Notes on Alaska’s Aleutian Arc Volcanoes

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. Faults: DF = Denali; KGF = Kaltag; INF = Iditarod-Nixon Fork; AFZ = Amlia Fracture zone (AFZ). Volcano location is from Global Volcanism Program.

There are far more active volcanoes in Alaska than the rest of the Lower 48 states combined. In fact, the 90 or so volcanoes running in an arc from the west shore of Cook Inlet to the far end of the Aleutian Islands, a distance of some 2,500 miles, are the most volcanoes in a single geologic region in the world.

What’s happening is that the Pacific Plate is subducting – being jammed under – the North American Plate. The Aleutian Trench marks the line where the subduction is occurring. That’s why that area has the most frequent and most severe earthquakes in North America; the Pacific Plate doesn’t slide down smoothly. It moves in sudden jumps. When that subjected Pacific Plate gets shoved down deep enough, it melts and the magma – melted rock – which is less dense, rises up and emerges as volcanic rock.1 That’s the process that has created the volcanoes of the Aleutian Islands and the Alaska Peninsula (collectively, the Alaska-Aleutian Arc).

Cook Inlet Quake Cross-Section
Cook Inlet Quake Cross-Section
Earthquake Activity, Cook Inlet, 1950 - 1996
Earthquake Activity, Cook Inlet, 1950 – 1996, viewed from the surface
Earthquakes of M4.5 and greater from 2010 to 2017 along the Aleutian Trench subduction zone (red line; teeth point in the direction of the subducting slab). White arrows show the directions of plate movement. Circle colours indicate the depths of earthquakes (see legend, lower left). Earthquakes become deeper moving north from the subduction zone. Source: Karla Panchuk (2017) CC BY 4.0. Base maps with epicentres generated using the U. S. Geological Survey Latest Earthquakes website. Visit Latest Earthquakes

The North American Plate is much thicker at mainland Alaska. You can see how steeply the Pacific Plate is forced downwards in the area of Cook Inlet. By contrast, the North American Plate is thinner under the Bering Sea and the Pacific Plate descends much less steeply. There’s volcanism all the way north from the Aleutian Islands to the Bering Straits.

It would be very useful if we could predict when and where a volcano was about to erupt. The advantage of studying the Alaska-Aleutian Arc is that the sheer number of active volcanoes means there is a lot of data. An embarrassment of riches, in fact. An average of about three eruptions a year, and a major earthquake ever 3-4 years.

Two geologists from the University of Milan, Italy, undertook to synthesize in one paper all of the published work, as well as their own research, on the volcanism of the Alaska-Aleutian Arc. It’s a pretty amazing effort. But before turning to that paper, WC must first disabuse some of his readers of how volcanoes erupt. The popular understanding is that there is a giant tube, a “conduit,” that conducts the melted rock from the melt zone deep below, to the surface. Popularizations talk about the “throat” of the volcano. That’s actually very rarely the case. The magma makes its way to the surface through a cluster of dikes. Dikes are vertical tabular structures, cracks in the rock with a usually vertical or steeply angled orientation. Eroded volcanic necks show the chilled dikes; the pattern of eruptive cones on the slopes of volcanos reveal the same dikes. In erupting volcanoes and during pre-eruptive earthquake swarms, geotechnics show the magma migrating along dike. So if you know the orientation of the dikes, you know better where future eruptions are more likely to occur.

What those Italian volcanologists, A. Tibaldi and F.L. Bonali, did was analyze the dike orientation for 47 of the 90 or so volcanoes in the Alaska-Aleutian Arc, focusing mostly on volcanoes that have been active in the Holocene, the last 11,700 years. And what they found is pretty remarkable. The overall orientation of the dikes of the Aleutian Arc is nearly perpendicular to the Aleutian Trench.

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.

That’s not intuitive; you’d expect faults and dikes to parallel the Aleutian Trench which is, after all, the master fault. Instead, the overall orientation of the dikes that feed the Alaska-Aleutian Arc are oriented north-northwest by south-southeast. The dikes seem to be pulled from perfectly perpendicular by the stress of the west-southwest motion of the North American Plate against the northwest movement of the Pacific Plate.

That begs the question of why the dikes – presumably expressions of the stresses in the rock they penetrate – are perpendicular and not parallel to the Aleutian Trench. There are lots of theories, but right now it’s unexplained. But that’s not to take away anything from A. Tibaldi and F.L. Bonali’s paper, an excellent summary of what we know and what is still left to puzzle out.

1 It’s a lot more complicated than that, of course. To a considerable extent, the water carried down with the ocean bottom lowers the melting point of both the subducted plate and the surrounding rock of the continental crust. That’s one of the reason there are so many different kinds of volcanic rocks. What melts depends on a number of variables; it’s not always just subducted plate.

6 thoughts on “Notes on Alaska’s Aleutian Arc Volcanoes

  1. Yup … I would have expected the dikes to be pretty much parallel and especially not perpendicular. While I remember discussions of dikes etc. back in my days at Idaho State I do not remember any discussion related to the dike orientation relative to subduction fault zones. But those duscussions were in “ancient” times, after all I took my last Geology or related Archaeology class in 1971 🙂


  2. Scratching my head over the Holocene volcano symbol north of the Denali Fault. Surely it’s not Donnelly? Seems to be too far west for that, and without looking it up, I believe that volcanism long predates that Era.


    • That’s a good question. It’s approximately in the area of the Cantwell Volcanics, but their last activity was late Miocene, as far as WC knows. Prindle Volcano is further east, near the Alaska-Canada border. WC will try n email to the paper’s authors, but don’t count on a response.



  3. Fascinating! As someone who lives near a more southern Pacific Plate-North American plate subduction zone, I’m always interested in the goings-on around the Ring of Fire.


    • Hayes Volcano is a different mountain (and in a different mountain range) than the Alaska Range’s Mt. Hayes. Hayes Volcano is about 25 miles north-northwest of Mt. Spurr. A possible volcanic neck is at the summit, but the caldera is filled with ice and snow, making it a bit hard to tell. It was active in the middle and late Holocene, evidenced by ash and tephra falls and a lamar. That’s about a 110 miles south of the Denali Fault. More Information:



Comments are closed.