Steam as Far as the Eye Could Reach
On July 31, 1916, Robert F. Griggs climbed over Katmai Pass and looked down into the Ukak River valley. What he saw stopped him where he stood. The entire floor of the valley was venting. Not one or two fumaroles, not a scattered handful, but thousands of steam columns rising from the ground at once, stretching to the horizon in every direction. He wrote later that the sight was "one of the most amazing visions ever beheld by mortal eye." He named it the Valley of Ten Thousand Smokes.
Four years earlier, none of it had been there. The Ukak River valley had been an ordinary tundra landscape on the Alaska Peninsula: scrub willow, tussock grass, a cold river running to the sea. On June 6, 1912, it ceased to exist. Within sixty hours, the valley had been buried under several hundred feet of superheated pyroclastic material. The river was gone. The ground itself was still hot enough to flash water into steam four years after the eruption. The landscape Griggs walked into had been created from nothing in three days.
The eruption that did this was the largest volcanic event of the entire twentieth century. It ejected more than three cubic miles of magma. It blanketed a town one hundred miles away in a foot of ash. It sent sulfurous haze across the Pacific Northwest, acid rain to British Columbia, and measurable ashfall as far as Virginia. For forty years after it happened, geologists were not even certain which volcano had caused it.
Five Days of Warning
The ground on the Alaska Peninsula had been restless for days before the eruption began. People in the Katmai region had felt earthquakes for at least five days before June 6. On June 4 and 5, tremors intensified, strong enough to be felt 160 miles away. The Alutiiq people of Katmai village read the signs correctly. They loaded their canoes and paddled along the coast toward Cold Bay, covering more than forty miles without stopping.
They were right to go. At approximately 1:00 p.m. Alaskan time on June 6, a vent opened in the floor of the Ukak River valley. Novarupta — the name would come later, meaning "newly erupted" in Latin — had not existed as a geographic feature before that moment. The eruption column shot to an altitude of twenty miles within hours. It was the first and largest of three distinct explosive phases that continued through June 9.
The first phase alone lasted roughly sixteen hours and emplaced about 2.6 cubic miles of pyroclastic material. A second phase followed after only a brief pause, lasting approximately twenty-nine hours. A third, shorter phase concluded the main eruption by June 9. In total, the eruption ejected between 3.1 and 3.4 cubic miles of magma. Mount St. Helens, in its famous 1980 eruption, produced about 0.25 cubic miles. Novarupta produced more than twelve times that volume, and it did so in sixty hours.
What It Did to Kodiak
Kodiak, Alaska sits about 100 miles southeast of the eruption site. By any ordinary measure of distance from a volcanic event, that should have been far enough. It was not.
Ash began falling on Kodiak within four hours of the eruption's start. Within a day, more than a foot of ash had accumulated. Buildings collapsed under the weight. Visibility dropped to near zero. Residents who stepped outside found the air thick with sulfur dioxide gas. Eyes burned. Breathing was painful. Water became undrinkable. The authorities evacuated refugees by boat, loading them onto the U.S. Customs cutter Manning while ash continued to fall.
For approximately sixty hours, complete darkness settled over Kodiak. A lantern held at arm's length could not be seen. One contemporary account described residents not knowing whether it was day or night. The ash that fell was not just heavy — it was chemically active. Sulfuric acid in the airborne material irritated skin and tissue on contact.
How Far the Ash Reached
The eruption column rose more than thirty kilometers into the stratosphere, where the jet stream caught it and carried it east. The scale of the dispersal was extraordinary.
In Juneau, 750 miles from the eruption, people could hear the blasts. The sound arrived more than an hour after the detonations occurred. Juneau also received measurable ashfall. By June 10, a steamer 100 miles off the coast of Vancouver, British Columbia was being covered in a steady shower of ash. Residents in Vancouver and the Puget Sound region reported sulfurous haze. At Port Townsend, Washington, airborne acid from the ash cloud corroded the brass fixtures on automobiles. In Vancouver, acid dissolved linen hung out to dry on clotheslines.
The ash cloud reached Virginia by June 10, North Africa by June 17. It circled the globe. In the stratosphere, sulfur aerosols from the eruption persisted through December 1914, reducing incoming solar radiation across the Northern Hemisphere. Surface temperatures in the Northern Hemisphere dropped by approximately 0.9 degrees Celsius in the months following the eruption. Crops failed across a broad swath of the continent. The summer of 1912 was measurably cooler from the Great Plains to Europe.
The destruction of the village of Savonoski was total. The eruption covered more than forty square miles of surrounding land with ash, in some places seven hundred feet deep. The Alutiiq communities that had survived by evacuating early did not return. Their ancestral lands had been transformed into something unlivable.
Yet no one died in the eruption itself. The five days of earthquakes had served as warning enough for those closest to the source. Miraculously, the largest volcanic eruption of the twentieth century produced zero direct fatalities.
The Valley That Was Filled Overnight
The pyroclastic flows from Novarupta traveled down Knife Creek and into the upper Ukak River valley. A pyroclastic flow is not lava. It is a dense, superheated mixture of volcanic gas, ash, and rock fragments that moves at extraordinary speed — in this case, fast enough to fill a wide river valley to a depth of several hundred feet before the ash had time to cool. Near the Novarupta vent, the deposit reached approximately 700 feet deep. The flows covered roughly 120 square kilometers of valley floor.
The Ukak River simply vanished beneath the deposit. The tundra and scrub vegetation beneath it were incinerated. By the time the eruption ended on June 9, the Ukak River valley had become a flat, featureless plain of hot grey pumice.
What happened next was what Robert Griggs found in 1916. Buried beneath the pyroclastic deposit were the remnants of a river system and its associated groundwater. Buried beneath it were glaciers and snowfields. As heat from the still-cooling deposit interacted with this trapped water, steam vented through thousands of cracks and fissures in the surface. The fumaroles that gave the valley its name were not volcanic gas rising from new magma below. They were water, vaporized by the residual heat of material that had landed there four years earlier. Griggs initially believed the smokes indicated ongoing volcanic activity at depth. Later research corrected this: the heat was surface-level, cooling slowly, and the fumaroles reflected it.
"The whole valley as far as the eye could reach was full of hundreds, no thousands — literally, tens of thousands — of smokes curling up from its fissured floor."
— Robert F. Griggs, National Geographic Society, 1916
By the 1930s, most of the fumaroles had gone cold. The residual heat had dissipated. The Valley of Ten Thousand Smokes became what it is today: a flat expanse of solidified ash, streaked with the mineral deposits left by extinct steam vents, cut through by new drainages that have spent a century carving channels into the pyroclastic plain.
The Wrong Volcano
Mount Katmai stands about six miles northeast of the Novarupta vent. It is a large stratovolcano, prominent on the Alaska Peninsula skyline. When the 1912 eruption occurred, the obvious assumption was that Katmai had caused it. The ash, the violence, the scale — it had to come from a major volcano. Katmai was the major volcano in the area.
This assumption went unchallenged for decades. The eruption was routinely called the Katmai eruption. Katmai National Monument, established by President Woodrow Wilson in 1918, took its name from the mountain. Scientific papers referenced Katmai as the source. No one had yet climbed to Katmai's summit to see what had actually happened there.
When Robert Griggs and his team finally reached the summit in 1916, they found a caldera. The entire top of the mountain had collapsed inward, leaving a crater roughly 1.5 miles wide and 500 meters deep. The interior cliffs were fresh, glacier-free, and fuming. This appeared to confirm that Katmai had been the eruption vent. Fumaroles ringed the caldera floor. The mountain looked like a volcano that had just erupted.
The problem was that the caldera showed no evidence of having ejected anything. A volcano that erupts produces outward evidence — ash deposits radiating from the vent, flow patterns, ballistic material. Katmai's caldera had collapsed inward. The geometry was wrong for a primary eruption vent. But recognizing this required detailed geological fieldwork that did not happen until the 1950s.
The Source Confirmed: 1953
Systematic fieldwork at Katmai did not resume until 1953, more than forty years after the eruption. What UC Berkeley geologist Garniss Curtis found when he mapped the area in detail fundamentally changed the picture.
The ash-flow deposits — the pyroclastic material that had filled the Ukak River valley — did not radiate from Katmai. They radiated from Novarupta. The distribution of fall deposits, the flow directions preserved in the solidified material, the location of the lava dome that formed at the close of the eruption: everything pointed to the small vent in the valley floor, not to the large mountain six miles away.
Curtis's work established what had actually happened. The magma that erupted had been stored beneath Mount Katmai. But it had not risen through Katmai's summit. Instead, it had migrated laterally through the crust, traveling approximately six miles underground, and broken the surface at Novarupta. As the magma drained from beneath Katmai's peak, the structural support for the summit disappeared. The mountain collapsed into the void left by its own evacuated magma chamber, creating the caldera that Griggs had mistaken for an eruption vent.
The process is now understood as a lateral magma migration event. The magma reservoir beneath Katmai fed an eruption six miles away. The erupting vent and the collapsing mountain were connected underground but appeared to be separate events on the surface. This was not, in 1912 or in 1916, a recognizable geological pattern. It has since become one of the most studied examples in volcanology of how a caldera can form at a location far removed from the actual eruption vent.
Novarupta had not existed before June 6, 1912. The vent that produced the largest eruption of the twentieth century was brand new when it opened. By the time it was correctly identified as the source, forty-one years had passed.
What the Eruption Is Still Called
Popular accounts of the 1912 eruption still frequently call it the Katmai eruption. The national park is named Katmai National Park and Preserve. The caldera lake at Katmai's summit is a famous landmark. The bear-viewing platform at Brooks Camp, a hundred miles from the eruption vent, is the most-visited attraction in the park. Katmai is the name people know.
Mount Katmai did not erupt in 1912. Katmai collapsed. The eruption vent was Novarupta, a new fissure that opened six miles to the southwest. Magma stored beneath Katmai traveled laterally underground and erupted at Novarupta. The loss of that magma removed the structural support for Katmai's summit, which then collapsed inward to form the caldera now visible in the park. The two events happened simultaneously and are causally linked — but they are not the same event. The eruption was Novarupta's. The caldera formation was Katmai's. Calling it the Katmai eruption is geologically imprecise and has contributed to four decades of scientific confusion about where the magma actually came from.
The fumaroles in the Valley of Ten Thousand Smokes were not evidence of ongoing volcanic activity. Robert Griggs initially believed the steam vents indicated active magma at shallow depth beneath the valley. Later research established that the fumaroles were produced by buried water and snow flashing to steam through the still-hot pyroclastic deposit. By the 1930s, as the deposit cooled, the fumaroles extinguished. The valley has been essentially cold and silent since.
This matters beyond mere terminology. The confusion about the eruption's source shaped scientific understanding of large volcanic events for half a century. It took detailed post-war fieldwork and a generation of advances in petrological analysis to establish that a magma chamber can feed an eruption miles from its own location. That discovery, made at Katmai in the 1950s, is now a foundational concept in how volcanologists model large eruptions. The misattribution was not a trivial error. It was a blind spot in the science that took four decades to correct.
The Park, the Valley, and What Remains
Robert Griggs returned to the Katmai region in 1917 and 1919, leading National Geographic Society expeditions that produced the first systematic scientific documentation of the valley and the eruption's effects. His advocacy for the area's protection led directly to the establishment of Katmai National Monument by President Woodrow Wilson in September 1918 — just six years after the eruption and two years after Griggs first saw the valley. In 1980, Congress upgraded the monument to Katmai National Park and Preserve. It now covers more than four million acres.
The Valley of Ten Thousand Smokes remains the park's most distinctive geological feature, though it is no longer steaming. Visitors reach it by bus from Brooks Camp, a forty-mile trip on unpaved road. What they find at the overlook is a vast flat plain of ash, cut by river channels that have formed since 1912, bounded by mountains that the eruption left unchanged. The pyroclastic deposit is still there, hardened and eroded but essentially intact, a permanent record of three days in June 1912.
Novarupta's lava dome sits at the far end of the valley, roughly 800 feet in diameter and 200 feet high, formed from viscous lava that pushed up through the vent after the explosive phase ended. It is the only surface feature that directly marks where the eruption occurred. Six miles away, the caldera lake at Katmai's summit reflects sky into a depression that did not exist before June 1912 — the cold, still record of a mountain that lost its interior in a matter of hours.
The Alutiiq communities displaced by the eruption never returned to the Katmai coast. The village of Savonoski was gone. Those who survived established New Savonoski on the Naknek River. The National Monument designation in 1918 formalized their exclusion from the ancestral lands the eruption had destroyed. Their oral traditions had preserved the protocols for surviving such events. Those protocols had worked. The land they contained the knowledge of was altered beyond return.
The largest volcanic eruption of the twentieth century killed no one at the source. It destroyed a landscape, displaced a people, altered the global climate for two years, sent acid rain to British Columbia, and created a valley that a botanist would walk into four years later and find so extraordinary he could not find words adequate to it. The vent that caused all of this had not existed the morning it opened.