are perpendicular to each other.

8

A. Tibaldi, F.L. Bonali / Earth-Science Reviews 167 (2017) 1–26

Table 1 Results of the present work on possible magma paths and other characteristics of the studied volcanoes. “Poor” means that data were not sufficiently complete to precisely reconstruct the magma path. Latest eruption, location and rock type are from the Global Volcanism Program (2013) and references therein. Latitude and longitude are in decimal degrees, DATUM WGS84. Nr.

Volcano name

Latitude

Longitude

Main rock type

Magma path (°)

Beginning of the latest eruption

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73

Adagdak Akutan Amak Amukta Aniakchak Atka Augustine Black Peak Bobrof Bogoslof Buldir Buzzard Creek Carlisle Chagulak Chiginagak Chuginadak Churchill Cleveland Coats Dana Denison Douglas Dutton East Cape Emmons Lake Fisher Fourpeaked Frosty Gareloi Great Sitkin Griggs Hayes Herbert Iliamna Imuruk Lake Ingakslugwat Hills Isanotski Kagamil Kaguyak Kanaga Kasatochi Katmai Kialagvik Kiska Koniuji Kookooligit Mt Korovin Kukak Kupreanof Little Sitkin Mageik Makushin Martin Moffett Novarupta Nunivak Island Okmok Pavlof Pavlof Sister Peulik Recheschnoi Redoubt Roundtop Sanford Seguam Segula Semisopochnoi Sergief Shishaldin Snowy Mountain Spurr St. Michael St. Paul Island

51.988 54.134 55.424 52.500 56.880 52.331 59.363 56.552 51.910 53.930 52.348 64.070 52.894 52.577 57.135 52.841 61.380 52.825 52.638 55.641 58.418 58.855 55.183 52.358 55.341 54.650 58.770 55.067 51.790 52.076 58.354 61.640 52.742 60.032 65.600 61.430 54.765 52.974 58.608 51.923 52.177 58.280 57.203 51.964 52.220 63.600 52.381 58.453 56.011 51.933 58.195 53.891 58.172 51.944 58.270 60.020 53.430 55.417 55.457 57.751 53.157 60.485 54.800 62.220 52.315 52.022 51.929 52.050 54.756 58.336 61.299 63.450 57.180

−176.592 −165.986 −163.149 −171.252 −158.170 −174.139 −153.430 −158.785 −177.438 −168.030 175.912 −148.420 −170.054 −171.130 −156.990 −169.756 −141.750 −169.944 −170.640 −161.214 −154.449 −153.542 −162.276 175.940 −162.073 −164.430 −153.672 −162.835 −178.794 −176.130 −155.092 −152.411 −170.111 −153,090 −163.920 −164.470 −163.723 −169.720 −154.028 −177.168 −175.508 −154.963 −156.745 177.460 −175.130 −170.430 −174.166 −154.355 −159.797 178.516 −155.253 −166.923 −155.361 −176.747 −155.157 −166.330 −168.130 −161.894 −161.854 −156.368 −168.539 −152.742 −163.589 −144.130 −172.510 −178.132 179.598 −174.950 −163.970 −154.682 −152.251 −162.120 −170.300

Basaltic Andesite Basalt/Picro-Basalt Andesite/Basaltic Andesite Andesite/Basaltic Andesite Andesite/Basaltic Andesite Basalt/Picro-Basalt Andesite/Basaltic Andesite Andesite/Basaltic Andesite Basalt/Picro-Basalt Trachyand./Basaltic trachyand. Basalt Basalt/Picro-Basalt Andesite/Basaltic Andesite Andesite/Basaltic Andesite Andesite/Basaltic Andesite No data Dacite Andesite/Basaltic Andesite Andesite/Basaltic Andesite Andesite/Basaltic Andesite Andesite/Basaltic Andesite Andesite/Basaltic Andesite Andesite/Basaltic Andesite Andesite Andesite/Basaltic Andesite Andesite/Basaltic Andesite Andesite/Basaltic Andesite Andesite/Basaltic Andesite Basalt/Picro-Basalt Andesite/Basaltic Andesite Andesite/Basaltic Andesite Dacite Andesite Andesite/Basaltic Andesite Basalt/Picro-Basalt Basalt/Picro-Basalt No data (checked) No data (checked) Dacite Andesite/Basaltic Andesite Basalt/Picro-Basalt Andesite/Basaltic Andesite Andesite/Basaltic Andesite

Poor N110 Poor N154 N121 N151/N45 Poor Poor Poor N165 N120 Poor Poor Poor NW-SE Poor Poor Poor N110 Poor Poor Poor NW-SE N120 N151/N60 N60 N00 N00 N140 N135 Poor Poor Poor NNW-SSE NW-SE N95 Poor N150 N150 Radial N-S? N70 Poor N40 Poor N100 N156 Poor Poor NW-SE NW-SE/N45 N135 NW-SE N147 N140/N70 N97 N143 N40 Poor N135 Poor NW-SE/NNE-SSW Poor Poor N105/N121 N135 Poor/radial Poor N135 Poor NNW-SSE N130 N73

Holocene Historical (1992) Historical (1796) Historical (1997) Historical (1931) Historical (1812) Historical (2005) Historical (1900 BCE ± 150 ys) Holocene Historical (1992) Pleistocene–Holocene (?) Historical (1050 BCE (?)) Historical (1828) Not precisely known Historical (1998) No data Historical (847 ± 1 y) Historical (2016) Historical (1937) Historical (1890 BCE (?)) Not precisely known Holocene Holocene Pleistocene-Holocene Pleistocene/Holocene Historical (1830) Historical (2006) Holocene Historical (1989) Historical (1974) Historical (1790 BCE ± 40 ys) Historical (1200 ± 300 ys) Holocene Historical (1876) Historical (0300) Holocene Holocene Historical (1929) Historical (3850 BCE (?)) Historical (2012) Historical (2008) Historical (1912) Holocene Historical (1990) Historical (1150 BCE ± 1900 ys) Pleistocene to Holocene Historical (2006) Holocene Historical (1987) Historical (1828) Historical (500 BCE ± 50 ys) Historical (1995) Historical (1953) Historical (1600 BCE (?)) Historical (1912) Pleistocene-Holocene Historical (2008) Historical (2016) Holocene (Unc) Historical (1814) Holocene Historical (2009) Historical (7600 BCE ± 500 ys) Holocene Historical (1993) Holocene Holocene–Historical (1873) Post-Miocene Historical (2014) Historical (1710 ± 200 ys) Historical (1992) Holocene Historical (1973)

Andesite/Basaltic Andesite Basalt/Picro-Basalt Basalt/Picro-Basalt Andesite/Basaltic Andesite Andesite/Basaltic Andesite Andesite, Basalt, Dacite Andesite/Basaltic Andesite Andesite/Basaltic Andesite Dacite Andesite/Basaltic Andesite Rhyolite Basalt/Picro-Basalt Basalt/Picro-Basalt Andesite/Basaltic Andesite Basalt/Picro-Basalt Andesite/Basaltic Andesite Andesite/Basaltic Andesite Andesite/Basaltic Andesite Rhyolite Andesite/Basaltic Andesite Andesite/Basaltic Andesite No available data Andesite/Basaltic No data Basalt/Picro-Basalt Andesite/Basaltic Andesite Andesite/Basaltic Andesite Basalt/Picro-Basalt Trachybasalt/Tephrite Basanite

A. Tibaldi, F.L. Bonali / Earth-Science Reviews 167 (2017) 1–26

9

Table 1 (continued) Nr.

Volcano name

Latitude

Longitude

Main rock type

Magma path (°)

Beginning of the latest eruption

74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91

Steller Stepovak Bay 2 Stepovak Bay 3 Stepovak Bay 4 Takawangha Tana Tanaga Trident Ugashik Ukinrek Maars Uliaga Unnamed Veniaminof Vsevidof Westdahl Wrangell Yantarni Yunaska West

58.430 55.913 55.929 55.954 51.873 52.830 51.885 58.236 57.751 57.832 53.065 57.870 56.170 53.130 54.518 62.000 57.019 52.610

−154.400 −160.041 −160.002 −159.954 −178.006 −169.770 −178.146 −155.100 −156.368 −156.510 −169.770 −155.411 −159.380 −168.693 −164.650 −144.020 −157.185 −170.778

Andesite/Basaltic Andesite Andesite/Basaltic Andesite Andesite/Basaltic Andesite No data (checked) Basalt/Picro-Basalt Rhyolite Basalt/Picro-Basalt Andesite/Basaltic Andesite Andesite/Basaltic Andesite Basalt/Picro-Basalt No data No data Andesite/Basaltic Andesite Andesite/Basaltic Andesite Basalt/Picro-Basalt Andesite/Basaltic Andesite Andesite/Basaltic Andesite Andesite/Basaltic Andesite

Poor Poor Poor Poor Poor N90 N70/N145 N140/N70 Poor Poor Poor Poor N145 Poor N120 Poor Poor N17

Holocene Holocene Holocene Late-Pleistocene/Holocene Historical (1550 (?)) Holocene Historical (1914) Historical (1974) Historical (1814) Historical (1977) Holocene Unc. Historical (2013) Historical (1878) Historical (1991) Historical (2002) Historical (800 BCE ± 500 ys) Holocene, no historical eruptions

3.2.10. Dutton Mount Dutton is a stratovolcano with a diameter of about 5 km, located east of Cold Bay, which consists of a summit multiple dome complex (Davies et al., 1988; Wallace et al., 2000; Global Volcanism Program, 2013). The geometric distribution of the dome complex cannot be used to assess magma paths because a volcano sector collapse towards the west destroyed the original vent arrangement. However, a swarm of shallow earthquakes that occurred beneath Mt. Dutton in 1984, July–August 1988, and 1991 (Smithsonian Institution, 1988; Miller et al., 1998; Miller et al., 1999) form a roughly linear zone, trending NW-SE, that extends from the western shoulder of the volcano towards the SE, for about 10 km (Miller et al., 1999). Even though the earthquake swarms weren't followed by any eruptions, the seismic events have been interpreted as evidence of magma movement at shallow depths beneath the volcano, resulting in a dyke intrusion of mafic

magma with associated fracturing of the country rock (Miller et al., 1999). The latest known eruption occurred prior to 10,000 years ago (Global Volcanism Program, 2013). 3.2.11. Emmons Lake The Emmons Lake Volcanic Center (ELVC) is located southwest of Pavlof volcano and north of Volcano Bay (Froese et al., 2002; Mangan et al., 2003; Global Volcanism Program, 2013). It is a large stratovolcano (Waythomas et al., 2006) made of ~ 350 km3 of Middle PleistoceneHolocene lavas arranged into a 30-km-long row of nested calderas and overlapping stratovolcanoes. The ELVC experienced up to five major caldera-forming eruptions. Holocene eruptions are mostly made of basaltic-andesites to andesites; from a historic point of view, at least 40 eruptions have occurred over the last 200 yr at Pavlof Volcano, that lies at the NE end of this volcanic row together with Pavlof Sister

Fig. 5. (A) Dykes striking NW-SE crop out along the northwestern coast of Amukta volcano (52°29′39.08″N, 171°15′17.14″W); this provides evidence for its shallow plumbing system. (B) The external flank of Coats caldera (52° 38' 16.80"N, 170° 38′ 23.99″W) has several eruptive fissures and aligned and elongated pyroclastic cones whose trend ranges N103–130°. (C) Most of Buldir Island (52°21′29″N, 175°55′29″E) is affected by a swarm of normal faults (red arrows) with strikes in the range N115–125°. (D) The 1912 Novarupta crater (58°16′ 0″N, 155°9′24″W) area is affected by a set of fissures striking N140° (red arrows), subparallel to the dominant bedrock joint set, and another set striking N70°. All images are from Google Earth™.

10

A. Tibaldi, F.L. Bonali / Earth-Science Reviews 167 (2017) 1–26

Fig. 6. (A) Mt. Adagdak is affected by a series of N111° to N137°, very recent normal faults creating a series of uphill facing scarps. (B) Series of N97°-striking eruptive fissures at Nunivak Island. (C) Pyroclastic cones aligned along an eruptive fissure at Seguam volcano. (D) NW-striking dykes cropping out southeast of Seguan volcano and (E) northwest of it. (F) At Westdahl volcano, morphometric data indicate a N120° trend, as can be noticed, for example, by observing the aligned craters.

volcano (see the distinct description of Pavlof and Pavlof Sister further on). Geochemical and geophysical observations suggest that, although all eruptions at the ELVC derive from a common basalt parent in the lower crust, magma follows one of two closely spaced, but distinct paths to the surface (Mangan et al., 2009). Under the northeastern end of the row, magma moves rapidly through a relatively young (~ 28 ka), plumbing system, made of a connected dyke plexus. Below the southwestern part of the row, magma moves through a long-lived (~ 500 ka) and complex plumbing system with several zones of magma mingling and fractioning (Mangan et al., 2009). Here there is the Emmons Lake Volcano that shows a main alignment of summit craters along a N60° direction. Nevertheless, the flank centers, mainly pyroclastic cones and craters, have morphometric characteristics indicating N141° to N161° trends. 3.2.12. Fisher Fisher is a stratovolcano with a large caldera located on the western part of the Unimak Island, northeast of Westdahl volcano (e.g. Miller and Smith, 1977; Wallace et al., 2000); it was formed about 9100 years ago (Miller et al., 1998). Six eruptions took place during the Holocene and the latest occurred in historical time (August 1830) (Global Volcanism Program, 2013). The edifice is strongly elliptical in plan view, measuring 10 × 15 km (e.g. Bindeman et al., 2001) and a N35° long axis. GPS measurements performed by Mann and Freymueller (2003) between 1998 and 2001 show deformation patterns that are best fitted by the deflation of a dyke-like source at a

shallow depth. The dyke is 14 km long, it strikes N35° and dips 80° to the NW. However, this solution is not fully constrained and the same GPS data can be explained in terms of a deeper and horizontal sill (Mann and Freymueller, 2003).Our morphometric analyses indicate a preferential N60° orientation of magma paths, based on the observation of several ramparts and pyroclastic cones (e.g. Fig. 3C). Taking all data into account, an average trend of N48° ± 12° can be proposed. 3.2.13. Fourpeaked Fourpeaked is a stratovolcano located in the northeastern part of Katmai National Park (eastern arc). Its first historical eruption occurred in September 2006, whereas no Holocene activity has been identified (Doukas and McGee, 2007; Global Volcanism Program, 2013).The vent area consists of nine craters or pits, perfectly aligned along a line trending north from the summit northern flank of the cone, which heavily crevassed and disrupted the glacial-ice cover (Neal et al., 2009). This N-S-trending, 1250 m long fissure produced minor ashfall as consequence of a single phreatomagmatic eruption. Based on seismicity, gas data, and phreatic activity that took place during the 2006 event, Gardine et al. (2011) suggested that new magma moved through the upper 10 km of crust beneath Fourpeaked volcano. Gas pooled in crack pathways at about 5 km depth and then breached the crust into the shallowest levels opening the way to further magma uprise. Seismicity clustered along preferred NNE-SSW to N-S trends located below the northern side of the volcano summit, mostly coinciding with the trend of the fracture and vents in the ice cover.

A. Tibaldi, F.L. Bonali / Earth-Science Reviews 167 (2017) 1–26

11

Fig. 7. Example of distribution of parasitic vents on two opposite flanks of a volcano. The Kookooligit shield volcano (63°35′38″N, 170°22′56″W) hosts more than one hundred pyroclastic cones (in red) on the eastern and western flanks. The land strip marked by most of the vents, the elongation of craters and cone bases, as well as their alignments, are consistent with possibly N100°-oriented magma paths.

3.2.14. Frosty Frosty Volcano is located on the Alaska Peninsula southwest of the village of Cold Bay (Global Volcanism Program, 2013); it is the youngest of two large volcanic structures of the Cold Bay volcanic complex. To the south, there is the glacially dissected, late Pliocene-early Pleistocene Morzhovoi volcano (Waldron, 1961). To the north, there is the late Pleistocene-Holocene Frosty Peak stratovolcano, which grew within a likely caldera structure. The N-S trend obtained by using the alignment of these two volcanoes coincides with the azimuth of two craters located on top of the Frosty Peak edifice, which in turn are N-S aligned with the Mt. Simeon small lava volcano located further north. Although a number of lava flows are well preserved, no Holocene eruptions have been officially reported for this volcano (Global Volcanism Program, 2013).

3.2.15. Gareloi Gareloi is a stratovolcano located at the western end of the Andreanof Islands; it is the northernmost volcano of the Delarof Group (Wallace et al., 2000; Global Volcanism Program, 2013). This volcano has been one of the most active in the Aleutian Arc since its discovery during the 1740 Bering expedition; twelve certain eruptions have been reported since 1790 CE (Global Volcanism Program, 2013). However, due to its remote location, the available data are very scanty. Our morphometric observations indicate a N140° trend for the possible magma paths. In April 1929, a phreatic eruption opened up a new crater just below the southern summit; the crater is strongly N170°-elongated with a 1600 m maximum diameter (Miller et al., 1998). Further explosions during the 1929 eruption resulted in the formation of another 12 coeval craters on the southeast flank of the volcano, from 80 to 1600 m in diameter, which show a clear N140° elongation and alignment (Coombs et al., 2008). This eruption was undoubtedly produced by a fissure striking NNW-SSE in the upper part of the cone and then bending to attain a NW-SE direction.

3.2.16. Great Sitkin Great Sitkin is a stratovolcano with a caldera, located near the central sector of the Aleutians, about 25 miles northeast of the Adak Island (Simons and Mathewson, 1955; Wallace et al., 2000). It is characterized by the presence of a series of five domes of similar age along the northwestern flank, aligned along a N135° direction (Waythomas et al., 2003). They are also aligned with the domes that were emplaced in the summit area during the eruptions of 1945 and 1974 (Waythomas et al., 2003). These domes were formed after a flank collapse of the volcano and within its depression, and thus their emplacement may have been facilitated by the removal of the rock mass as a consequence of the failure (Tibaldi, 2004; Acocella and Tibaldi, 2005; Tibaldi et al., 2008a, 2008b; Vezzoli et al., 2008). Nevertheless, their strong rectilinear alignment points to a possible control by dyking along the preferential NW-SE trending zone of weakness. According to the Global Volcanism Program (2013) six certain eruptions have happened here since 1792 CE. 3.2.17. Kagamil Kagamil island (52° 59′ 35″ N, 169° 42′ 43″ W – 1504 km SW of Anchorage (Wallace et al., 2000)) hosts the eponymous stratovolcano in its southern sector as well as another, older volcano in its northern sector (Miller et al., 1998). Kagamil Volcano features a number of partially eroded craters in the summit zone, which are aligned along a N150° trend. In the northwestern part of the island there is a series of N-S-striking dykes. Just one eruption is reported for this volcano in historical times and the Holocene: It happened in 1929 CE (Global Volcanism Program, 2013) 3.2.18. Kaguyak Mount Kaguyak is a volcano complex with a 2.5 × 3 km-wide caldera located in the NE part of Katmai National Park (Global Volcanism Program, 2013). Two post-caldera lava domes are present within the

12

A. Tibaldi, F.L. Bonali / Earth-Science Reviews 167 (2017) 1–26

Fig. 8. Orientation of magma pathways (black bars in the map) reconstructed for each volcano. Rose diagrams show azimuths for the whole set of magma pathways, for the volcanoes located along the volcanic arc, and for the volcanoes located in the back-arc area. Volcanoes are reported in Table 1.

caldera, whereas several pre-caldera domes can be observed. The precaldera edifice was not a stratovolcano but was made of nine lava dome clusters (Fierstein and Hildreth, 2008). Four extra-caldera clusters are mid-to-late Pleistocene, and the other five clusters are b60 ka. Most of these clusters were fed by N150°-striking dykes (Fierstein and Hildreth, 2008). Two eruptions are reported in historical times (Global Volcanism Program, 2013) 3.2.19. Kanaga Kanaga stratovolcano is a 1307 m-high symmetric, composite cone with a circular shape at sea level; it is the most notable feature of Kanaga Island, 1951 km southwest of Anchorage (Andreanof Islands) (Wallace et al., 2000). A circular summit crater, approximately 200 m in diameter, hosts several active fumaroles. The most significant eruption in historical times was observed in 1906, when four lava flows were outpoured from the summit zone (Coats, 1956b); this eruption was possibly preceded by another one, in 1904. The andesitic flows extend from fissures located on the south, southwestern, and northeastern flanks of the Kanaga cone: Hence, they can be regarded as radial fissures. For this volcano, 17 confirmed eruptions are reported in historical times (Global Volcanism Program, 2013). 3.2.20. Kasatochi Kasatochi Island represents the emergent part of a predominately submarine volcano. The only known eruption happened in 2008; however, previous eruptions may have occurred here in historical times (Waythomas et al., 2010). A study of the seismic swarm associated

with the 2008 eruption, with special regard to earthquake locations and source parameters, showed that the first earthquakes occurred along a N-S direction, either directly underneath the volcano, or within 5–10 km south of it (Ruppert et al., 2011). After the occurrence of a MW 5.8 jolt, the earthquakes occurred in a new crustal volume slightly east and north of the previous events. Ruppert et al. (2011) argue that, should the N-S trend of earthquake epicenters indicates the dyke strike, this would correspond to a direction subparallel to the σHmin. However, dyke strike should be parallel to the σHmax. Ruppert et al. (2011) regard this possibility as unlikely and hence propose that the seismicity cluster defines the region surrounding a magma diaper or a volume of critically stressed crust. 3.2.21. Katmai, Novarupta, Trident Mt. Katmai and Mt. Trident, together with the secondary center, Novarupta, belong to the 65-km-long SW-NE-trending Katmai Volcanic Cluster (KVC), located about 440 km southwest of Anchorage (Wallace et al., 2000). The KVC includes, from the southwest, the major Martin, Mageik, Trident, Katmai, Snowy and Steller stratovolcanoes. Katmai is a stratovolcano with a caldera, while Novarupta is a pyroclastic vent with a plug dome and Mt. Trident is a cluster of stratovolcanoes with a dome complex (Wallace et al., 2000). The activity within the KVC shifted gradually from NE to SW over time (Hildreth et al., 2001). Lu et al. (1997), by using synthetic-aperture-radar interferometry of 1993–95 images, reported 7 cm of uplift beneath the southwestern sector of Mt. Trident. Lu et al. (1997) interpreted the apparent ground deformation to reflect inflation of a pressure source at 0.8- to 2-km depth,

A. Tibaldi, F.L. Bonali / Earth-Science Reviews 167 (2017) 1–26

which they attributed to either magma intrusion or pressurization of the hydrothermal system. Wallmann et al. (1990) described a set of fissures crossing the margins of the 1912 Novarupta crater and striking N140°, subparallel to the dominant bedrock joint set (Fig. 5D). These authors suggest that fissures and joints, local plate-motion vector and the inferred regional stress orientation are consistent with a N140°trending feeder dyke that propagated from a reservoir beneath Mt. Trident volcano towards the eruptive Novarupta vent, The contemporaneous creation of Mt. Katmai caldera represents a possible magmatic connection between the reservoir beneath this volcano and Mt. Trident; this connection has been interpreted as a propagating sill or the failure of septa between N70°-aligned batches of magma (Wallmann et al., 1990; Lowenstern et al., 1991; Hildreth and Fierstein, 2000). It worth noting that in the area there is another set of joints, striking N70°, exactly parallel to the line connecting Mt. Katmai to Mt. Trident. All these data suggest two possible shallow magma paths in the area: a N140° one affecting Novarupta and Mt. Trident, and a N70° affecting all three volcanoes. Regarding the eruptive activity, only one eruption has been reported for Novarupta in historical times (in 1912), as well as for Katmai, whereas at Mt. Trident volcano, 15 eruptions have happened since 1900 (Global Volcanism Program, 2013) 3.2.22. Kiska Kiska is a stratovolcano located on Kiska Island, 1898 km SW of Anchorage (the westernmost of the Rat Islands) (Wallace et al., 2000). We assigned a N50° trend for magma paths at Kiska volcano, based on the distribution and morphometric characteristics of pyroclastic cones (Fig. 3E) and the elongation of the main cone. Here, also Nakamura et al. (1980) assigned a magma path trend of N50° and geological research by Coats et al. (1961) showed the presence of fractures in the surrounding substratum, striking about N30° and N120°; for a few of these it has been possible to individuate normal fault movements. We also recognized high-angle faults striking from N-S to NE-SW. Based on these data, we assigned to Kiska a magma path of N40° ± 10°. This volcano has experienced four certain eruptions since 1900 (Global Volcanism Program, 2013). 3.2.23. Kookooligit The Kookooligit Mountains are represented by a PleistoceneHolocene shield volcano with a N97° major axis; the volcano is located in the north-central part of St. Lawrence Island, in the Bering Sea (63°24′ N, 170°10′W) (Patton et al., 2011), (Fig. 7). The basaltic shield is 30 × 40 km wide and is punctuated by N 100 pyroclastic cones. Most of the cones are elliptical; an analysis of their morphometric characteristics, together with the observation of the two zones of cone clustering, trending N98° on the two opposite E and W flanks of the shield volcano, enable to assess an average N100° ± 10° trend. The Global Volcanism Program (2013) has not reported any Holocene eruptions in the Kookooligit area. 3.2.24. Iliamna Mt. Iliamna is a stratovolcano located 219 km SW of Anchorage, Aleutian Range, in Lake Clark National Park (Wallace et al., 2000). Two main earthquake swarms occurred beneath Iliamna volcano in May and August 1996. Roman et al. (2004a) located 88 and 2833 events respectively, at a depth of 1–4 km b.s.l. Increases in SO2 and CO2 were coincident with the second swarm but, at the end, no eruptions occurred. The hypocenters of the swarms form a NNW-SSE elongated cluster, south of Mt. Iliamna's summit. Fault-plane solutions calculated for 24 earthquakes at the top of the August 1996 cluster, indicate normal and strike-slip faulting with p-axes parallel to the axis of the regional σHmax (Roman et al., 2004a). A similar conclusion was reached also by Statz-Boyer et al. (2009) after precise relocation of about 3000 earthquakes. More recently, a study of seismicity extended to the period May 1996–June 1997, favours two possible scenarios involving either a NW-SE-striking dyke or a N-S-striking one (Roman and Power,

13

2011). We wish to underscore that, although in previous papers the authors suggested possible preferred orientations of the crater zone of this volcano, data are not sufficiently numerous to propose any preferential trend on the basis of field morphometry or geological structures. In regard to eruption activity, the Global Volcanism Program (2013) reports seven certain eruption as having occurred here since 5050 BCE. 3.2.25. Imuruk The Oligocene-to-Holocene Imuruk monogenetic volcanic field, situated in the central Seward Peninsula north of the Bendeleben Mountains, features around 75 small basaltic vents surrounded by voluminous lava flows (Hopkins, 1959, 1963). The largest and most recent vent is the Lost Jim cone, a 30-m-high cinder cone near Imuruk Lake that produced the only Holocene lava flow in the Imuruk field. This massive lava flow, which was erupted about 1655 years ago, extends 35 km west and 9 km north of the vent, covering about 230 sq. km. The late-Pleistocene Camille lava flow, the second youngest in the volcanic field, flowed as far as 39 km from its vent. For this volcanic field, a NW-striking magma pathway is suggested here, based on normal fault geometry and Holocene cone alignment (Hopkins, 1959, 1963); however, the same authors found out minor faults and minor cone alignments striking also NE-SW. 3.2.26. Ingakslugwat Ingakslugwat Hills are 40 cinder cones and craters spread across N500 km2; this is one of the most recently active areas in the YukonKuskokwin delta region of SW Alaska (Coonrad, 1957; Hoare and Condon, 1971). A first definition of pathway direction for this volcano was given by Nakamura (1977), who inferred an about N70–75° direction based on eruptive fissures and cones distribution. As a result of our analysis, we slightly changed this direction, up to N95°.The latest activity is considered to have occurred during the Holocene (Wood and Kienle, 1990). A low cone, containing a 400-m-wide lake, may be a maar. 3.2.27. Little Sitkin Little Sitkin is an active stratovolcano with a nested caldera (Snyder, 1959; Wallace et al., 2000), located on the eponymous Island – Rat Islands – (located 1963 km SW of Anchorage). One relatively recent flow originated from a fissure along the western trace of the older caldera boundary fault (Snyder, 1959). This vent is located on the northwestern flank of the volcano, along with other parasitic vents, suggesting possible dominant NW-SE magma paths. Two confirmed historical eruptions (1776 and 1828) are reported in the Global Volcanism Program (2013). 3.2.28. Makushin A possible N135° trend for magma paths can be inferred for Makushin volcano, based on morphometric characteristics of secondary vents. Anomalies of seismic high b-value (Bridges and Gao, 2006) and InSAR data (Lu et al., 2002) suggest the presence of a magma chamber located 5 km east of the volcano summit. The close correspondence between InSAR data and the seismic anomaly has been explained by Bridges and Gao (2006) in terms of the presence of highly fractured rocks above and around the magma chamber, due to applied pressure from the magma chamber expansion. A plausible interpretation is that the NW-SE-elongated pattern of earthquake epicenters from July 1996 to August 2000, with hypocenter cluster dipping to the NE shown by Lu et al. (2002), as well as the presence of the magma chamber at a depth of 7 km (displaced 5 km laterally from the main summit vent), suggest a planar conduit that dips 35° to the NE. This is consistent with the N135° trend determined by way of morphometric observations. Field data show the presence of several dykes, in the upper volcano flanks, which mostly strike NW-SE, secondarily NE-SW, and N-S in one case (McConnell et al., 1998). NE-SW-striking dykes are more abundant

14

A. Tibaldi, F.L. Bonali / Earth-Science Reviews 167 (2017) 1–26

in the older volcanic stratigraphic units (Tertiary), whereas NW-SEstriking dykes are more frequent in the Quaternary deposits (McConnell et al., 1998). Based on all the aforementioned data, we suggest that the more recent plumbing system is dominated by about N135°-striking dykes. 3.2.29. Mageik Mageik is a typical example of a stratovolcano that bears evidence of shallow magma paths with two different orientations. It is located 452 km SW of Anchorage (Wallace et al., 2000). The volcano summit features several craters that can correspond to two alignments, respectively NW-SE and N55° oriented. The morphometric characteristics of single craters indicate a N35° trend. In summary, the data suggest two possible trends with the same probability: NW-SE and N45° ± 10°. Jolly and McNutt (1999) found that during 1995–1997, an intrusive event with magmatic degassing occurred below the Martin-Mageik volcanoes; however, it has not been possible to define magma path geometry, although seismicity was elongated WSW-ENE below the two edifices, similarly to what had been found by Matumoto and Ward (1967), who analysed 1800 microearthquakes below the Katmai Volcanic Cluster. Eight, certain eruptions, are reported for this volcano since 8670 BCE ± 300 years (Global Volcanism Program, 2013). 3.2.30. Martin Mt. Martin is a stratovolcano located 458 km SW of Anchorage, near the southwestern end of the SW-NE-trending Katmai Volcanic Cluster (KVC) (Wallace et al., 2000). O'Brien et al. (2012) calculated 134 double-couple fault plane solutions (FPS) for a pre-swarm period (Jan 2004–Dec 2005), a swarm period (January 2006), and a post-swarm period (Feb 2006–Dec 2008); they also analysed the temporal changes in FPS orientations by means of directional statistics. O'Brien et al. (2012) showed significant changes in the degree of FPS heterogeneity and the mean P-axis azimuth between the pre-, syn-,and post-swarm FPS orientations. The local stress field was heterogeneous with no well-defined preferred P- and/or T-axis orientations during the pre-swarm period. The P-axis orientations became aligned to a NNE-SSW trend (i.e. about perpendicular to regional σ1) during the 2006 swarm. Finally, the Paxis orientation became vertical during the post-swarm period. Normal faulting events dominated during the minor earthquake swarms in 2007 and 2008, defining a different state of stress/kinematics with respect to the major 2006 swarm. O'Brien et al. (2012) also showed that the January 2006 P-axis orientations are consistent with a model of shallow injection of magma into a NW-SE-striking dyke located at ~ 0–3 km depth beneath Mt. Martin's summit. Four certain eruptions are reported for this volcano since 1750 BCE, two of which happened after 1900 CE (Global Volcanism Program, 2013). 3.2.31. Moffett Mt. Moffett stratovolcano is located in the northern part of Adak Island together with another two eruptive centers (Mt. Adagdak and the informally designated “Andrew Bay volcano”, Waythomas, 1995). Mt. Moffett is a composite volcanic cone of Quaternary age and is not known to have been active in historical times (Wood and Kienle, 1990); however, it is considered to have had eruptions during Holocene times (Waythomas, 1995). Mt. Moffet features a number of vents aligned in a N147° direction, comprising parasitic vents on the northwestern and southeastern flanks and one more eroded peak, corresponding to an older summit carter, and the younger vent. It is worth noting the presence of several, recent normal faults affecting Mt. Adagdak, with strikes ranging N111° to N137° (Fig. 6A). For this volcano, three historical (1600 BCE–7850 BCE) eruptions are reported by the Global Volcanism Program (2013). 3.2.32. Nunivak Nunivak Island is entirely made of pahoehoe lava flows punctuated by sixty cinder cones and four maars (Hoare et al., 1968; Settle, 1979).

The lava flows originate from small shield volcanoes, some of which show an about E-W elongation. Radiometric dating suggests that basaltic volcanism was active in the eastern part of Nunivak Island between 0.70 Ma and 0.15 Ma (Mukasa et al., 2007). Analysis of morphometric values, especially relative to parallel fissure eruptions (e.g. in Fig. 6B) and pyroclastic cones, enables to obtain a N97° trend. In the Global Volcanism Program (2013), no Holocene eruptions are reported at Nunivak Island. 3.2.33. Okmok Okmok is a central shield complex with a nested caldera located on Umnak Island, 1397 km SW of Anchorage (Wallace et al., 2000). Morphometric characteristics do not allow determinining a definitive dominant direction for possible magma paths. This complex situation has been underscored also by Mann et al. (2002), who used geodetic data associated with the 1997 eruption and argued that the plumbing system of this volcano may be transient, as evidenced by the presence of multiple cones along the caldera rim. Based on GPS data, Fournier et al. (2009) suggested that the plumbing system consists of a central magma chamber located below the center of the caldera. Magma rises to the surface along ring fractures inside the caldera, with inclined paths and circular in plan view. More recently, Haney (2010) suggested that the source of VLP tremors at Okmok caldera can be explained by the presence of a vertical, N143°-striking crack-like conduit. Masterlark et al. (2010) applied FEM modelling with the purpose of reconstructing stress state variations around a magma chamber and compared this with InSAR data, concluding that an inclined sheet is consistent with the stress orientation. For this volcano 18 certain eruptions are reported in the Global Volcanism Program (2013) since 6310 BCE. 3.2.34. Pavlof Pavlof is stratovolcano located near the end of the Alaska Peninsula, 953 km SW of Anchorage (Wallace et al., 2000); it is one of the most active volcanoes in North America (McNutt and Beavan, 1981; McNutt and Jacob, 1986) with N30 eruptions since 1900 (Global Volcanism Program, 2013). The volcano lies along a line of vents extending NE from the ELVC (Global volcanism program, 2013). We were able to define a N40°-oriented trend for magma paths at this volcano, on the basis of the distribution and morphometric characteristics of pyroclastic cones (as shown in Fig. 3D), in agreement with the observations of Nakamura (1977) who based his results on edifice elongation. 3.2.35. Redoubt Redoubt is a stratovolcano located 175 km SW of Anchorage, Aleutian Range, Lake Clark (Wallace et al., 2000). At this volcano, the interpretation of the geometry of the plumbing system is particularly complex. Based on the inversion of fault plane solutions of 420 earthquakes below the volcano, Sánchez et al. (2004) showed that, during the 1989–90 eruptions, the σ1 and σ2 were typically sub-horizontal, respectively trending ESE-WNW and NNE-SSW, with a near-vertical σ3. During July 1991–January 1998, the same authors found a nearvertical σ1 and sub-horizontal σ2 and σ3, respectively striking SSENNW and ENE-WSW. This was explained in terms of the initial expansion of a plexus-like magma body, with no-preferred orientation, followed by the 1991–98 period of non-magmatic activity with very little horizontal stress beneath the volcano. In this period the vertical load was the largest principal stress and the σHmax of tectonic origin was σ2 with a SSE-NNW trend. Lahr et al. (1994) showed that an earthquake swarm from the 1989–90 eruption had locations consistent with activity occurring along a narrow conduit dipping 75° to the NE, and in the depth range of 1 to 7 km. Roman and Gardine (2013), studying the shear-wave splitting during the 2005–2006 period of quiescence at Redoubt, found a stress field characterized by a NW-SE σHmax that should represent the tectonic stress. The analysis of fault-plane solutions of 210 volcano-tectonic events indicates periods of complex intrusions in

A. Tibaldi, F.L. Bonali / Earth-Science Reviews 167 (2017) 1–26

2009, followed by a period with a stable dominant NE-SW-trending Paxis orientation. This trend is perpendicular to the regional σHmax and is also characterized by shallow earthquakes located directly below Redoubt's summit; it has been interpreted as the effect of horizontal displacement of the conduit walls in response to intruding magma (Roman and Gardine, 2013). After that, plane solutions indicate again a period of stable NW-SE-trending P-axis that represents the re-establishment of post-eruptive effect of tectonic stress. A magma path should be perpendicular to the syn-intrusive P-axis and thus we infer that the most probable dyke should strike NW-SE, hence being consistent with one of the two solutions proposed by Buurman et al. (2013). A plexus of dykes and sills was recognized beneath the northern volcano flank between 1 and 6 km below sea level, and a conduit zone extends from this area to below the volcano summit (Benz et al., 1996). The same authors showed the distribution of 930 volcano-tectonic earthquakes that took place in July 1991; these, in plan view, suggest a N15°E trend for the conduit zone. Based on all these data it is possible to conclude that the tectonic σHmax trends between NNW-SSE and NW-SE and magma pathways can strike NW-SE and also NNE-SSW; moreover, they can alternate over time. Several eruptions are reported in the Global Volcanism Program (2013) since 9310 BCE. 3.2.36. Seguam The Seguam stratovolcano of the Smithsonian Institute Catalogue corresponds to the Seguam Island, that hosts several volcanoes (Global Volcanism Program, 2013); there are two late Quaternary calderas, spaced 8 km apart, and each caldera features one Holocene cone (Jicha and Singer, 2006). The eastern caldera affects the Wilcox volcano; further east, there is another Holocene cone along the eastern shore of the island, known as Moundhill volcano. Although the simple alignment of the main cones may suggest a N75° trend for magma paths, actually a more precise information can be derived from the observation of eruptive fissures, dykes and morphometric parameters. In the western part of the island, the fissure from which lava was erupted in 1977 has a N146° strike (Jicha and Singer, 2006) and originated at the southern rim of the western caldera. Inside this caldera there is a historically active cone, named Pyre Peak, which represents the highest point on the island, with an elevation of 1054 m. The summit crater area shows two depressed points of the crater rim, indicating a N97° trend. The whole surrounding area, west of the volcano, hosts older volcanic rocks cut by several dykes (e.g. Fig. 6D–E). They strike mostly N121–142°, with a few scattered N55–90° and an average trend of N119°. In the central part of the island, different morphometric parameters at several pyroclastic cones allow to define a N100° trend, and eruptive fissures of dacite and rhyodacite lavas dated to b8.4 ka BP (Jicha and Singer, 2006) strike N100–105° (Fig. 6C). In the southeastern part of the island, in correspondence of the Wilcox volcano lavas dated at 98–49 ka BP (Jicha and Singer, 2006), there is a series of dykes striking N110°. InSAR data from 1993 to 2000 allowed to recognize the presence of point pressure magma sources at a depth of 2–4 below the southern margins of the western and eastern caldera (Price, 2004). In summary, the data indicate an average N105° trend in the central-western part of the island and an average N121° trend in its eastern part. Eleven confirmed eruptions are reported since 7300 BCE (Global Volcanism Program, 2013). 3.2.37. Segula The Segula stratovolcano is located east of Kiska volcano (Nelson, 1959; Wood, 1992; Global Volcanism Program, 2013) and, as suggested by Wood and Kienle (1990), it shows a series of flank cones whose morphometric characteristics allow us to assign a N135° trend to possible magma paths. 3.2.38. Shishaldin Shishaldin is a stratovolcano located on Unimak Island, 1098 km SE of Anchorage (Wallace et al., 2000); the volcano is characterized by

15

the presence of a swarm of pyroclastic cones that grew on its NW flank, most of which have morphometric features compatible with the emplacement of dykes striking N135° ± 5° (e.g. Fig. 3F). Moran et al. (2002) studied a sequence of tectonic earthquakes preceding the April–May 1999 eruption and simulated a dyke oriented according to this strike. This preferred orientation for magma feeding paths was also considered by Nakamura et al. (1980). A total of 38 confirmed eruptions are reported in the Global Volcanism Program (2013) since 7550 BCE. 3.2.39. Semisopochnoi Semisopochnoi is the largest subaerial stratovolcano with intracaldera cones (Wallace et al., 2000) in the western AleutiansSemisopochnoi Island (Global Volcanism Program, 2013). For this volcano, we could not find any information useful for defining magma pathway orientations (e.g. Coats, 1959). Two historical eruptions are reported for this volcano, respectively in 1873 and 1987 CE (Global Volcanism Program, 2013). 3.2.40. Spurr Mount Spurr is a stratovolcano characterized by an active vent (Crater Peak) and a caldera; it is located 127 km SW of Anchorage (Wallace et al., 2000). The seismic pattern preceding and accompanying the 1992 eruption at this volcano is described in Power et al. (1992). During the time interval October 1981–July 1991, representing the background normal seismicity, most of the activity concentrated at depths b3 km beneath the volcano summit (Crater Peak), with very few events at depths b5 km. During the precursory period, between August 1991 and June 1992, the activity concentrated along a NNE-SSW belt crossing the caldera and reaching down to major depths. During the eruptive period, July–December 1992, the seismicity propagated towards the SSW and from depths in the range 38–0 km. By way of tomographic inversion, Power et al. (1998) showed an old conduit below Mt. Spurr summit and confirmed the NNE-SSW dyke below Crater Peak; however, more recently the same authors revised the model of the shallow portion, suggesting a pipe-like conduit dipping to the SE and extending from the reservoir upward to the surface (Power et al., 2002). Koulakov et al. (2013), based on the distribution of the Vp/Vs values, recognized a conduit reaching the surface in correspondence of the vent of the 2004–2005 unrest, and the second one, locked at a depth of 20 km, beneath the same location where an explosive eruption occurred in 1992. Seven confirmed eruptions are reported for this volcano in the Global Volcanism Program (2013) since 6050 BCE. 3.2.41. St. Paul St. Paul Island, located in the Bearing Sea, contains several well preserved cinder cones and small basaltic-to-trachybasaltic shield volcanoes (Barth, 1956; Feeley and Winer, 1999). A more recent series of lava flows originated from a N70°-trending row of vents in the western part of the island, whereas another N75° row is present in the southern part. An E-W-striking series of fractures/fissures is present in the central part of the island. In the northern part, there are several pyroclastic cones with morphometric characteristics suggesting an average N56° trend. All the data enable to obtain an average N73° trend. Volcanism started about 400 ka ago (Mukasa et al., 2007); the recentmost confirmed eruption, reported for this volcano in the Global Volcanism Program (2013), occurred in 1280 BCE. 3.2.42. Tana The Tana volcanic complex, on Chuginadak Island, is composed of at least two prominent E-W-trending volcanoes (Wood and Kienle, 1990; Global Volcanism Program, 2013). The eastern volcano shows evidence of glacial erosion and looks older; the western edifice features two possible coalescent craters, the observation of which enables to obtain a N95° trend. The summits of the two volcanoes are connected by an EW depression. Along the western foothills of the Tana volcanic complex

16

A. Tibaldi, F.L. Bonali / Earth-Science Reviews 167 (2017) 1–26

there are at least seven youthful pyroclastic cones, and another one lies immediately west of the isthmus between Cleveland and Tana volcanoes. Five monogenic cones have morphometric characteristics that suggest a N90° trend, whereas another three cones, located further west, indicate a N135° trend. Based on available information, the Global Volcanism Program (2013) has not reported any eruptions during the Holocene. 3.2.43. Tanaga Tanaga is a stratovolcano complex made of three main cones, located in the Andreanof Islands, 2003 km SW of Anchorage (Wallace et al., 2000; Coombs et al., 2007). The main cones are aligned ENE-WSW and, although they have different ages, the morphometric characteristics of the single craters allow to assign a N70° trend to possible magma paths. The westernmost cone, named Sajaka, shows one main Holocene sector collapse to the NW and another successive minor collapse in the same direction, interleaved by reconstruction of the cone within the collapse depression. The craters of the youngest cone are aligned NE-SW and each single crater is elongated in the same direction; however, this trend can be explained as being influenced by the effect of gravitational instability towards the NW. A number of dykes found in the southern part of the Tanaga complex strike from N130° to N160° (Jicha et al., 2012), suggesting that the area was subjected to distinct plumbing systems, which were not active at the same time. Tanaga is regarded as historically active (Miller et al., 1998), having erupted six times since 1050 BCE (Global Volcanism Program, 2013). 3.2.44. Ugashik-Mount Peulik The Ugashik-Mount Peulik volcanic complex, located 533 km SW of Anchorage, is considered as a stratovolcano with summit and flank domes (Wallace et al., 2000; Hildreth et al., 2007). It composed of the late Quaternary 5-km-wide Ugashik caldera and the Peulik stratovolcano, which grew on the northern flank of Ugashik. Peulik has been entirely formed in Holocene times and erupted in 1814 and 1845 (Miller, 2004). The morphometry and location of some satellite centers, with respect to the volcano summit (mostly lava domes), indicate a N135° trend for magma pathways at Peulik volcano. Data for Ugashik caldera are instead too scanty. Four confirmed eruptions are reported for this volcano since 6550 BCE (Miller et al., 1998; Global Volcanism Program, 2013). 3.2.45. Veniaminof Veniaminof is a stratovolcano with an 8-km-diameter, ice-filled summit caldera, located 7854 km SW of Anchorage (Wallace et al., 2000). Holocene pyroclastic flow and lahar deposits, as young as 3700 yr BP, surround the volcano (Miller and Smith, 1987; Bacon et al., 2003). InSAR data indicate a source of magma inflation located in the depth range 3–17 km below the caldera (Fournier and Freymueller, 2008), but they are not sufficient to geometrically constrain a possible dyke orientation at shallower levels. Nevertheless, tens of pyroclastic comes are concentrated along a NW-SE land strip across the volcano (Fig. 2B); their morphometric characteristics indicate a N145° ± 10° average trend. N 20 eruptions are reported for this volcano in historical times, since 1750 BCE (Global Volcanism Program, 2013) 3.2.46. Vsevidof and Recheshnoy Vsevidof is a stratovolcano located on Unmak Island, 1443 km SW of Anchorage (Wallace et al., 2000). Recheschnoi is another stratovolcano located ENE of a roughly 900-m-high saddle across from Vsevidof volcano (Global Volcanism Program, 2013). Nakamura et al. (1980) assigned to Vsevidof and Recheshnoy volcanoes a N65 ± 20° and a N 110° ± 10° trend for magma pathways, respectively, Based on our observations, available information and consequent evaluation, we believe that the data are not numerous enough to allow the definition of possible preferential orientations.

No Holocene eruptions are reported for Recheschnoi volcano in the Global Volcanism Program (2013), whereas 3 confirmed eruptions occurred at Vsevidof volcano during 1800–1900 CE. 3.2.47. Westdahl At Westdahl Volcano, regarded as a possibly pyroclastic cone on a truncated ancestral stratovolcano or shield volcano (Miller et al., 1998; Wallace et al., 2000), morphometric data indicate a N120° trend, as can be seen for example at three coeval and aligned centers in Fig. 6F. It is located 1147 km SW of Anchorage. By using radar and optical satellite imagery, Lu et al. (2003, 2004) showed that the lava flows of the 1991–1992 eruption were emitted from a system of fissures striking from E-W to WNW-ESE, the latter orientation being consistent with our morphometric observations. GPS data acquired from 1998 to 2001 show active deformation linked to a central source at 7.2 km depth beneath the caldera (Mann and Freymueller, 2003). Seven eruptions have occurred since 1795 CE (Global Volcanism Program, 2013) 4. Earthquake-induced stress changes Several authors suggested that large subduction earthquakes could trigger eruptions due to induced stress changes in the hanging-wall block (Linde and Sacks, 1998; Hill et al., 2002; Marzocchi, 2002; Marzocchi et al., 2002; Manga and Brodsky, 2006; Walter and Amelung, 2007; Watt et al., 2009; Delle Donne et al., 2010; Bebbington and Marzocchi, 2011). This is more evident in the nearfield (b 250 km from the epicenter) (Eggert and Walter, 2009) and within a few days after the earthquake (Linde and Sacks, 1998; Manga and Brodsky, 2006; Eggert and Walter, 2009); we need to stress that the delay earthquake/eruption can be from seconds to years, depending on the complexity of volcanic systems (Linde and Sacks, 1998; Nostro et al., 1998; McLeod and Tait, 1999; Walter and Amelung, 2007; Eggert and Walter, 2009). For example, Watt et al. (2009) suggested that the overall eruption rate in the Southern Volcanic Zone of the Andes increased after the two Chilean subduction earthquakes of August 1906 (Mw 8.2) and May 1960 (Mw 9.5); Bonali (2013) suggested that the 2010 Mw 8.8 Chile earthquake was capable of promoting eruptions up to three years after the event due to crustal extension perpendicular to magma pathway (unclamping). Three principal modes for earthquake-induced stress transfer capable of promoting eruptions have been proposed: static, quasi-static, and dynamic stress changes (Hill et al., 2002; Marzocchi et al., 2002; Manga and Brodsky, 2006). The static stress change is the difference in the stress field from just before an earthquake to shortly after the seismic waves have decayed (Hill et al., 2002). It may explain processes leading to eruption in regions close to the fault rupture by changing the state of stress on magma pathways (e.g., Walter and Amelung, 2007; Bonali, 2013; Bonali et al., 2013; Bonali et al., 2015; Bonali et al., 2016) and up to five years after an earthquake (Marzocchi, 2002) (Fig. 9). From the point of view of magma pathways, much in-depth research has been conducted to understand how a large subduction earthquake could influence magma rise along structures (Sepúlveda et al., 2005; Walter, 2007; Bonali, 2013; Bonali et al., 2013, Bonali et al., 2015). Particularly, Sepúlveda et al. (2005) observed that a major earthquake is capable of influencing the structural mechanism of eruptions. The Cordon Caulle volcano, in fact, erupted after the great Mw 9.5 Chile earthquake of May 22, 1960, due to a dyke intrusion oriented normal to the direction of the subduction earthquake. The same authors suggested that this dyke intrusion was favoured by volumetric expansion and consequent unclamping for pathway with a suitable orientation. At the arc scale, Bonali et al. (2015) demonstrated that earthquake-induced magma pathway unclamping in the Southern Volcanic Zone, close to the fault rupture of the Mw 8.8 2010 Maule earthquake, was greater for pathways striking subparallel to the trench (N-S) than for those striking perpendicular (E-W) to the trench. Such observation suggests that also in this case,

A. Tibaldi, F.L. Bonali / Earth-Science Reviews 167 (2017) 1–26

17

earthquake, and occurs over a period of years to decades (Marzocchi, 2002; Freed and Lin, 2002; Marzocchi et al., 2002), favouring fluid upwelling (Chaussard et al., 2013). Lupi and Miller (2014) proposed a new mechanism related to quasi-static stress change where postseismic relaxation of the compressional stress regime initiates shortlived strike-slip kinematics in the volcanic arc (Fig. 12). Finally, dynamic stress associated with the passage of seismic waves is often proposed as a possible eruption trigger, also over great distances (Linde and Sacks, 1998; Manga and Brodsky, 2006; Delle Donne et al., 2010; Namiki et al., 2016), due to nucleation, growth, and ascent, of the volatile phases through shaking. Fig. 9. Sketch summarizing triggering mechanisms by earthquake-induced static deformation (after Walter and Amelung, 2007). A large earthquake along the subduction slip plane induces different stress changes at different rock volumes: the upper crustal wedge is characterized by compression towards the trench (blue area) and extension below the volcanic arc (red area). This crustal extension favours pathway unclamping, dyke formation and propagation, associated with processes like volatile exsolution, density and viscosity decrease, bubble growth, as well as mingling of different magmas.

dyke intrusion is encouraged in a direction perpendicular to the fault slip and parallel to the trench. Such finding is also valid for Alaska, as shown in Figures10-11, in which we have modelled the earthquake-induced effect of the 1964 Mw 9.2 Good Friday earthquake in terms of static stress change normal to NW-SE and NE-SW vertical dykes (Bonali et al., 2015), instead of modelling the volumetric expansion. Such directions are respectively perpendicular and parallel to the trench as well as are parallel and perpendicular to the overall maximum horizontal stress. The numerical models were performed using the Coulomb 3.3 software (Lin and Stein, 2004; Toda et al., 2005); calculations were made in an elastic half-space with uniform isotropic elastic properties following Okada's (1992) formulae. The upper crust was modelled with a Young's modulus E = 80 GPa and a Poisson's ratio ν = 0.25 based on King et al. (1994), Mithen (1982), Lin and Stein (2004) and Toda et al. (2005); a lower value of Young's modulus would only have the effect of reducing the magnitude of static stress changes. A finite fault model used to simulate the earthquake-induced effects is based on tsunami wave form inversion (Table 2). The input stress field is based on data reported in the World Stress Map database (Heidbach et al., 2010). The subvolcanic magma feeding systems that link the volcanoes with their magma reservoirs are assumed as vertical surfaces for all volcanoes. The stress changes across the area of interest have been calculated at a representative depth of 2 km below the volcanic arc, based on the fact that previous analyses show minimum variations of stress change with depths in the range of possible magma chambers (e.g. Bonali et al., 2015). Quasi-static stress change is associated with slow viscous relaxation of the lower crust and upper mantle beneath the epicenter of a large Table 2 Characteristics of the Mw 9.2 earthquakes and related finite fault model: date of the event, magnitude, fault rupture length, strike and dip angle of the fault plane, number of patches, averaged fault slip and rake angle, depth of top and bottom of the fault plane, trend and plunge of σ1, σ2 and σ3 (Johnson et al., 1996; Ichinose et al., 2007; Harding and Algermissen, 1969). Name Date (dd/mm/year) Mw Rupture length (km) Fault geometry (strike/dip, °) N. of patches Average slip (m) Average rake angle (°) Fault top (km) Fault bottom (km) σ1 (azimuth/plunge, °) σ2 (azimuth/plunge, °) σ3 (azimuth/plunge, °)

Good Friday 28/03/1964 9.2 ~8001; 540 ÷ 740 216 ÷ 237°/6 ÷ 12° 95 3.8 83.0 5.2 46.8 165°/35° 333°/0° 243°/63°

5. Discussion 5.1. Magma pathway geometry An integrated review of geophysical and geological-structural data allowed us to reconstruct magma path orientations at 49 volcanoes (out of 91), totalling 57 pathway directions (8 volcanoes have a double orientation), whereas for 42 volcanoes data are not sufficient to precisely assess magma paths. The data are uniformly distributed over the whole Aleutian-Alaska arc and back-arc (Fig. 8), enabling a discussion at the local scale as well as in terms of regional volcano-tectonics. Most volcanoes show volcano-tectonic rift zones cutting across the edifice, following Nakamura's (1977) model. A total of 40 volcanoes show shallow magma paths trending NW-SE, 6 volcanoes have a NESW trend, and 3 volcanoes a N-S trend. Seven of the 40 volcanoes display both the NW-SE and NE-SW trend. As a comparison, data from the World Stress Map Project (Heidbach et al., 2010), mainly based on focal mechanism solutions of the tectonic earthquakes, together with data from boreholes and faults, indicate that the horizontal greatest principal stress σHmax trends between NNW-SSE and NW-SE across the whole arc (Fig. 13). This is consistent with the model of plate motions that indicates a relative NW-SE convergence across the AleutianAlaska volcanic arc, and with the direction of subduction that tends to be NNW-SSE towards the eastern part of the arc, and NW-SE towards its western part (DeMets et al., 1994; Buurman et al., 2014).The same conclusions have been reached also by modelling of Quaternary faults and GPS data (Finzel et al., 2011) and further geodynamic modelling (Finzel et al., 2014). These data suggest that at the studied volcanic arc, magma mostly intrudes along preferential directions parallel to the regional NNW- to NW-trending σHmax; however, magma can rise to shallow levels also along a direction that is perpendicular to the σHmax. Since σHmax coincides with σ1, most frequently dilation for dyke injection takes place along a direction parallel to σ3 or to σ2. On the other hand, in the case of magma intrusion along a direction parallel to the volcanic arc, dyking occurs perpendicularly to the σ1. Considering that recent normal faulting occurred along planes parallel to these two directions, it has to be concluded that two trends of possible dilation might coexist, at least locally, in the studied volcanic arc. The explanations of these trends will be discussed later. The fact that at several volcanoes it has not been possible to reconstruct magma paths derives, in part, from the scarcity of studies due to the remoteness of the areas. However, it is worth pointing out that the reconstruction of the plumbing system, even at shallow levels, represents a very difficult task. Plumbing systems can be very complicated, as they are made of a plexus of intrusive sheets with different orientations. As it is well known, volcanoes can exhibit non-preferred orientation of magma paths, such as in the case of magma radial forces dominating over tectonic forces (Nakamura, 1977; Gudmundsson, 1984, 1988, 1995, 2006; Gautneb and Gudmundsson, 1992; Gudmundsson and Loetveit, 2005). An example of radial magma paths may be represented by the four lava flows that were outpoured from fissures near the summit of the Kanaga (51°55′24″N, 177°10′05″W) cone on the south, southwest, and northeast flanks (Coats, 1956b).

18

A. Tibaldi, F.L. Bonali / Earth-Science Reviews 167 (2017) 1–26

Another complication may derive from the presence of several volcanoes that were affected by one or more caldera collapses. A caldera may be marked by a complex morphostructure, with several monogenic edifices (pyroclastic cones or lava domes) inside the depression and on its outer flanks. At some of the studied calderas, morphometric indicators of cones within the collapse depression have various orientations, that make it difficult to assess a definitive dominant trend. This is coherent with the fact that, frequently, caldera formation occurs through failure of the subsiding block in several secondary blocks separated by fractures. These fractures may guide magma uprising during subsequent unrest, without a specific control by remote stresses. Some volcanoes also suffered flank or sector collapses, which also perturb the magma plumbing system due to unbuttressing favoured by the removal of the collapsed rock volume (Tibaldi, 2001; Tibaldi et al., 2009b). Both caldera and lateral collapses may destroy pre-existing eruptive fissures and centers, hence decreasing the amount of information needed for reconstructing shallow magma paths. Another consequence is that post-collapse centers and fissures may locate following the local stress field that has been modified by the new morphology of the cone (reviews in: Acocella and Neri, 2009; Tibaldi, 2015).

5.2. Magma paths parallel to the subduction direction NNW- to NW-striking dykes and normal faults are coherent with the regional trend of the σHmin that ranges between NE-SW and ENE-WSW (Figs. 13 and 14). The NW-SE-striking structures can develop along the volcano slopes at distance from the central plumbing system, as exemplified by the March 1996 intense swarm of volcano-tectonic earthquakes beneath Akutan Island (54°07′41″N, 165°55′05″W). A dyke intrusion produced extensive ground cracks in the north-western sector of the island along the lower volcano slope (Lu et al., 2000). The reactivation of Holocene NW-striking normal faults in the eastern part of the island was interpreted as the effect of magma intrusion beneath the volcano and increased extensional strain and/or decreased pore pressure (Lu et al., 2000). This kind of intrusion is coherent with Nakamura's (1977) model, which postulates that, at some distance from the central conduit, dykes tend to intrude parallel to the regional σHmax. Another example of Nakamura's (1977) model can be seen at Gareloi Volcano (51°47′18″N, 178°47′39″W); the eruption of April 1929 started with the opening of a fracture in the uppermost portion of the cone, producing a N170°, very elongated crater (Miller et al., 1998). The downslope propagation of this fracture was accompanied by formation of another 12 craters and by the re-orientation of fractures, that bended, becoming parallel to the regional subduction direction, which here is NW-SE. Based on the Nakamura et al.'s (1980) model, the σHmin that favours the emplacement of the dykes parallel to plate convergence, is represented by σ3. In a convergence setting like the crustal wedge above a subduction zone, since σ1 is horizontal too, this implies the development of conjugate strike-slip faults oblique to the convergence direction and T fractures parallel to it (Fig. 15A). Nakamura et al.'s (1980) model hence predicts the emplacement of volcanoes along the T fractures. However, focal mechanism solutions across the Aleutian-Alaska arc provide more indications of strike-slip motions occurring along faults parallel to the arc, especially in its western sector (Fig. 14B) (e.g. Ekström and Engdahl, 1989; Lallemant and Oldow, 2000). Based on Lallemant and Oldow (2000), also field geological-structural data and Global Positioning System geodesy indicate arc-parallel extension of the Aleutian volcanic arc linked to displacement partitioning with arc-parallel transcurrent faults and arc-perpendicular normal faults and dykes. This may be a plausible mechanism for explaining the along-arc extension required for emplacement of NW-SE-striking dykes. Along-arc extension (i.e. oriented NE-SW) with normal faults has been detected also by focal mechanism solutions along the eastern volcanic arc by Ruppert et al. (2012). This is consistent with the required along-arc extension for emplacement of NW-SE-striking dykes.

If σHmin = σ2 and σ3 is vertical, reverse faulting occurs as observed at several sites along the Aleutian-Alaska arc (Fig. 14C). In this case, the theory suggests that magma propagates in the form of a sill, because the axis of minimum buttressing is vertical. During sill inflation, the magma produces a force on the sheet walls capable of modifying the surrounding local stress state. This is consistent with several authors who have suggested that local stress reorientations may be a consequence of magma intrusions (Parsons and Thompson, 1991; Vigneresse et al., 1999; Kühn and Dahm, 2008; Geoffroy et al., 1993; Roman et al., 2004b; Lehto et al., 2010; Chaput et al., 2014). We argue that sill inflation creates a local stress increase in a direction normal to the sill wall, i.e. along the vertical stress axis. Since σ3 is the vertical stress in a reverse tectonic setting, sill inflation may lead to a switch between σ2 and σ3, the latter attaining a horizontal position (Fig. 15B). Successive magma inflation leads to the emplacement of a dyke perpendicular to the trench, following the new local stress state. In alternative, magma can rise along reverse faults also in a tectonic setting marked by horizontal σ1 and σ2, as already observed, for example, in the Ecuador (Tibaldi, 2005, 2008) and Japan (Acocella et al., 2008) volcanic arcs. The NE-striking dykes and normal faults instead developed perpendicularly to the regional σ1. This implies different possible explanations: i) the upper crustal magma upwelling follows a trend guided by the deep zone of magma production; ii) previously intruded dykes alter the local state of stress; iii) some effect of large slip along the subduction zone. We will discuss these scenarios in the following sections. 5.3. Magma paths perpendicular to subduction direction and magma flux The site of magma upwelling is associated with the magma production zone that, in turn, depends upon various factors mainly linked with pressure (P) and temperature (T) conditions. Due to the inherent geometry of a subducting plate, the deep zone of magma generation is elongated and follows the linear zone where P-T conditions are reached for fluid production. Magma upwelling is thus originally focused in zones elongated parallel to the trench. In case the uppermost crust is mostly dissected by fractures parallel to the arc, magma batches are favoured to follow this trend passively instead of creating new fractures perpendicular to the arc (i.e. NNW- to NW-striking fractures in the Aleutian-Alaska case). This can be mechanically explained as follows: Magma can intrude a new fracture when this forms at the tip of a dyke if magmatic pressure (pm) exceeds the lithostatic pressure (pl), plus the host rock tensile strength, plus the horizontal compressive stress in the host rocks perpendicular to the dyke (Gudmundsson, 1995, 2012). If the host rock is characterized by the presence of fractures inherited from previous deformation events, these planes have no cohesion (or very poor cohesion for sealing effect). In this case magma propagates along a preexisting fracture if magmatic overpressure (po = pm − pl) exceeds the compressive stress acting perpendicular to that plane. Since along the Aleutian-Alaska arc we have also magma paths parallel to the arc, this implies that the magmatic overpressure can be locally larger than the tectonic stress acting normal to the fracture planes parallel to the arc. At Redoubt volcano for example, Sánchez et al. (2004) based on the inversion of fault plane solutions of 420 earthquakes, showed that during the 1991–98 period of nonmagmatic activity, the σ1 was vertical and the σHmax = σ2 = NW-SE. Thus the magmatic stress sufficient to intrude and dilate the NEstriking planes must have been N σ2. An analysis of the distribution along the arc of the volcanoes with arc-parallel magma path, shows that only three of these are located in the western arc (Atka (52° 19″ 51.24″N, 174° 8″ 20.40″W), Kiska (51°57′51″N, 177°27′36″E), Yunaska (52°37′53″N, 170°41′49″W)), whereas the other nine are located in the central and eastern sections of the arc. The presence of coalescent stratovolcanoes, which compose rows of volcanoes parallel to the arc, is also more diffuse eastwards. From a tectonic point of view, there are some relevant differences along the Aleutian-Alaska arc. Oblique subduction dominates the

A. Tibaldi, F.L. Bonali / Earth-Science Reviews 167 (2017) 1–26

19

and Engdahl, 1989). This change in regional seismicity occurs in correspondence of the Amlia Fracture zone (AFZ) (Fig. 1), which is a bathymetric high at N173°W on the Pacific Plate that subducts below the central Aleutian Islands (Ryan et al., 2012).The location of the AFZ coincides also with the change from dominant shallow hypocenters of volcano-tectonic (VT) earthquakes in the west to deep VT seismicity in the east. All these data, together with the presence of a larger number of volcanoes and of several coalescent rows of volcanic cones in the central-eastern part of the arc, are consistent with less coupling to the east where magma is able to ascend more easily through the crust. A more voluminous magmatic flux and consequently more frequent deep VT earthquakes have been already suggested for the region east of the AFZ by Buurman et al. (2014). 5.4. Effect of previous dyke intrusions on magma paths

Fig. 10. 1964 Earthquake-induced normal stress change resolved on (A) NE-SW and (B) NW-SE-striking vertical magma pathways. Depth = 2 km below the volcanic arc. White Triangles locate volcanoes. Red colours represent a normal static stress reduction on the receiver plane (unclamping), blue colours represent an increase (clamping). Unclamping is positive for convention.

western part of the arc, with partitioning of fault motions above the subduction zone (McCaffrey, 1992, 2002). Here, the tectonics of the intraarc zone and the surroundings is mostly expressed by strike-slip motions (Ekström and Engdahl, 1989; Lallemant and Oldow, 2000) and clockwise rotation of a number of blocks (Geist et al., 1988; Krutikov et al., 2008). On one side, this produces the local formation of crustal normal faults, leaving extensional basins at the trailing edges of rotating blocks (Spence, 1977; Geist et al., 1988), as well as shear zones. The result is the presence of a horizontal σ3 that facilitates magma upwelling along a direction that is different from the convergence trend. For instance, Seguam Island (52°19′24″N, 172°27′58″W) is located between two of the largest basins, Amlia and Amukta basins, created by rotating blocks (Geist et al., 1988). Also Ruppert et al. (2012), on the basis of focal mechanism solutions, showed different tectonics along the arc: alongarc strike-slip faulting dominates in the western part, strike-slip faulting perpendicular to the arc dominates in the central arc, and normal faulting is the most frequent in the eastern part. This complex structural architecture of the crust is reflected in the presence of dominant paths with different orientation, namely the N105° trend in the centralwestern part of the island and the N121° trend in the eastern part, plus a very secondary magma path trending NE-SW. On the other side, the production of magma along the western part of the volcanic arc is hindered, with respect to the eastern part, by a stronger coupling between the subducting plate and the overriding plate; this contributes to explaining the more frequent presence of huge volcano rows in the eastern part. Stronger coupling is shown by the presence of more frequent tectonic earthquakes with M N 6 occurring in the western arc compared to the central-eastern arc (Ekström

The volcanoes for which both NW- and NE-striking magma paths have been identified, might be the expression of transient magma stresses dominating over tectonic stresses. In this case, magma intrudes along different directions facilitated by preexisting mechanical discontinuities in the surrounding rocks. This might preferentially occur in the central conduit zone where the magma force is larger. This interpretation is consistent, for example, with the results coming from the inversion of fault plane solutions of 420 earthquakes at Redoubt volcano (60°29′07″N, 152°44′35″W) by Sánchez et al. (2004). These authors showed that during the 1989–90 eruptions, stress distribution was coherent with the expansion of a plexus-like magma body with different orientations. Another possibility comes from Mt. Spurr where Koulakov et al. (2013) have proposed a representation of the magma conduits following the 2005–2005 unrest that is different from what previously found at the same volcano after the 1992 eruption by Power et al. (1998), both using tomography. This suggests that seismic structures may change in time in consequence of reallocation of volatiles, partial melt and stress. A more detailed picture is here suggested based on the comparison between dyke injection geometries at Redoubt volcano during the 1989–90, 1991 and 2009 eruptions. Lahr et al. (1994) showed that the location of an earthquake swarm of the 1989–90 eruption is consistent with a NE-dipping narrow conduit (i.e. a NW-SE-striking dyke), whereas Benz et al. (1996) showed that the successive July 1991 eruption was accompanied by earthquakes compatible with a N15°-striking conduit zone. The successive eruption of 2009 was immediately followed by a period with volcano-tectonic events

Fig. 11. 1964 Earthquake-induced normal stress change (Fig. 10) resolved on NE-SW and NW-SE-striking magma pathways plotted versus epicentral distance. Unclamping for NESW-striking dykes (parallel to the trench and perpendicular to the overall σHmax) is always greater than for NW-SE-striking dykes (perpendicular to the trench and parallel to the overall σHmax).

20

A. Tibaldi, F.L. Bonali / Earth-Science Reviews 167 (2017) 1–26

Fig. 12. Simplified model of the stress regime acting in the hanging wall block pre- and post-large subduction earthquake (after Lupi and Miller, 2014). During the compression state, before the event (A), thrust faulting is encouraged and vertical fluid flow is limited. Particularly, beneath the volcanic arc, the value of σv (σ3) is greater (and close to σΗmin (σ2)) than at the same depths in the fore-arc region. (B) Due to post-seismic unloading, σv becomes the σ2 initiating strike-slip faulting in the arc, together with a reduction of the σHmax (σ1). Furthermore, at the same time, the vertical permeability increases due to the reduction in fault normal stress and the σ3 corresponds to σΗmin.

indicating a stable NE-SW-trending P-axis orientation corresponding to the horizontal displacement of the conduit walls (Roman and Gardine, 2013) along a possible NW-SE dyke. This alternation of intrusions along the two main NW-SE and NE-SW trends suggests that each intrusion might locally modify the stress state within the surrounding rocks, favouring the successive emplacement of a dyke in a perpendicular direction (Fig. 15C). This suggestion is consistent with the model of Roman (2005) and Roman and Cashman (2006) who propose that VT compressional events occur in the wall block of a dyke in consequence of its expansion in the direction of the regional σHmin. Dyke inflation induces a ~ 90° reorientation of the local stress field near the walls of the dyke with respect to the regional stress field orientation. We can arrive at a similar conclusion also by looking at the seismicity of Mt. Martin (58.172°N, 155.361°W) during the 2006 event. The absence of preferred P- and T-axis orientations during the pre-swarm period indicates that the local stress field beneath Mt. Martin was nearisotropic with about identical magnitudes of the three principal stress axes (O'Brien et al., 2012). The P-axis orientations became aligned in a NNE-SSW orientation, which is perpendicular to the regional σ1, during the shallow injection of a NW-SE-striking dyke. Finally, P-axis orientation became vertical during the post-swarm period. We suggest that the perturbation of the local stress field induced during dyke inflation may persist if faulting in the dyke wall block does not relieve the accumulated compression. If another dyke event occurs before the stress is relieved, the new dyke may be emplaced following the local σHmax, which in turn is perpendicular to the regional σHmax. This process results in the presence of two sets of magma paths, trending perpendicular to each other.

Fig. 13. Strike of the maximum horizontal stress in the study area (black arrows = arc zone; blue arrows = back-arc zone), data from the World Stress Map Project (Heidbach et al., 2010), based on focal mechanism solutions, boreholes and fault geometry. Rose diagrams show plots of maximum horizontal stress azimuths for the whole data set, along the volcanic arc, and in the back-arc area.

A. Tibaldi, F.L. Bonali / Earth-Science Reviews 167 (2017) 1–26

5.5. Effect of major subduction fault slip on magma pathways The major subduction earthquakes that take place along the Aleutian-Alaska arc can be accompanied by regional, vertical, differential positive and negative deformations, just like it happened during the 1964 earthquake (Stauder and Bollinger, 1966). The vertical motions can be accompanied by normal faulting along planes parallel to the elongation of the uplifting or subsiding blocks, as supported by the presence of Holocene normal faults very close to some of the studied volcanoes (e.g. Akutan, Pavlof, Kiska, Redoubt). The large subduction earthquakes produce static stress and strain changes in the uppermost part of the crust that comprise dilatation of the rock volume below the volcanic arc and consequent unclamping on suitably-oriented feeder dykes (Fig. 9). Two main theories have been proposed to explain seismic-induced changes in a magma body: i) magma is squeezed upward by increased compressional stress in the crust surrounding a magma chamber close to its critical state (e.g. Bautista et al., 1996), but we are inclined to rule out this process in the studied area because the magnitudes of the calculated earthquake-induced stress changes are quite low and in particular the very weak values of clamping exclude a “toothpaste” or “flask” process due to permanent deformation. ii) A decrease in compressional stress can promote additional melting, formation of bubbles and volatiles exsolution, encouraging eruptions in a time frame of years (Hill et al., 2002; Walter and Amelung, 2007). In fact, earthquake-induced static (permanent) normal stress reduction on magma pathway could directly promote dyke intrusion (Hill et al., 2002; Walter, 2007). Some authors suggested that for basaltic magmas, an overpressure b1 MPa is sufficient to generate tensile deviatoric stresses capable of promoting dyke formation and propagation without cooling, whereas for more silicic magmas this overpressure has been estimated at 10–100 MPa (Tait et al., 1989; McLeod and Tait, 1999; Jellinek and DePaolo, 2003). In our example, the magnitude of the calculated unclamping is always b 1 MPa, consistent with Manga and Brodsky (2006) who suggested values in the order of 10−2–10−1 MPa for seismic-induced dynamic and static stress changes. The same authors suggested that the overpressure of volcanic system must be within 99%–99.9% of the maximum overpressure for the earthquake to initiate a new eruption. On the contrary, volcanoes showing a longer response may be farther from this tipping point, but stress changes must still be significant to induce permanent pressure changes that could initiate a new eruption in a time frame of years, after an incubation period, where the eruption is the “failure event” (e.g. Jupp et al., 2004). We cannot exclude that, in some cases, both static and dynamic triggering might play a role in favouring new eruptions. Dynamic shaking, in fact, is another important factor in triggering eruptions over greater distances than the static effect (Manga and Brodsky, 2006). Dynamic shaking may promote ascent of gas bubbles, and consequently magma ascent (Manga and Brodsky, 2006); however, the details of feedback mechanisms are still unclear. Dynamic effects can also favour bubble growth, including adjective overpressure (Linde et al., 1994), rectified diffusion (Brodsky et al., 1998; Ichihara and Brodsky, 2006) and shear strain (Sumita and Manga, 2008). The example of unclamping produced by the 1964 Mw 9.2 Good Friday earthquake (Figs. 10–11) that was larger at NE-SW-striking planes (parallel to the trench) at different distances from the epicenter, indicates that a large subduction earthquake can facilitate magma upwelling along paths perpendicular to the overall regional maximum horizontal stress (Fig. 15D).

21

(63.45°N, 162.12°W) and St. Paul (57° 11′ 0″ N, 170° 16′ 0″ W). The distance, calculated as the shorter line connecting the center of the volcanic field with the nearest point of the subduction trench, is from 590 to 1230 km. If, however, we consider the distance as the line connecting the center of the volcanic field with the trench along the subduction direction, the values are 940–1660 km. It is worth noting that four out of six volcanoes are located at a distance of about 960 km from the trench. All these centers, in fact, are represented by flat shield volcanoes, with tens to more than one hundred pyroclastic cones and some maars (e.g. Fig. 7). They are made of very mafic products that indicate the absence of magma differentiation. All these centers are located in the Bering Strait region that does not correspond to a typical back-arc basin because it never experienced widespread spreading in recent geologic times (Nakamura and Uyeda, 1980). Anyway, the region has some Quaternary-Holocene structures as the Bering Strait Faults that are marked by E-W strike and normal dip-slip motions, and other parallel normal faults that affect the Chirikov Basin (Grim and McManus, 1970; Nelson et al., 1974). Also seismicity, structural and topographic data show Pliocene to Recent N-S extension across the region (Dumitru et al., 1995). A particularly large (Ms = 6.8) shallow

5.6. Magma paths at back-arc region Some of the Holocene volcanoes are located at distance from the arc; these are Imuruk (65° 35′ 59.99″N, 163° 55′ 11.99W), Ingakslugwat (61° 25′ 47.9994″N, 164° 28′ 11.99″W), Kookooligit (63°35′38″N, 170°22′56″W), Nunivak (60° 1′ 12″N, 166° 19′ 48″W), St. Michael

Fig. 14. Directions of σHmax divided by kinematics of earthquake focal mechanism solutions (435): (A) normal (64 solutions), (B) strike-slip (67 solutions), and (B) reverse (371 solutions).

22

A. Tibaldi, F.L. Bonali / Earth-Science Reviews 167 (2017) 1–26

earthquake did occur in 1991 in the eastern Bering Sea (Abers et al., 1993). The earthquake, with a N-S T-axis, was located along a NW-SE normal fault that follows the outer shelf of the Bering Sea. More recently, a continuous model strain rate field by Finzel et al. (2011) shows a NE-SW direction of extension south of the Bering Strait, where the six volcanic fields are located. The data mentioned so far indicate that the Holocene stress field in the back-arc region of the Bering Strait is expressed by an E-W σ2 and a N-S σ3 north of the strait, and a NW-SE σ2 and a NE-SW σ3 south of the strait, always with a vertical σ1. This implies a rotation of the stress tensor from the Aleutian-Alaska subduction zone where σ1 is horizontal. The extension in the Bering Strait region is thus in contrast to a well-documented N-S contraction in inner Alaska. This has been explained by a model in which westward extrusion of crustal blocks from interior Alaska by transcurrent faulting was accommodated by the N-S extension to the west (Dumitru et al., 1995). This extension may explain the presence of mafic magmas and their undifferentiated upwelling along crustal discontinuities. A generalized crustal extension and thinning is compatible with the presence of several conduits that supply magmas to hundreds of monogenic cones, as seen at the six volcanic fields. Nevertheless, it is still unexplained why crustal discontinuities used as magma paths at all these back-arc volcanoes range between N73° and N130°, i.e. oblique to the regional σ3. 6. Conclusions A review of all the available published geological, volcanological, structural and geophysical data, integrated with new observations, has allowed us to reconstruct the possible shallow magma paths for about half of the Holocene volcanoes of the Aleutian-Alaska arc and back-arc regions. The magma-feeding planes strike NNW-SSE to NW-SE at 40 volcanoes; they strike NE-SW at 6 volcanoes, and N-S at 3 volcanoes. Seven volcanoes have both the NW-SE and NE-SW trend.

Focal-mechanism solutions of tectonic earthquakes indicate that the horizontal greatest principal stress σHmax trends between NNW-SSE and NW-SE all over the arc, and subduction direction is NNW-SSE in the eastern arc and NW-SE in the western arc. Most volcanoes thus have magma paths consistent with the regional state of stress, i.e. dykes are emplaced perpendicular to the σHmin and to the volcanic arc. Nakamura's (1977) model would imply that σHmin = σ3 under a strike-slip fault regime with transcurrent faults oblique to the volcanic arc. Intra-arc, transcurrent faults are parallel to the arc and a horizontal σ3 is more linked to local block reorganization. The presence of intra-arc reverse faulting implies that σHmin may also be =σ2; in this case, possible sill emplacement may locally increase the vertical stress leading to a switch between σ2 and σ3 with consequent σHmin = σ3 again, and emplacement of dykes parallel to σHmax. Three possible explanations for the presence of magma paths perpendicular to the direction of subduction are proposed, taking into consideration that these mechanisms can be active at the same time: i) one involves the effect of magma flux and related large pressure through a crust already affected by arc-parallel fractures. An eastward increase in the number of stratovolcanoes and the presence of rows of coalescent cones, together with a decrease in frequency of large magnitude tectonic events and an increase in number and depth of VT events in the same direction, may be indicative of a higher magma flux above the eastern subduction zone. The uprising of large magma batches produces a large enough stress to open preexisting arc-parallel fractures. ii) Another explanation suggests that once that magma has intruded along a dyke parallel to the regional σHmax, inflation produces an increase of the stress perpendicular to the dyke wall, resulting in a reorientation of the local σHmax. The successive dyke emplacement will take place following the new direction of the local stress field.

Fig. 15. (A) Sketch illustrating the relationships between the orientation of the principal stress axes in the forearc, arc, and back-arc zones, and the trend of the NW-SE dykes (normal to trench) in the Aleutian-Alaska arc and NE-SW dykes in the back-arc, following Nakamura et al.'s (1980) model. (B) Under a reverse faulting tectonic setting, a sill may be emplaced and then induce an increase in the vertical stress resulting in a switch between σ2 and σ3 and emplacement of a trench-normal dyke. (C) NE-SW dyke (trench-parallel) in the arc zone induced by local push of a NW-SE dyke. (D) NE-SW dykes (trench-parallel) in the arc zone favoured by earthquake-induced unclamping. Black arrows simply indicate the σ3.

A. Tibaldi, F.L. Bonali / Earth-Science Reviews 167 (2017) 1–26

iii) A third explanation involves the unclamping effect of large subduction earthquakes that decrease the stress acting on fracture planes parallel to the volcanic arc, facilitating dyke emplacement along this direction.

The activation of one preferential trend mainly depends upon the balance between magmatic stress, tectonic stress and static stress change, in connection with the presence of preferential vertical discontinuities in the rocks; the substrate surrounding the studied volcanoes, in fact, shows preferential NW- and/or NE-striking fracture fields. Acknowledgments We wish to acknowledge the Editor, Arturo Gómez-Tuena for his useful comments and Michelle Coombs for her very insightful review, which helped us improve the quality of our paper. This work was carried out in the framework of the International Lithosphere ProgramTask Force II “Volcanoes and society: environment, health and hazards”. The Coulomb 3.2 software is provided freeware for research and teaching purposes from www.coulombstress.org. References Abers, G.A., Ekström, G., Marlow, M.S., Geist, E.L., 1993. Bering Sea earthquake of February 21, 1991: active faulting along the Bering shelf edge. J. Geophys. Res. Solid Earth 98 (B2), 2155–2165. Acocella, V., Neri, M., 2009. Dike propagation in volcanic edifices: overview and possible developments. Tectonophysics 471 (1), 67–77. Acocella, V., Tibaldi, A., 2005. Dike propagation driven by volcano collapse: a general model tested at Stromboli, Italy. Geophys. Res. Lett. 32 (8). Acocella, V., Yoshida, T., Yamada, R., Funiciello, F., 2008. Structural control on late Miocene to Quaternary volcanism in the NE Honshu arc, Japan. Tectonics 27 (5). Alaska Seismic Hazards Safety Commission, 2012. Annual Report to the Governor and State Legislature. http://seismic.alaska.gov/. Aloisi, M., Bonaccorso, A., Gambino, S., 2006. Imaging composite dike propagation (Etna, 2002 case). J. Geophys. Res. Solid Earth 111 (B6). Annen, C., Lénat, J.F., Provost, A., 2001. The long-term growth of volcanic edifices: numerical modelling of the role of dyke intrusion and lava-flow emplacement. J. Volcanol. Geotherm. Res. 105 (4), 263–289. Bacon, C.R., Calvert, A.T., Nye, C.J., Sisson, T.W., 2003. History and Eruptive Style of Mount Veniaminof, a Huge Alaskan Basalt-to-Dacite Volcano With Pleistocene and Holocene Caldera-forming Eruptions. 1. AGU, p. 1048 Fall Meeting Abstracts. Bacon, C.R., Neal, C.A., Miller, T.P., McGimsey, R.G., Nye, C.J., 2014. Postglacial eruptive history, geochemistry, and recent seismicity of Aniakchak Volcano, Alaska. U.S. Geological Survey Professional Paper 1810:74 p. Available online at http://pubs.usgs.gov/pp/ 1810/. Barth, T.F., 1956. Geology and petrology of the Pribilof Islands, Alaska. U.S. Geol. Surv. Bull. 101–160. Bautista, B.C., Bautista, M.L.P., Stein, R.S., Barcelona, E.S., Punongbayan, R.S., Laguerta, E.P., Rasdas, A.R., Ambubuyog, G., Amin, E.Q., 1996. Relationship of regional and local structures to Mount Pinatubo activity. Fire and Mud: Eruptions and Lahars of Mount Pinatubo, Philippines, pp. 351–370. Bebbington, M.S., Marzocchi, W., 2011. Stochastic models for earthquake triggering of volcanic eruptions. J. Geophys. Res. Solid Earth 116 (B5). Beck, S.L., Christensen, D.H., 1991. Rupture process of the February 4, 1965, Rat Islands earthquake. J. Geophys. Res. 96 (B2), 2205–2221. Benz, H.M., Chouet, B.A., Dawson, P.B., Lahr, J.C., Page, R.A., Hole, J.A., 1996. Threedimensional P and S wave velocity structure of Redoubt Volcano, Alaska. J. Geophys. Res. Solid Earth 101 (B4), 8111–8128. Bindeman, I.N., Fournelle, J.H., Valley, J.W., 2001. Low-δ18O tephra from a compositionally zoned magma body: Fisher Caldera, Unimak Island, Aleutians. J. Volcanol. Geotherm. Res. 111 (1), 35–53. Bistacchi, A., Tibaldi, A., Pasquarè, F.A., Rust, D., 2012. The association of cone–sheets and radial dykes: data from the Isle of Skye (UK), numerical modelling, and implications for shallow magma chambers. Earth Planet. Sci. Lett. 339, 46–56. Bonali, F.L., 2013. Earthquake-induced static stress change on magma pathway in promoting the 2012 Copahue eruption. Tectonophysics 608, 127–137. Bonali, F.L., Corazzato, C., Tibaldi, A., 2011. Identifying rift zones on volcanoes: an example from La Réunion Island, Indian Ocean. Bull. Volcanol. 73 (3), 347–366. Bonali, F.L., Tibaldi, A., Corazzato, C., 2015. Sensitivity analysis of earthquake-induced static stress changes on volcanoes: the 2010 Mw 8.8 Chile earthquake. Geophys. J. Int. 201 (3), 1868–1890. Bonali, F.L., Corazzato, C., Bellotti, F., Groppelli, G., 2016. Active Tectonics and Its Interactions with Copahue Volcano. In Copahue VolcanoSpringer, Berlin Heidelberg, pp. 23–45. Bonali, F.L., Tibaldi, A., Corazzato, C., Tormey, D.R., Lara, L.E., 2013. Quantifying the effect of large earthquakes in promoting eruptions due to stress changes on magma pathway: the Chile case. Tectonophysics 583, 54–67.

23

Bridges, D.L., Gao, S.S., 2006. Spatial variation of seismic b-values beneath Makushin Volcano, Unalaska Island, Alaska. Earth Planet. Sci. Lett. 245 (1), 408–415. Brodsky, E.E., Sturtevant, B., Kanamori, H., 1998. Earthquakes, volcanoes, and rectified diffusion. J. Geophys. Res. B 103 (B10), 23827–23838. Buurman, H., West, M.E., Thompson, G., 2013. The seismicity of the 2009 Redoubt eruption. J. Volcanol. Geotherm. Res. 259, 16–30. Buurman, H., Nye, C.J., West, M.E., Cameron, C., 2014. Regional controls on volcano seismicity along the Aleutian arc. Geochem. Geophys. Geosyst. 15 (4), 1147–1163. Cameron, C.E., Nye, C.J., 2014. Preliminary database of Quaternary vents in Alaska: Alaska Division of Geological & Geophysical Surveys Miscellaneous Publication. 153:p. 11. http://dx.doi.org/10.14509/27357. Carracedo, J.C., 1994. The Canary Islands: an example of structural control on the growth of large oceanic-island volcanoes. J. Volcanol. Geotherm. Res. 60 (3–4), 225–241. Chaput, M., Famin, V., Michon, L., 2014. Deformation of basaltic shield volcanoes under cointrusive stress permutations. J. Geophys. Res. Solid Earth 119 (1), 274–301. Chaussard, E., Amelung, F., Aoki, Y., 2013. Characterization of open and closed volcanic systems in Indonesia and Mexico using InSAR time series. J. Geophys. Res. Solid Earth 118 (8), 3957–3969. Chouet, B., Saccorotti, G., Martini, M., Dawson, P., De Luca, G., Milana, G., Scarpa, R., 1997. Source and path effects in the wave fields of tremor and explosions at Stromboli Volcano, Italy. J. Geophys. Res. Solid Earth 102, 15–129. Chouet, B., Dawson, P., Martini, M., 2008. Shallow-conduit dynamics at Stromboli Volcano, Italy, imaged from waveform inversions. Geol. Soc. Lond., Spec. Publ. 307 (1), 57–84. Christensen, D.H., Beck, S.L., 1994. The rupture process and tectonic implications of the great 1964 Prince William Sound earthquake. Shallow Subduction Zones: Seismicity, Mechanics and Seismic Potential. Birkhäuser, Basel, pp. 29–53. Coats, R.R., 1956a. Reconnaissance Geology of Some Western Aleutian Islands, Alaska. US Government Printing Office. Coats, R.R., 1956b. Geology of northern Kanaga Island, Alaska. U.S. Geol. Surv. Bull. 1028D, 69–81. Coats, R.R., 1959. Geologic Reconnaissance of Semisopochnoi Island, Western Aleutian Islands, Alaska (No. 1028). US Government Printing Office. Coats, R.R., Nelson, W.H., Lewis, R.Q., Powers, H.A., 1961. Geologic reconnaissance of Kiska Island, Aleutian Islands, Alaska. U.S. Geol. Surv. Bull. 1028-R, 563–581. Connor, C.B., 1990. Cinder cone clustering in the TransMexican Volcanic Belt: implications for structural and petrologic models. J. Geophys. Res. Solid Earth 95 (B12), 19395–19405. Coombs, M.L., White, S.M., Scholl, D.W., 2007. Massive edifice failure at Aleutian arc volcanoes. Earth Planet. Sci. Lett. 256 (3), 403–418. Coombs, M.L., McGimsey, R.G., Browne, B.L., 2008. Preliminary Volcano-Hazard Assessment for Gareloi Volcano, Gareloi Island, Alaska. US Department of the Interior, US Geological Survey. Coonrad, W.L., 1957. Geologic Reconnaissance in the Yukon-Kuskokwim Delta Region, Alaska (No. 223). Corazzato, C., Tibaldi, A., 2006. Fracture control on type, morphology and distribution of parasitic volcanic cones: an example from Mt. Etna, Italy. J. Volcanol. Geotherm. Res. 158 (1), 177–194. Corazzato, C., Francalanci, L., Menna, M., Petrone, C., Renzulli, A., Tibaldi, A., Vezzoli, L., 2008. What does it guide sheet intrusion in volcanoes? Petrological and structural characters of the Stromboli sheet complex, Italy. J. Volcanol. Geotherm. Res. 173, 26–54. Davies, J.N., Taber, John, Miller, T.P., Power, J.A., Yount, M.E., 1988. Seismicity and Geology of Mt. Dutton, a Volcano Near the Tip of the Alaska Peninsula. American Geophysical Union, p. 1507 Abstracts with Programs, v. 69, n. 44. Debari, S., Kay, S.M., Kay, R.W., 1987. Ultramafic xenoliths from Adagdak volcano, Adak, Aleutian Islands, Alaska: deformed igneous cumulates from the Moho of an island arc. J. Geol. 329–341. Delle Donne, D., Harris, A.J., Ripepe, M., Wright, R., 2010. Earthquake-induced thermal anomalies at active volcanoes. Geology 38 (9), 771–774. DeMets, C., Gordon, R.G., Argus, D.F., Stein, S., 1994. Effect of recent revisions to the geomagnetic reversal time scale on estimates of current plate motions. Geophys. Res. Lett. 21 (20), 2191–2194. Dieterich, J.H., 1988. Growth and persistence of Hawaiian volcanic rift zones. J. Geophys. Res. Solid Earth 93 (B5), 4258–4270. Doukas, M.P., McGee, K.A., 2007. A Compilation of Gas Emission-Rate Data From Volcanoes of Cook Inlet (Spurr, Crater Peak, Redoubt, Iliamna, and Augustine) and Alaska Peninsula (Douglas, Fourpeaked, Griggs, Mageik, Martin, Peulik, Ukinrek Maars, and Veniaminof), Alaska, from 1995–2006. U.S. Geological Survey (Open-File Report 2007-1400). Dumitru, T.A., Miller, E.L., O'Sullivan, P.B., Amato, J.M., Hannula, K.A., Calvert, A.T., Gans, P.B., 1995. Cretaceous to recent extension in the Bering Strait region, Alaska. Tectonics 14 (3), 549–563. Eggert, S., Walter, T.R., 2009. Volcanic activity before and after large tectonic earthquakes: observations and statistical significance. Tectonophysics 471 (1), 14–26. Ekström, G., Engdahl, E.R., 1989. Earthquake source parameters and stress distribution in the Adak Island region of the central Aleutian Islands, Alaska. J. Geophys. Res. Solid Earth 94 (B11), 15499–15519. Feeley, T.C., Winer, G.S., 1999. Evidence for fractionation of Quaternary basalts on St. Paul Island, Alaska, with implications for the development of shallow magma chambers beneath Bering Sea volcanoes. Lithos 46 (4), 661–676. Fierstein, J., Hildreth, W., 2008. Kaguyak dome field and its Holocene caldera, Alaska Peninsula. J. Volcanol. Geotherm. Res. 177 (2), 340–366. Finzel, E.S., Flesch, L.M., Ridgway, K.D., 2011. Kinematics of a diffuse North America–Pacific–Bering plate boundary in Alaska and western Canada. Geology 39 (9), 835–838. Finzel, E.S., Flesch, L.M., Ridgway, K.D., 2014. Present-day geodynamics of the northern North American Cordillera. Earth Planet. Sci. Lett. 404, 111–123.

24

A. Tibaldi, F.L. Bonali / Earth-Science Reviews 167 (2017) 1–26

Fiske, R.S., Jackson, E.D., 1972, August. Orientation and growth of Hawaiian volcanic rifts: the effect of regional structure and gravitational stresses. Proceedings of the Royal Society of London A: Mathematical, Physical and Engineering Sciences. Vol. 329. The Royal Society, pp. 299–326 No. 1578. Fournier, T., Freymueller, J., 2008. Inflation detected at Mount Veniaminof, Alaska, with campaign GPS. Geophys. Res. Lett. 35 (20). Fournier, T., Freymueller, J., Cervelli, P., 2009. Tracking magma volume recovery at Okmok volcano using GPS and an unscented Kalman filter. J. Geophys. Res. Solid Earth 114 (B2). Freed, A.M., Lin, J., 2002. Accelerated stress buildup on the southern San Andreas Fault and surrounding regions caused by Mojave Desert earthquakes. Geology 30 (6), 571–574. Froese, D., Westgate, J., Preece, S., Storer, J., 2002. Age and significance of the late Pleistocene Dawson tephra in eastern Beringia. Quat. Sci. Rev. 21 (20), 2137–2142. Gardine, M., West, M., Werner, C., Doukas, M., 2011. Evidence of magma intrusion at Fourpeaked volcano, Alaska in 2006–2007 from a rapid-response seismic network and volcanic gases. J. Volcanol. Geotherm. Res. 200 (3), 192–200. Gautneb, H., Gudmundsson, A., 1992. Effect of local and regional stress fields on sheet emplacement in West Iceland. J. Volcanol. Geotherm. Res. 51 (4), 339–356. Geist, E.L., Childs, J., Scholl, D., 1988. The origin of summit basins of the Aleutian ridge: implications for block rotation of an arc massif. Tectonics 7, 327–341. Geoffroy, L., Bergerat, F., Angelier, J., 1993. Modification d'un champ de contrainte regional par un champ de contraintes magmatiques local; Exemple de l'ile de Skye (Ecosse) au Paleocene. Bull. Soc. Geol. Fr. 164 (4), 541–552. Germa, A., Connor, L.J., Cañon-Tapia, E., Le Corvec, N., 2013. Tectonic and magmatic controls on the location of post-subduction monogenetic volcanoes in Baja California, Mexico, revealed through spatial analysis of eruptive vents. Bull. Volcanol. 75 (12), 1–14. Global Volcanism Program, 2013. In: Venzke, E. (Ed.), Volcanoes of the World, v. 4.5.0. Smithsonian Institution (Downloaded 30 Aug 2016). Grim, M.S., McManus, D.A., 1970. A shallow seismic-profiling survey of the northern Bering Sea. Mar. Geol. 8 (5), 293309–306320. Gudmundsson, A., 1984. Formation of dykes, feeder-dykes, and the intrusion of dykes from magma chambers. Bull. Volcanol. 47 (3), 537–550. Gudmundsson, A., 1987. Formation and mechanics of magma reservoirs in Iceland. Geophys. J. Int. 91 (1), 27–41. Gudmundsson, A., 1988. Effect of tensile stress concentration around magma chambers on intrusion and extrusion frequencies. J. Volcanol. Geotherm. Res. 35 (3), 179–194. Gudmundsson, A., 1990. Emplacement of dikes, sills and crustal magma chambers at divergent plate boundaries. Tectonophysics 176 (3), 257–275. Gudmundsson, A., 1995. Infrastructure and mechanics of volcanic systems in Iceland. J. Volcanol. Geotherm. Res. 64 (1), 1–22. Gudmundsson, A., 2002. Emplacement and arrest of sheets and dykes in central volcanoes. J. Volcanol. Geotherm. Res. 116 (3), 279–298. Gudmundsson, A., 2006. How local stresses control magma-chamber ruptures, dyke injections, and eruptions in composite volcanoes. Earth Sci. Rev. 79 (1), 1–31. Gudmundsson, A., 2012. Magma chambers: formation, local stresses, excess pressures, and compartments. J. Volcanol. Geotherm. Res. 237–238, 19–41. Gudmundsson, A., Loetveit, I.F., 2005. Dyke emplacement in a layered and faulted rift zone. J. Volcanol. Geotherm. Res. 144 (1), 311–327. Haeussler, P.J., Schwartz, D.P., Dawson, T.E., Stenner, H.D., Lienkaemper, J.J., Sherrod, B., Cinti, F.R., Montone, P., Craw, P.A., Crone, A.J., Personius, S.F., 2004. Surface rupture and slip distribution of the Denali and Totschunda faults in the 3 November 2002 M 7.9 earthquake, Alaska. Bull. Seismol. Soc. Am. 94 (6B), S23–S52. Haney, M.M., 2010. Location and mechanism of very long period tremor during the 2008 eruption of Okmok volcano from interstation arrival times. J. Geophys. Res. Solid Earth 115 (B10). Harding, S.T., Algermissen, S.T., 1969. Focal mechanism of the Prince William Sound, Alaska earthquake of March 28, 1964. Bull. Seismol. Soc. Am. 59 (2), 799–811. Hastie, L.M., Savage, J.C., 1970. A dislocation model for the 1964 Alaska earthquake. Bull. Seismol. Soc. Am. 60 (4), 1389–1392. Heidbach, O., Tingay, M., Barth, A., Reinecker, J., Kurfeß, D., Müller, B., 2010. Global crustal stress pattern based on the World Stress Map database release 2008. Tectonophysics 482 (1), 3–15. Hildreth, W., Fierstein, J., 2000. Katmai volcanic cluster and the great eruption of 1912. Geol. Soc. Am. Bull. 112 (10), 1594–1620. Hildreth, W., Fierstein, J., Lanphere, M.A., Siems, D.F., 2001. Trident Volcano: Four Contiguous Stratocones Adjacent to Katmai Pass, Alaska Peninsula. Studies by the US Geological Survey in Alaska. pp. 153–180. Hildreth, W., Fierstein, J., Calvert, A.T., 2007. Blue Mountain and the Gas Rocks: Rear-Arc Dome Clusters on the Alaska Peninsula. US Department of the Interior, US Geological Survey, pp. 1–27. Hill, D.P., Pollitz, F., Newhall, C., 2002. Earthquake-volcano interactions. Phys. Today 55 (11), 41–47. Hoare, J.M., Condon, W.H., 1971. Geologic Map of the Marshall Quadrangle, Western Alaska (No. 668). Hoare, J.M., Condon, W.H., Cox, A., Dalrymple, G.B., 1968. Geology, paleomagnetism, and potassium-argon ages of basalts from Nunivak Island, Alaska. Geol. Soc. Am. Mem. 116, 377–414. Hopkins, D.M., 1959. History of Imuruk Lake, Seward Peninsula, Alaska. Geol. Soc. Am. Bull. 70 (8), 1033–1046. Hopkins, D.M., 1963. Geology of the Imuruk Lake Area, Seward Peninsula, Alaska (No. 1141-C). US Govt. Print. Off. Ichihara, M., Brodsky, E.E., 2006. A limit on the effect of rectified diffusion in volcanic systems. Geophys. Res. Lett. 33 (2). Ichinose, G., Somerville, P., Thio, H.K., Graves, R., O'Connell, D., 2007. Rupture process of the 1964 Prince William Sound, Alaska, earthquake from the combined inversion of seismic, tsunami, and geodetic data. J. Geophys. Res. Solid Earth 112 (B7).

Institution, Smithsonian, 1988. Scientific Event Alert Network Bulletin, v. 13, Nos. 2. pp. 4–8. Jaggar, T.A., 1908. The evolution of Bogoslof Volcano. Bull. Am. Geogr. Soc. 40 (7), 385–400. Jellinek, A.M., DePaolo, D.J., 2003. A model for the origin of large silicic magma chambers: precursors of caldera-forming eruptions. Bull. Volcanol. 65 (5), 363–381. Jicha, B.R., Singer, B.S., 2006. Volcanic history and magmatic evolution of Seguam Island, Aleutian Island arc, Alaska. Geol. Soc. Am. Bull. 118 (7–8), 805–822. Jicha, B.R., Coombs, M.L., Calvert, A.T., Singer, B.S., 2012. Geology and 40Ar/39Ar geochronology of the medium-to high-K Tanaga volcanic cluster, western Aleutians. Geol. Soc. Am. Bull. 124 (5–6), 842–856. Johnson, C.A., Harrison, C.G.A., 1990. Neotectonics in central Mexico. Phys. Earth Planet. Inter. 64 (2–4), 187–210. Johnson, J.M., Satake, K., 1994. Rupture extent of the 1938 Alaskan earthquake as inferred from tsunami waveforms. Geophys. Res. Lett. 21 (8), 733–736. Johnson, J.M., Satake, K., 1997. Estimation of seismic moment and slip distribution of the April 1, 1946, Aleutian tsunami earthquake. J. Geophys. Res. 102 (11), 11765–11774. Johnson, J.M., Satake, K., Holdahl, S.R., Sauber, J., 1996. The 1964 Prince William Sound earthquake: joint inversion of tsunami and geodetic data. J. Geophys. Res. Solid Earth 101, 523–532. Jolly, A.D., McNutt, S.R., 1999. Seismicity at the volcanoes of Katmai National Park, Alaska; July 1995–December 1997. J. Volcanol. Geotherm. Res. 93 (3), 173–190. Jupp, T.E., Pyle, D.M., Mason, B.G., Dade, W.B., 2004. A statistical model for the timing of earthquakes and volcanic eruptions influenced by periodic processes. J. Geophys. Res. Solid Earth 109 (B2). King, G.C., Stein, R.S., Lin, J., 1994. Static stress changes and the triggering of earthquakes. Bull. Seismol. Soc. Am. 84 (3), 935–953. Koehler, R.D., Burns, R.E., Combellick, R.A., 2012. Quaternary Faults and Folds in Alaska: A Digital Database. Koulakov, I., West, M., Izbekov, P., 2013. Fluid ascent during the 2004–2005 unrest at Mt. Spurr inferred from seismic tomography. Geophys. Res. Lett. 40 (17), 4579–4582. Krutikov, L., Stone, D.B., Minyuk, P., 2008. New paleomagnetic data from the central Aleutian Arc: evidence and implications for block rotations. Active Tectonics and Seismic Potential of Alaska, pp. 135–149. Kühn, D., Dahm, T., 2008. Numerical modelling of dyke interaction and its influence on oceanic crust formation. Tectonophysics 447 (1), 53–65. Lahr, J.C., Chouet, B.A., Stephens, C.D., Power, J.A., Page, R.A., 1994. Earthquake classification, location, and error analysis in a volcanic environment: implications for the magmatic system of the 1989–1990 eruptions at Redoubt Volcano, Alaska. J. Volcanol. Geotherm. Res. 62 (1), 137–151. Lallemant, H.G.A., Oldow, J.S., 2000. Active displacement partitioning and arc-parallel extension of the Aleutian volcanic arc based on Global Positioning System geodesy and kinematic analysis. Geology 28 (8), 739–742. Lamb, D., Linneman, S.R., Myers, J.D., Nicolaysen, K.E., 1992. Caldera formation on Yunaska Island, central Aleutian arc. Eos 73 (43), 645. Lehto, H.L., Roman, D.C., Moran, S.C., 2010. Temporal changes in stress preceding the 2004–2008 eruption of Mount St. Helens, Washington. J. Volcanol. Geotherm. Res. 129–142. Linde, A.T., Sacks, I.S., 1998. Triggering of volcanic eruptions. Nature 395, 888–890. Lin, J., Stein, R.S., 2004. Stress triggering in thrust and subduction earthquakes and stress interaction between the southern San Andreas and nearby thrust and strike-slip faults. J. Geophys. Res. Solid Earth 109 (B2). Linde, A.T., Sacks, I.S., Johnston, M.J.S., Hill, D.P., Bilham, R.G., 1994. Increased pressure from rising bubbles as a mechanism for remotely triggered seismicity. Nature 371, 408–410. Lowenstern, J.B., Wallmann, P.C., Pollard, D.D., 1991. The West Mageik Lake sill complex as an analogue for magma transport during the 1912 eruption at the Valley of Ten Thousand Smokes, Alaska. Geophys. Res. Lett. 18 (8), 1569–1572. Lu, Zhong, Dzurisin, Daniel, 2014. InSAR Imaging of Aleutian Volcanoes: Chichester. Springer-Praxis, UK (390 pp). Lu, Z., Fatland, R., Wyss, M., Li, S., Eichelberger, J., Dean, K., Freymueller, J., 1997. Deformation of New Trident Volcano measured by ERS-1 SAR interferometry, Katmai National Park, Alaska. Geophys. Res. Lett. 24 (6), 695–698. Lu, Z., Wicks, C., Power, J.A., Dzurisin, D., 2000. Ground deformation associated with the March 1996 earthquake swarm at Akutan volcano, Alaska, revealed by satellite radar interferometry. J. Geophys. Res. Solid Earth 105 (B9), 21483–21495. Lu, Z., Power, J.A., McConnell, V.S., Wicks, C., Dzurisin, D., 2002. Preeruptive inflation and surface interferometric coherence characteristics revealed by satellite radar interferometry at Makushin Volcano, Alaska: 1993–2000. J. Geophys. Res. Solid Earth 107 (B11). Lu, Z., Masterlark, T., Dzurisin, D., Rykhus, R., Wicks, C., 2003. Magma supply dynamics at Westdahl volcano, Alaska, modeled from satellite radar interferometry. J. Geophys. Res. Solid Earth 108 (B7). Lu, Z., Rykhus, R., Masterlark, T., Dean, K.G., 2004. Mapping recent lava flows at Westdahl Volcano, Alaska, using radar and optical satellite imagery. Remote Sens. Environ. 91 (3), 345–353. Lupi, M., Miller, S.A., 2014. Short-lived tectonic switch mechanism for long-term pulses of volcanic activity after mega-thrust earthquakes. Solid Earth 5 (1), 13. Lutz, T.M., 1986. An analysis of the orientations of large-scale crustal structures: a statistical approach based on areal distributions of pointlike features. J. Geophys. Res. Solid Earth 91 (B1), 421–434. Lutz, T.M., Gutmann, J.T., 1995. An improved method for determining and characterizing alignments of pointlike features and its implications for the Pinacate volcanic field, Sonora, Mexico. J. Geophys. Res. Solid Earth 100 (B9), 17659–17670. Manga, M., Brodsky, E., 2006. Seismic triggering of eruptions in the far field: volcanoes and geysers. Annu. Rev. Earth Planet. Sci. 34, 263–291.

A. Tibaldi, F.L. Bonali / Earth-Science Reviews 167 (2017) 1–26 Mangan, M.T., Waythomas, C.F., Miller, T.P., Trusdell, F.A., 2003. Emmons lake volcanic center, Alaska Peninsula: source of the Late Wisconsin Dawson tephra, Yukon territory, Canada. Can. J. Earth Sci. 40 (7), 925–936. Mangan, M., Miller, T., Waythomas, C., Trusdell, F., Calvert, A., Layer, P., 2009. Diverse lavas from closely spaced volcanoes drawing from a common parent: Emmons Lake Volcanic Center, Eastern Aleutian Arc. Earth Planet. Sci. Lett. 287 (3), 363–372. Mann, D., Freymueller, J., 2003. Volcanic and tectonic deformation on Unimak Island in the Aleutian Arc, Alaska. J. Geophys. Res. Solid Earth 108 (B2). Mann, D., Freymueller, J., Lu, Z., 2002. Deformation associated with the 1997 eruption of Okmok volcano, Alaska. J. Geophys. Res. Solid Earth 107 (B4). Marsh, B.D., 1980. Geology of northern Atka, Aleutian Islands, Alaska. Geol. Soc. Am. Abstr. Programs 12, 476. Marzocchi, W., 2002. Remote seismic influence on the large explosive eruptions. J. Geophys. Res. 107 (EPM 6-1). Marzocchi, W., Casarotti, E., Piersanti, A., 2002. Modeling the stress variations induced by great earthquakes on the largest volcanic eruptions of the 20th century. J. Geophys. Res. Solid Earth 107 (B11). Massonnet, D., Sigmundsson, F., 2000. Remote sensing of volcano deformation by radar interferometry from various satellites. Remote Sensing of Active Volcanism, pp. 207–221. Masterlark, T., Haney, M., Dickinson, H., Fournier, T., Searcy, C., 2010. Rheologic and structural controls on the deformation of Okmok Volcano, Alaska: FEMs, InSAR, and ambient noise tomography. J. Geophys. Res. Solid Earth 115 (B2). Mattia, M., Patane, D., Aloisi, M., Amore, M., 2007. Faulting on the western flank of Mt Etna and magma intrusions in the shallow crust. Terra Nova 19 (1), 89–94. Matumoto, T., Ward, P.L., 1967. Microearthquake study of Mount Katmai and vicinity, Alaska. J. Geophys. Res. 72 (10), 2557–2568. McCaffrey, R., 1992. Oblique plate convergence, slip vectors, and forearc deformation. J. Geophys. Res. Solid Earth 97 (B6), 8905–8915. McCaffrey, R., 2002. Crustal block rotations and plate coupling. In: Stein, S., Freymueller, J. (Eds.), Plate Boundary Zones, AGU Geodyn. Ser. vol. 30. AGU, Washington, D. C, pp. 101–122. McConnell, V.S., Beget, J.E., Roach, A.L., Bean, K.W., Nye, C.J., 1998. Geologic map of the Makushin volcanic field, Unalaska Island, Alaska. Alaska Division of Geological & Geophysical Surveys Report of Investigations RI 97-20 unpaged, 2 sheets, scale 1:63,360. McLeod, P., Tait, S., 1999. The growth of dykes from magma chambers. J. Volcanol. Geotherm. Res. 92 (3), 231–245. McNutt, S.R., Beavan, R.J., 1981. Volcanic earthquakes at Pavlof Volcano correlated with the solid earth tide. Nature 294 (5842), 615–618. McNutt, S.R., Jacob, K.H., 1986. Determination of large-scale velocity structure of the crust and upper mantle in the vicinity of Pavlof Volcano, Alaska. J. Geophys. Res. Solid Earth 91 (B5), 5013–5022. Miller, T.P., 2004. Geology of the Ugashik-Mount Peulik Volcanic Center, Alaska. USGS (No. 2004-1009). Miller, T.P., Smith, R.L., 1977. Spectacular mobility of ash flows around Aniakchak and Fisher Calderas, Alaska. Geology 5 (3), 173–176. Miller, T.P., Smith, R.L., 1987. Late Quaternary caldera-forming eruptions in the eastern Aleutian Arc, Alaska. Geology 15 (5), 434–438. Miller, T.P., McGimsey, R.G., Richter, D.H., Riehle, J.R., Nye, C.J., Yount, M.E., Dumoulin, J.A., 1998. Catalog of the Historically Active Volcanoes of Alaska. US Geological Survey, pp. 98–582 Open-File Report. Miller, T.P., Chertkoff, D.G., Eichelberger, J.C., Coombs, M.L., 1999. Mount Dutton volcano, Alaska: Aleutian arc analog to Unzen Volcano, Japan. J. Volcanol. Geotherm. Res. 89 (1), 275–301. Mithen, D.P., 1982. Stress amplification in the upper crust and the development of normal faulting. Tectonophysics 83 (3), 259–273. Moore, J.G., Normark, W.R., Holcomb, R.T., 1994. Giant Hawaiian landslides. Annu. Rev. Earth Planet. Sci. 22, 119–144. Moran, S., Stihler, S., Power, J., 2002. A tectonic earthquake sequence preceding the April– May 1999 eruption of Shishaldin Volcano, Alaska. Bull. Volcanol. 64 (8), 520–524. Motyka, R.J., Liss, S.A., Nye, C.J., Moorman, M.A., 1993. Geothermal Resources of the Aleutian Arc. Alaska Div Geol Geophys Surv (Prof Rpt, no. 114, 17 p and 4 map sheets). Mukasa, S.B., Andronikov, A.V., Hall, C.M., 2007. The 40Ar/39Ar chronology and eruption rates of Cenozoic volcanism in the eastern Bering Sea Volcanic Province, Alaska. J. Geophys. Res. Solid Earth 112 (B6). Myers, J.D., Marsh, B.D., Frost, C.D., Linton, J.A., 2002. Petrologic constraints on the spatial distribution of crustal magma chambers, Atka volcanic center, central Aleutian Arc. Contrib. Mineral. Petrol. 143 (5), 567–586. Nakamura, K., 1977. Volcanoes as possible indicators of tectonic stress orientation—principle and proposal. J. Volcanol. Geotherm. Res. 2 (1), 1–16. Nakamura, K., Uyeda, S., 1980. Stress gradient in arc–back arc regions and plate subduction. J. Geophys. Res. Solid Earth 85 (B11), 6419–6428. Nakamura, K., Plafker, G., Jacob, K.H., Davies, J.N., 1980. A Tectonic Stress Trajectory Map of Alaska Using Information From Volcanoes and Faults. 55. Bull. Earth. Res. Center, pp. 89–100. Namiki, A., Rivalta, E., Woith, H., Walter, T.R., 2016. Sloshing of a bubbly magma reservoir as a mechanism of triggered eruptions. J. Volcanol. Geotherm. Res. 320, 156–171. Neal, C.A., McGimsey, R.G., Miller, T.P., Riehle, J.R., Waythomas, C.F., 2000. Preliminary Volcano-Hazard Assessment for Aniakchak Volcano, Alaska (No. 00-519). US Geological Survey. Neal, C., McGimsey, R., Dixon, J., Manevich, A., Rybin, A., 2009. 2006 volcanic activity in Alaska, Kamchatka, and the Kurile Islands: summary of events and response of the Alaska Volcano Observatory. U.S. Geological Survey Scientific Investigations Report 2008-5214 http://pubs.usgs.gov/sir/2008/5214/.

25

Nelson, W.H., 1959. Geology of Segula, Davidof and Khvostof Islands, Alaska. U.S. Geol. Surv. Bull. 257–266. Nelson, C.H., Hopkins, D.M., Scholl, D.W., 1974. Cenozoic sedimentary and tectonic history of the Bering Sea. Oceanography of the Bering Sea. Institute of Marine Science, University of Alaska Fairbanks, pp. 485–516. Nicolaysen, K.E., Myers, J.D., Linneman, S.R., Lamb, D., 1992. Geologic relations of the Yunaska volcanic complex, central Aleutian Arc. EOS Trans. Am. Geophys. Union. Nostro, C., Stein, R.S., Cocco, M., Belardinelli, M.E., Marzocchi, W., 1998. Two-way coupling between Vesuvius eruptions and southern Apennine earthquakes, Italy, by elastic stress transfer. J. Geophys. Res. 103 (24), 24487–24504. Nye, C.J., McGimsey, G., Power, J., 1998. Volcanoes of Alaska. Alaska Div. Geol. Geophys. Surv., Inf. Circ., p. 38. O'Brien, J.F., Roman, D.C., Dixon, J.P., Power, J.A., Arnold, R., 2012. Multiple causes for noneruptive seismic swarms at Mt. Martin, Katmai Volcanic Cluster, Alaska (2004–2008). J. Volcanol. Geotherm. Res. 229, 13–22. Okada, Y., 1992. Internal deformation due to shear and tensile faults in a half-space. Bull. Seismol. Soc. Am. 82, 1018–1040. Page, R.A., Biswas, N.N., Lahr, J.C., Pulpan, H., 1991. Seismicity of continental Alaska. In: Slemmons, D.B., Engdahl, E.R., Zoback, M.D., Blackwell, D.D. (Eds.), Neotectonics of North America. Geological Society of America, Boulder, Colorado, p. 1 Decade Map. Parsons, T., Thompson, G.A., 1991. The role of magma overpressure in suppressing earthquakes and topography: worldwide examples. Science 253 (5026), 1399–1402. Pasquaré, F.A., Tibaldi, A., 2003. Do transcurrent faults guide volcano growth? The case of NW Bicol volcanic arc, Luzon, Philippines. Terra Nova 15 (3), 204–212. Pasquarè, F.A., Tibaldi, A., 2007. Structure of a sheet-laccolith system revealing the interplay between tectonic and magma stresses at Stardalur Volcano, Iceland. J. Volcanol. Geotherm. Res. 161 (1), 131–150. Pasquarè, G., Francalanci, L., Garduno, V.H., Tibaldi, A., 1993. Structure and geological evolution of the Stromboli Volcano, Aeolian islands, Italy. Acta Volcanol. 3, 79–89 Pisa. Patton Jr., W.W.W., Taylor, F.H., Theresa, A., 2011. Geologic Map of Saint Lawrence Island, Alaska. USGS. Peltier, A., Ferrazzini, V., Staudacher, T., Bachelery, P., 2005. Imaging the dynamics of dyke propagation prior to the 2000–2003 flank eruptions at Piton de La Fournaise, Reunion Island. Geophys. Res. Lett. 32. Plafker, G., 1969. Tectonics of the March 17, 1964, Alaska earthquake. U.S. Geological Survey Professional Paper, 543-I, I1–I74, 2 sheets, scale 1:500,000. Pollard, D.D., Delaney, P.T., Duffield, W.A., Endo, E.T., Okamura, A.T., 1983. Surface deformation in volcanic rift zones. Tectonophysics 94, 541–584. Power, J.A., Jolly, A.D., Page, R.A., McNutt, S.R., 1992. Seismicity and forecasting of the 1992 eruptions of Crater Peak vent, Mount Spurr Volcano, Alaska: an overview. In: Keith, T.E.C. (Ed.), The 1992 Eruptions of Crater Peak Vent, Mount Spurr Volcano, Alaska. 2139. U.S. Geological Survey Bulletin, pp. 149–160. Power, J.A., Villasenor, A., Benz, H.M., 1998. Seismic image of the Mount Spurr magmatic system. Bull. Volcanol. 60 (1), 27–37. Power, J., Jolly, A., Nye, C., Harbin, M., 2002. A conceptual model of the Mount Spurr magmatic system from seismic and geochemical observations of the 1992 Crater Peak eruption sequence. Bull. Volcanol. 64 (3–4), 206–218. Price, E.J., 2004. Dynamic deformation of Seguam Island, Aleutian Islands, Alaska, 1993–2000: implications for magmatic and hydrothermal processes. J. Geophys. Res. 109, B04202. http://dx.doi.org/10.1029/2003JB002671. Reed, B.L., Lanphere, M.A., 1973. Alaska-Aleutian Range batholith: geochronology, chemistry, and relation to circum-Pacific plutonism. Geol. Soc. Am. Bull. 84 (8), 2583–2610. Richter, D.H., Waythomas, C.F., McGimsey, R.G., Stelling, P.L., 1998. Geology of Akutan Island, Alaska. US Geological Survey, pp. 98–135 Open-File Report. Rivalta, E., Taisne, B., Bunger, A.P., Katz, R.F., 2015. A review of mechanical models of dike propagation: schools of thought, results and future directions. Tectonophysics 638, 1–42. Roman, D.C., 2005. Numerical models of volcanotectonic earthquake triggering on nonideally oriented faults. Geophys. Res. Lett. 32 (2). Roman, D.C., Cashman, K.V., 2006. The origin of volcano-tectonic earthquake swarms. Geology 34 (6), 457–460. Roman, D.C., Gardine, M.D., 2013. Seismological evidence for long-term and rapidly accelerating magma pressurization preceding the 2009 eruption of Redoubt Volcano, Alaska. Earth Planet. Sci. Lett. 371, 226–234. Roman, D.C., Power, J.A., 2011. Mechanism of the 1996–97 non-eruptive volcano-tectonic earthquake swarm at Iliamna Volcano, Alaska. Bull. Volcanol. 73 (2), 143–153. Roman, D.C., Moran, S.C., Power, J.A., Cashman, K.V., 2004a. Temporal and spatial variation of local stress fields before and after the 1992 eruptions of Crater Peak vent, Mount Spurr Volcano, Alaska. Bull. Seismol. Soc. Am. 94 (6), 2366–2379. Roman, D.C., Power, J.A., Moran, S.C., Cashman, K.V., Doukas, M.P., Neal, C.A., Gerlach, T.M., 2004b. Evidence for dike emplacement beneath Iliamna Volcano, Alaska in 1996. J. Volcanol. Geotherm. Res. 130 (3), 265–284. Rubin, A.M., Pollard, D.D., 1987. Origins of blade-like dikes in volcanic rift zones. In: Decker, R.W., Wight, T.L., Stuffer, P.H. (Eds.), Volcanism in Hawaii. U.S. Geological Survey Professional Paper 1350, pp. 1449–1470. Ruppert, N.A., Prejean, S., Hansen, R.A., 2011. Seismic swarm associated with the 2008 eruption of Kasatochi Volcano, Alaska: earthquake locations and source parameters. J. Geophys. Res. Solid Earth 116 (B2). Ruppert, N.A., Kozyreva, N.P., Hansen, R.A., 2012. Review of crustal seismicity in the Aleutian Arc and implications for arc deformation. Tectonophysics 522, 150–157. Ryan, H.F., Draut, A.E., Keranen, K., Scholl, D.W., 2012. Influence of the Amlia fracture zone on the evolution of the Aleutian Terrace forearc basin, central Aleutian subduction zone. Geosphere 8 (6), 1254–1273. Sánchez, J.J., Wyss, M., McNutt, S.R., 2004. Temporal-spatial variations of stress at Redoubt Volcano, Alaska, inferred from inversion of fault plane solutions. J. Volcanol. Geotherm. Res. 130 (1), 1–30.

26

A. Tibaldi, F.L. Bonali / Earth-Science Reviews 167 (2017) 1–26

Schaefer, J.R., Scott, W.E., Evans, W.C., Jorgenson, J., McGimsey, R.G., Wang, B., 2008. The 2005 catastrophic acid crater lake drainage, lahar, and acidic aerosol formation at Mount Chiginagak volcano, Alaska, USA: field observations and preliminary water and vegetation chemistry results. Geochem. Geophys. Geosyst. 9 (7). Sepúlveda, F., Lahsen, A., Bonvalot, S., Cembrano, J., Alvarado, A., Letelier, P., 2005. Morpho-structural evolution of the Cordón Caulle geothermal region, Southern Volcanic Zone, Chile: insights from gravity and 40Ar/39Ar dating. J. Volcanol. Geotherm. Res. 148 (1), 165–189. Settle, M., 1979. The structure and emplacement of cinder cone fields. Am. J. Sci. 279 (10), 1089–1107. Simons, F.S., Mathewson, D.E., 1955. Geology of Great Sitkin Island, Alaska. US Government Printing Office. Smith, P.A., 1937. The submarine topography of Bogoslof. Geogr. Rev. 27 (4), 630–636. Snyder, G.L., 1959. Geology of Little Sitkin Island, Alaska (No. 1028-H). US Govt. Print. Off. Spence, W., 1977. The Aleutian arc: tectonic blocks, episodic subduction, strain diffusion, and magma generation. J. Geophys. Res. 82 (2), 213–230. Statz-Boyer, P., Thurber, C., Pesicek, J., Prejean, S., 2009. High precision relocation of earthquakes at Iliamna Volcano, Alaska. J. Volcanol. Geotherm. Res. 184 (3), 323–332. Stauder, W., 1968. Mechanism of the Rat Island earthquake sequence of February 4, 1965, with relation to island arcs and sea-floor spreading. J. Geophys. Res. 73 (12), 3847–3858. Stauder, W., Bollinger, G.A., 1966. The focal mechanism of the Alaska earthquake of March 28, 1964, and of its aftershock sequence. J. Geophys. Res. 71 (22), 5283–5296. Strecker, M., Bosworth, W., 1991. Quaternary stress-field change in the Gregory Rift, Kenya. Eos 72 (3). Sumita, I., Manga, M., 2008. Suspension rheology under oscillatory shear and its geophysical implications. Earth Planet. Sci. Lett. 269 (3), 468–477. Sykes, L.R., 1971. Aftershock zones of great earthquakes, seismicity gaps, and earthquake prediction for Alaska and the Aleutians. J. Geophys. Res. 76 (32), 8021–8041. Syracuse, E.M., Maceira, M., Zhang, H., Thurber, C.H., 2015. Seismicity and structure of Akutan and Makushin volcanoes, Alaska, using joint body and surface wave tomography. J. Geophys. Res. Solid Earth 120 (2), 1036–1052. Tadini, A., Bonali, F.L., Corazzato, C., Cortés, J.A., Tibaldi, A., Valentine, G.A., 2014. Spatial distribution and structural analysis of vents in the Lunar Crater Volcanic Field (Nevada, USA). Bull. Volcanol. 76 (11), 1–15. Tait, S., Jaupart, C., Vergniolle, S., 1989. Pressure, gas content and eruption periodicity of a shallow, crystallising magma chamber. Earth Planet. Sci. Lett. 92 (1), 107–123. Tibaldi, A., 1995. Morphology of pyroclastic cones and tectonics. J. Geophys. Res. Solid Earth 100 (B12), 24521–24535. Tibaldi, A., 1996. Mutual influence of dyking and collapses at Stromboli Volcano, Italy. Geol. Soc. Lond., Spec. Publ. 110 (1), 55–63. Tibaldi, A., 2001. Multiple sector collapses at Stromboli Volcano, Italy: how they work. Bull. Volcanol. 63 (2–3), 112–125. Tibaldi, A., 2003. Influence of volcanic cone morphology on dykes, Stromboli, Italy. J. Volcanol. Geotherm. Res. 126, 79–95. Tibaldi, A., 2004. Major changes in volcano behaviour after a sector collapse: insights from Stromboli, Italy. Terra Nova 16 (1), 2–8. Tibaldi, A., 2005. Volcanism in compressional tectonic settings: is it possible? Geophys. Res. Lett. 32 (6). Tibaldi, A., 2008. Contractional tectonics and magma paths in volcanoes. J. Volcanol. Geotherm. Res. 176 (2), 291–301. Tibaldi, A., 2015. Structure of volcano plumbing systems: a review of multi-parametric effects. J. Volcanol. Geotherm. Res. 298, 85–135. Tibaldi, A., Corazzato, C., Kozhurin, A., Lagmay, A.F., Pasquarè, F.A., Ponomareva, V.V., Rust, D., Tormey, D., Vezzoli, L., 2008a. Influence of substrate tectonic heritage on the evolution of composite volcanoes: predicting sites of flank eruption, lateral collapse, and erosion. Glob. Planet. Chang. 61 (3), 151–174. Tibaldi, A., Pasquarè, F.A., Papanikolaou, D., Nomikou, P., 2008b. Discovery of a huge sector collapse at the Nisyros volcano, Greece, by on-land and offshore geological-structural data. J. Volcanol. Geotherm. Res. 177 (2), 485–499. Tibaldi, A., Corazzato, C., Rovida, A., 2009a. Miocene–quaternary structural evolution of the Uyuni–Atacama region, Andes of Chile and Bolivia. Tectonophysics 471 (1), 114–135.

Tibaldi, A., Corazzato, C., Marani, M., Gamberi, F., 2009b. Subaerial-submarine evidence of structures feeding magma to Stromboli Volcano, Italy, and relations with edifice flank failure and creep. Tectonophysics 469 (1), 112–136. Tibaldi, A., Pasquarè, A.F., Rust, D., 2011. New insights into the cone sheet structure of the Cuillin Complex, Isle of Skye, Scotland. J. Geol. Soc. 168 (3), 689–704. Tibaldi, A., Bonali, F.L., Pasquaré, F.A., Rust, D., Cavallo, A., D'urso, A., 2013. Structure of regional dykes and local cone sheets in the Midhyrna-Lysuskard area, Snaefellsnes Peninsula (NW Iceland). Bull. Volcanol. 75 (11), 1–16. Tibaldi, A., Bonali, F.L., Corazzato, C., 2017. Structural control on volcanoes and magma paths from local- to orogen-scale: The central Andes case. Tectonophysics 699, 16–41. Toda, S., Stein, R.S., Richards-Dinger, K., Bozkurt, S.B., 2005. Forecasting the evolution of seismicity in southern California: animations built on earthquake stress transfer. J. Geophys. Res. Solid Earth 110 (B5). Vezzoli, L., Tibaldi, A., Renzulli, A., Menna, M., Flude, S., 2008. Faulting-assisted lateral collapses and influence on shallow magma feeding system at Ollagüe volcano (Central Volcanic Zone, Chile-Bolivia Andes). J. Volcanol. Geotherm. Res. 171 (1), 137–159. Vigneresse, J.L., Tikoff, B., Améglio, L., 1999. Modification of the regional stress field by magma intrusion and formation of tabular granitic plutons. Tectonophysics 302 (3), 203–224. Wadge, G., Cross, A., 1988. Quantitative methods for detecting aligned points: an application to the volcanic vents of the Michoacan-Guanajuato volcanic field, Mexico. Geology 16 (9), 815–818. Waldron, H.H., 1961. Geologic Reconnaissance of Frosty Peak Volcano and Vicinity, Alaska (No. 1028-T). US Govt. Print. Off. Walker, G.P., 1999. Volcanic rift zones and their intrusion swarms. J. Volcanol. Geotherm. Res. 94 (1), 21–34. Wallace, K.L., McGimsey, R.G., Miller, T.P., 2000. Historically Active Volcanoes in Alaska-A Quick Reference (No. 118–00). Alaska Volcano Observatory. Wallmann, P.C., Pollard, D.D., Hildreth, W., Eichelberger, J.C., 1990. New structural limits on magma chamber locations at the Valley of Ten Thousand Smokes, Katmai National Park, Alaska. Geology 18 (12), 1240–1243. Walter, T.R., 2007. How a tectonic earthquake may wake up volcanoes: Stress transfer during the 1996 earthquake–eruption sequence at the Karymsky Volcanic Group, Kamchatka. Earth Planet. Sci. Lett. 264 (3), 347–359. Walter, T.R., Amelung, F., 2007. Volcanic eruptions following M ≥ 9 megathrust earthquakes: implications for the Sumatra-Andaman volcanoes. Geology 35 (6), 539–542. Walter, T., Schmincke, H.U., 2002. Rifting, recurrent landsliding and Miocene structural reorganization on NW-Tenerife (Canary Islands). Int. J. Earth Sci. 91 (4), 615–628. Watt, S.F., Pyle, D.M., Mather, T.A., 2009. The influence of great earthquakes on volcanic eruption rate along the Chilean subduction zone. Earth Planet. Sci. Lett. 277 (3), 399–407. Waythomas, C.F., 1995. Surficial Geologic Map of Northern Adak Island, Alaska. U.S. Geological Survey. Waythomas, C.F., Miller, T.P., Nye, C., 2003. Preliminary Geologic Map of Great Sitkin Volcano, Alaska (No. 2003–36). US Geological Survey. Waythomas, C., Miller, T.P., Mangan, M.T., 2006. Preliminary volcano hazard assessment for the Emmons Lake volcanic center, Alaska. Scientific Investigations Report 20065248. U.S. Geological Survey. Waythomas, C.F., Scott, W.E., Prejean, S.G., Schneider, D.J., Izbekov, P., Nye, C.J., 2010. The 7–8 August 2008 eruption of Kasatochi Volcano, central Aleutian Islands, Alaska. J. Geophys. Res. Solid Earth 115 (B12). Wood, C.A., 1992. Volcanoes of North America: United States and Canada. Cambridge University Press. Wood, C.A., Kienle, J., 1990. Volcanoes of North America. Cambridge University Press (354 pp). Yamaoka, K., Kawamura, M., Kimata, F., Fujii, N., Kudo, T., 2005. Dike intrusion associated with the 2000 eruption of Miyakejima Volcano, Japan. Bull. Volcanol. 67, 231–242.