Most mornings, Katia Krafft would wake up, eat breakfast, and then hike directly toward something that could vaporize her in seconds. She called it love. Her husband Maurice called it science. They both died doing it in 1991 when Mount Unzen’s pyroclastic flow—a 700-degree avalanche of gas and rock—swallowed them whole.
When Your Office Is Literally Trying to Kill You Every Single Day
Here’s the thing about volcanology: it’s one of the few professions where “close enough” means you’re probably already dead. The Kraffts got within a hundred meters of lava flows regularly, closer than anyone thought sane. They filmed Nyiragongo’s lava lake in the Democratic Republic of Congo in 1977, just before it drained and killed around 600 people. Their footage remains some of the most stunning—and disturbing—volcano documentation ever captured.
But wait—maybe the danger isn’t even the interesting part.
Modern volcanologists spend more time staring at computer screens than dodging molten rock. They analyze seismic data, sulfur dioxide emissions, ground deformation measurements. It’s less Indiana Jones, more data analyst with a death wish. When Mount Pinatubo showed signs of waking up in 1991, scientists had to convince 60,000 people to evacuate based on squiggly lines on graphs and gas readings. They were right—the eruption killed around 350 people, but would have killed tens of thousands without the warning.
Turns out predicting volcanic eruptions is like predicting when your temperamental aunt will finally lose it at Thanksgiving dinner. You know it’s coming. You see the warning signs. But the exact moment? That’s where things get messy.
The Part Where Everything You Think You Know Is Wrong
Most people imagine volcanologists as adventurers in heat-resistant suits, but the reality involves grant applications, peer review, and arguing about milenia-old rock samples. David Johnston spent years studying Mount St. Helens, measuring gases, taking samples, building predictive models. On May 18, 1980, he was manning an observation post six miles from the summit when the entire north face of the mountain collapsed. His last words, radioed to colleagues: “Vancouver! Vancouver! This is it!” They never found his body.
The mountain had given warnings—earthquakes, bulging, steam vents—but nobody had seen a lateral blast like that before. It traveled at 300 miles per hour.
That’s the cruel joke of volcanology: every volcano writes its own rulebook, and you only get to read it after the test is over. Kilauea in Hawaii had been erupting continuously from 1983 to 2018—thirty-five years of relatively predictable behavior. Then in 2018 it suddenly drained its summit lava lake, opened new fissures in residential areas, and destroyed over 700 homes. Volcanologists saw it coming days before, but “days before” doesn’t help much when your house is in the path.
Why Anyone Would Voluntarily Stand Next to Geological Blowtorches
Jess Phoenix runs an organization that takes people on volcano expeditions. She’s been doing fieldwork since 2004, collecting samples from active vents, mapping lava flows, occasionally running from unexpected explosions. When asked why, she talks about understanding Earth’s plumbing system—how the planet recycles its crust, how volcanic soils feed half the world’s agriculture, how geothermal energy could power entire nations.
But honestly? Most volcanologists are chasing something else entirely. That moment when you’re standing on fresh lava—rock that didn’t exist yesterday—watching the Earth literally create new land. Paricutin volcano in Mexico emerged in 1943 in a farmer’s cornfield, grew to 1,100 feet in a year, and volcanologists documented every stage. That’s about as dramatic as geological birth gets.
The science itself reads like chaos theory had a baby with chemistry. Magma composition determines eruption style—rhyolitic magma is thick and explosive (think Mount St. Helens), basaltic magma flows easily (think Hawaiian volcanoes). But then you add dissolved gases, crystal content, conduit geometry, and suddenly you’re juggling twenty variables that all interact in unpredictable ways.
The Tedious Parts Nobody Mentions Because They’re Aggressively Boring
For every dramatic helicopter flight over a lava lake, there are months of lab work. Analyzing debre from past eruptions under microscopes. Running chemical assays on volcanic gases. Building computer models that crash repeatedly. Writing papers that get rejected. Rewriting papers. Teaching undergraduates who think volcanology is all adventure and no calculus.
Clive Oppenheimer spent years studying Erebus volcano in Antarctica—one of Earth’s few permanent lava lakes—which meant camping in minus-40-degree temperatures while simultaneously being near 1,700-degree lava. He’s also spent countless hours in libraries studying historical eruption records, because sometimes the best data comes from 17th-century ship logs describing mysterious fogs and blood-red sunsets.
The 1815 eruption of Mount Tambora in Indonesia killed approximately 71,000 people directly, but its global climate effects caused the “Year Without a Summer” in 1816, leading to crop failures and famines that killed hundreds of thousands more. Volcanologists studying Tambora today are trying to understand whether we’d see warning signs if something similar was brewing. The answer, frustratingly, is: maybe?
Twenty-four people died studying volcanoes between 1600 and 2017. Not many, considering how many scientists work near active vents. But each death reshapes the field’s safety protocols, its ethics, its understanding of acceptable risk. After the Kraffts died, volcano tourism became more regulated. After David Johnston died, monitoring stations moved farther back.
The volcanologists who survive aren’t necessarily the most careful—they’re often just the lucky ones who happened to be somewhere else when the mountain decided to rewrite its rules.
