Many concepts and models used for explaining past caldera-forming eruptions and subsequent collapses were formulated by USGS scientist Robert L. Smith and his colleagues, mainly based on their work at Valles caldera in New Mexico. Subsequently, USGS scientist Peter Lipman’s research at Tertiary calderas of the San Juan Mountains in Colorado illuminated many fundamental processes linking large-volume eruptions and caldera collapses. Geologic mapping by USGS scientist Bob Christiansen suggested that the collapse of the Yellowstone caldera 631,000 years ago in response to the eruption of about 1,000 cubic kilometers (240 cubic miles) of magma occurred within hours or days.

However, recent field observations indicate a more complicated story of large caldera forming eruptions.

Schematic illustration showing the 1912 Novarupta eruption in Alaska and the caldera collapse at Mount Katmai ,10.6 km (about 6.6 miles) to the northeast. The illustration was modified from Eichelberger and Izbekov, Philosophical Transactions of the Royal Society of London, 358, 1465–1485 (2000)/Public domain

We can learn more about past volcanic eruptions and caldera collapses from observations of more recent eruptions. The 1912 Novarupta eruption on the Alaska Peninsula was the largest in the 20th century, with a Volcanic Explosivity Index value of 6. There were plenty of eyewitness accounts of the eruption, including observations made a few years later by Robert Griggs. About 15 cubic kilometers (3.5 cubic miles) of magma erupted over the course of 60 hours. An interesting observation was that the caldera did not collapse under Novarupta, but rather magma propagated laterally from beneath Mount Katmai, about 10.6 km (about 6.6 miles) to the northeast and which collapsed due to lateral drainage of its magma chamber. While unraveling details of the Novarupta eruption and the geologic history of the region several decades later, USGS scientists Wes Hildreth and Judy Fierstein were able to use information from the eyewitness accounts, and in contrast to geologic mapping at Yellowstone, they mapped landscapes that were less eroded and rocks that were less altered.

In the late 20th and early 21st centuries, the establishment of volcano monitoring networks allowed scientists to capture several caldera collapses in unprecedented detail. One of the best examples is from the 2018 eruption of Kīlauea Volcano on the Island of Hawai’i. Data collected with seismometers, tiltmeters, high-rate Global Positioning System (GPS) stations, and gas sensors helped document precursory activity, growth of the caldera, and subsurface lateral magma transport from Kīlauea‘s summit to the eruption site in the volcano’s Lower East Rift Zone, not unlike what occurred at Katmai and Novarupta 106 years prior. Over the three months of the eruption, the caldera at Kīlauea‘s summit subsided by more than 500 meters (1640 feet).

Novarupta Dome, with Falling Mountain and the upper valley portion of the Valley of Ten Thousand Smokes in the background. Photo by Tom Miller, June 1979/Public domain

Many types of monitoring instruments captured the onset of subsurface magma transport, the evolving  geometry  of the subsiding caldera and its faults, and the amount of magma erupted during the 2014–2015 collapse of Bárðarbunga Caldera under the Vatnajökull glacier in Iceland. As at Katmai and Kīlauea, magma propagated laterally in the subsurface away from the caldera, erupting 1.5 cubic kilometers (0.35 cubic miles) of basaltic magma at Holuhraun, about 40 kilometers (25 miles) to the northeast.

Other caldera collapses in the 21st century that were captured by volcano monitoring networks include the 2000 collapse at Miyakejima Volcano in Japan and the 2007 collapse of Dolomieu crater at Piton de la Fournaise Volcano on Réunion Island in the Indian Ocean. Because of the increasing capacity to monitor volcanoes from space, even the 2018 eruption and caldera resurgence (in contrast to collapse) at the remote Sierra Negra volcano in the Galápagos Islands (Pacific Ocean) were documented in great detail by multiple instruments.

The ability to capture and document caldera collapse with modern monitoring networks provides scientists with diverse datasets to improve upon the early models proposed by Smith, Lipman, Christiansen, and others. Although a massive caldera-forming eruption like the one that resulted in Yellowstone caldera has never been witnessed, geologists can use modern examples, coupled with insights from the geologic history of the Yellowstone region, to better understand how caldera collapse occurs.

Schematic displaying the general processes associated with collapse of Yellowstone Caldera. (A) Pre-caldera volcanism includes the eruption of dome complexes from the underlying magma chambers. (B) The caldera-forming eruption evacuates a significant amount of magma from the chamber, causing the overlying crustal block to subside into the void space. (C) Post-caldera phase displaying the region immediately after the caldera-forming eruption, with the definition of the topographic rim of the caldera and the highly faulted subsided crustal block. (D) Post-caldera volcanism initiates with the intrusion of magma into the caldera ring faults and faults within the subsided crustal block, producing eruptions of lava flows and dome complexes/Public domain