When it comes to volcanoes of the western U.S., the towering peaks of the Cascade Range poking through the clouds of Oregon and Washington may come to mind. Among these volcanoes stands the iconic Mount Saint Helens, which violently erupted on May 18, 1980.

But there are other volcanoes that dot the surface farther east, like the San Francisco Volcanic Field of northern Arizona. This is a distributed volcanic field that sprawls across a few thousand square kilometers, pocked with signs of mostly diminutive basaltic eruptions.

Yet, distributed volcanic fields are among the most widespread forms of volcanism. They occur around the world, in a variety of tectonic settings. And should one begin to erupt — particularly if the outpouring is of the silica-rich, explosive type — this would pose a hazard in populated regions. 

In a new study published in Geology, scientists led by Ryan Porter of Northern Arizona University investigate details of the San Francisco Volcanic Field by seismically imaging the subsurface. They found two zones of melt, as well as distinct changes in crustal thickness. These findings are focused beneath San Francisco Mountain — a curiously felsic (silica-rich) volcano in a field of mostly basalt.

A stratovolcano among cinder cones

The college town of Flagstaff, Arizona, sits on the southern flanks of the San Francisco Peaks, which rise above the rest of the San Francisco Volcanic Field. The entirety of the volcanic field, which began to erupt about 6 million years ago, sits atop the southern margin of the Colorado Plateau.

Surrounding the San Francisco Peaks are more than 600 mafic (rock rich in both iron and magnesium) eruptive centers that spurted just once — they’re monogenetic. Distributed volcanic systems often feature these single-use cones because limited magma supply cannot maintain a continuous path to each vent. Instead, an isolated dike may feed an eruptive center, producing a cinder cone where melt meets surface.

In the San Francisco Volcanic Field, these cones become younger toward the northeast. The volcanic field is moving at roughly the rate at which North America moves over the mantle, says Porter, assuming a reference frame in which the mantle holds steady and everything else moves.

In addition to the basaltic bumps, eight silicic dome complexes punctuate the landscape, having migrated in the same direction and at the same rate. But these volcanic complexes are much bigger than their cinder cone cousins.

The San Francisco Peaks of today mark the largest of these silicic volcanic complexes. Here, Arizona’s highest point, Mount Humphreys, hosts hikers climbing skyward much of the year, with winter bringing on skiers and snowboarders. But in the not-so-distant past, there was once a much higher single stratovolcano called San Francisco Mountain, built by multiple eruptions spanning 0.9 and 0.4 million years ago. San Francisco Mountain spewed about 20% of the eruptive volume of the field, with a total erupted volume estimated at around 110 square kilometers.

There are other such edifices in the region, like Sugarloaf Mountain and O’Leary Peak — the youngest silicic volcanoes in the field, younger than 91,000 years old. But these and other felsic domes in the area were never stratovolcanoes like the great San Francisco Mountain. What could have caused such a massive volcano in a field of otherwise petite prominences?

A Sunset Crater deployment

Sunset Crater Volcano National Monument contains the youngest and biggest of the cinder cones. This volcano erupted around 1085 C.E., affecting the Indigenous populations that lived in the area. Like its less dramatic cinder cone brethren, it erupted just once.

To study the subsurface of the San Francisco Volcanic Field, scientists led by Eric Kiser at University of Arizona deployed an array of nodal seismometers with an approximately 4 kilometer station spacing centered on Sunset Crater. An inner array of 90 stations were deployed for about a year. Additional stations were installed for about a month. There are an additional 31 stations installed in the region for other reasons, Porter says. These data are available via the NSF National Geophysical Facility archive.

Using these data, Porter and colleagues mapped velocity contrasts beneath the field using receiver functions. This process involves examining seismic waves from distant earthquakes to see where P-waves are converted to S-waves in the subsurface. These conversion zones denote velocity contrasts. These velocity contrasts can help scientists map the location of low-velocity zones, which may indicate the presence of melt. Scientists can also map the Mohorovičić discontinuity, or Moho, which marks the boundary between Earth’s crust and the dense mantle rocks below.

Moho depths

Across the area, Porter and colleagues found several major velocity contrasts. For example, they identified the base of a regional high-velocity layer in the upper crust, likely the Proterozoic southern Yavapai block. This is similar to the crust of the nearby Four Corners region.

However, beneath San Francisco Mountain, they found a previously identified low-velocity zone. This zone, originally discovered in 1982, was originally estimated to span depths of 8 to 30 kilometers. “The top of this low velocity zone is at 15 to 18 kilometers,” says Porter, and the bottom is about 8 to 10 kilometers below the top. At even greater depths, the team found the Moho — the crust-mantle boundary — at about 40 kilometers depth, with another low-velocity layer at about 50 kilometers.

In general, the depth to Moho increases eastward, from about 36 kilometers to about 57 kilometers. The team points out that in the west, the thinner crust may have allowed for higher-relief topography than the thicker crust on the eastern side of the San Francisco Volcanic Field, which features more a more subdued landscape.

The thickness of the crust does not, however, gradually increase eastward. Instead, east of the San Francisco Peaks, the crustal thickness changes somewhat suddenly, with a stepwise increase of about 5 kilometers.

Melt, missing crust, and mountains

Porter and colleagues infer that the two low-velocity zones beneath the San Francisco Peaks represent zones of partial melt in the mid- and lower crust. These zones likely fed the eruptions that formed the original San Francisco Mountain. That we can see the remnant melt zones today suggests that ascent may have stalled. But why?

The mid-crustal melt may be stuck at the brittle-ductile transition — where the crust tends to flow instead of break. Additionally, this low-velocity zone aligns with a 2009 magmatic swarm of seismicity, further supporting the presence of melt instead of anomalously warm crust.

But what about the other low-velocity zone? This could be another zone of partial melt that’s stuck at the base of the crust, perhaps resulting from the pooling of mafic melt. The top of this zone is shallower than the adjacent Moho, suggesting that the melt has penetrated the lower crust. A similar relationship between melt and Moho has been observed beneath Mount Saint Helens, the authors note.

Then there is the 5-kilometer jump in crustal thickness that occurs just east of the San Francisco Peaks. There’s neither an obvious surface expression of this change, nor is there a major terrane boundary. The authors argue that this could be the eastward extent of lithospheric removal, an idea supported by xenoliths collected about 50 kilometers south of the San Francisco Volcanic Field. So, something caused the dense lower crust and underlying mantle lithosphere of this part of the Colorado Plateau to disappear.

One possibility is that the former Farallon plate, which subducted at a low angle toward the end of its existence at these latitudes, may have eroded away the lower crust from below. Another possibility is downwelling, whereby the dense, cold root sank via delamination in one go, or over time as a drip.

Regardless of why the lower crust and underlying mantle lithosphere is gone, the result is a much thinner crust that is easier for melt to invade. This removal of the underside of this part of the Colorado Plateau could also explain why crustal thickness negatively correlates with surface elevation. The high topography to the west, rising above its eastern counterparts, could be caused by a now-buoyant crust no longer tethered to its high-density root.

This feature could also explain the northeastward march of volcanism, and the existence of the San Francisco Mountain. The silica-rich volcanism may have focused itself where downwelling stalled, sustaining the connections needed between melt and surface for multiple eruptions. 

Future volcanism?

Porter and colleagues have demonstrated that San Francisco Mountain sits at the boundary between intact and removed crust. The boundary itself seems to have encouraged the concentration of melt into mid- and lower-crustal reservoirs. These reservoirs likely fed felsic volcanism — the type that tends to be explosive — in a volcanic field of mostly runny basalt.

What happens should the melt residing in the crust start to move again? The melt sits not far from vents that are less than 100,000 years old. Should melt migrate upward, a future felsic eruption is a slim possibility.

“A future eruption is likely to be a cinder cone in the eastern part of the field,” Porter says. “I do think the odds of the San Francisco peaks erupting are very low.”