Years ago, I submitted to a literary agent this prologue and three chapters of a biography of a friend of mine, called “Fire and Ice: Real-life Adventures of a Volcanologist”, who worked for the US Geological Survey and had pioneered predicting the eruptions of certain kinds of volcanoes. NOVA even did an episode on his work. The agent came on board but for various reasons the project fell apart. Still, I was always proud of the opening. This is a true story.

++++++++++

This world…ever was, and is, and shall be, ever-living Fire,in measures being kindled and in measures going out.        Heraclitus On the Universe

***

“It’s going to blow, and it’s going to blow big. If they don’t evacuate that terminal soon, it’s going to be a disaster.”

Tom Miller, the Scientist in Charge at the Alaska Volcano Observatory, listened intently. He knew deep in his gut that this man, sitting thousands of miles away in an office of the U.S. Geological Survey in Menlo Park, California—this man, whom he had known for only 19 days—was speaking the truth.

“Alright. I don’t have the authority to tell them to evacuate. But I’ll explain the data and give them our conclusions.”

“Let’s hope they listen.”

Tom hung up the phone and stared out his office window at the distant Chugach Mountains. It was only mid-afternoon in Anchorage, but in Alaska in January, mid-afternoon was getting close to twilight time. There was not much daylight left for an evacuation. Tom turned his gaze to that day’s RSAM data lying on his desk. That scientist in Menlo Park had opened his eyes to what this data indicated.

Redoubt was a cone-shaped, ice-mantled stratovolcano located on the west side of the Cook Inlet, 120 miles southwest of Anchorage, a radius that encompassed over half of Alaska’s population. Tom had studied Redoubt’s history. Rising over 10,000 feet in Lake Clark National Park, Redoubt had erupted six times since 1778 when Captain James Cook had observed it “emitting white smoke but no fire.” More smoking had occurred in 1819, 1902, and 1933. Then in 1966, Redoubt had a major eruption, though not a life-threatening one. During the last eruption in 1968, instruments had recorded several explosions of ash clouds that had lasted only minutes or even seconds. Then for the next 21 years Redoubt had slumbered quietly.

That had changed seven weeks earlier on November 20th. A pilot flying a private plane reported seeing wisps of steam emitting from the crater. On December 8th a steam plume, visible from Anchorage, poured out of the crater for almost six hours. But no seismic activity was detected, so the AVO team thought that the plume merely reflected a renewal of geothermal activity. Then five days later, on December 13th, the real show began. Redoubt began generating 23 straight hours of vigorous steaming and an intensifying seismic swarm.

The Alaska Volcano Observatory was a shoestring operation established by the USGS just over a year before. Until then, Tom was the only volcanologist in Alaska. When he was appointed Scientist in Charge of AVO, his team consisted of one assistant volcanologist in Anchorage, and a few seismologists and faculty members at the Geophysical Institute at the University of Alaska in Fairbanks. Redoubt was one of over 40 active volcanoes forming the “Aleutian volcanic arc,” and with three to five of those volcanoes erupting every year, Tom and his team had plenty of activity to focus on. Nevertheless, Redoubt was an active volcano surrounded by populated areas, and it merited serious attention. Even though the AVO staff was small, they did manage to install a seismic net of five instruments on Redoubt the previous October. Fortunately.

The seismic swarm had begun at 10:30 that December morning. As the activity increased into the afternoon, Tom had realized that he had to give out some official notifications. He had called the Alaska Division of Emergency Services, the weather service, the local news media, and the FAA. Not knowing that he already had hard data demonstrating that an eruption was likely within 24 hours, Tom had simply explained that the seismic activity at Redoubt was intensifying in ways that could possibly and eventually result in an eruption.

Tom and his staff had stayed up all that night watching the increasing seismic activity. By 5:00 a.m., December 14th, the seismic swarm had coalesced into high-amplitude tremor. Then at 9:47 a.m., Redoubt erupted.

Within 30 minutes of the eruption, Tom received a call from the USGS in Menlo Park. A volcano seismologist named Bernard Chouet was on the phone introducing himself. Tom explained that he didn’t have much time to talk; he had an erupting volcano on his hands.

“Redoubt?” asked Bernard.

“That’s right.”

“Yeah, last night I could see that it was going to erupt.”

Tom paused. Whoa, he thought.  I’m the Scientist in Charge. Why isn’t this information getting back to me?

“Look, Bernard, I can’t talk to you now, but I do want to talk to you. Give me your phone number.”

Bernard gave him both work and home numbers. Tom didn’t realize then how this man would radically transform his ability to interpret apparently inconsequential data.

Redoubt began ejecting thousands of tons of tephra, volcanic particles ranging in size from extremely fine ash with the consistency of confectioner’s sugar, to large chunks several feet in diameter. On that first morning, the lava dome that had formed in the crater in 1968 exploded, sending a volcanic ash plume over 30,000 feet into the air. That evening, Tom’s staff measured strong volcanic tremor for over three hours. On the morning of the second day, December 15th, Redoubt erupted three times, at 1:52 a.m., 3:38 a.m., and 10:13 a.m.  It was that third eruption that had demonstrated Redoubt’s deadly potential.

That morning, a Boeing 747-400 jetliner from Amsterdam was flying at 28,000 feet inbound to Anchorage, following the path taken by another 747 only 20 minutes before. KLM flight 867 carried 231 passengers and a flight crew of 14. The pilots had been notified of Redoubt’s third eruption 90 minutes earlier, which had sent a volcanic ash plume over 35,000 feet into the air. The prevailing winds had carried the ash cloud over 180 miles northeast of Redoubt where it encountered the KLM jetliner.

Tom reflected on how often people thought that lava is the most dangerous part of a volcanic eruption, in part because of film and television. In reality, volcanic ash is far more dangerous. A volcanic ash cloud is often indistinguishable from ordinary clouds, both visually and on radar. But the KLM pilots recognized the brown ash cloud for what it was, and they received permission from Anchorage to begin climbing above the cloud. Unfortunately, not fully realizing the dangers involved, the pilots plotted their climb through the ash. Upon entering the cloud, the four turbojet engines began sucking in the fine particles, and the silicon in the ash began melting, forming a ceramic-like coating around the hot turbine components.

A modern turbojet engine has three parts: the compressor, the combustors, and the turbine. As air enters the front of the engine, the rotating series of compressor blades raise the incoming air to high pressures. The compressed air then enters the combustors, where it mixes with jet fuel, ignites, and is forced across the turbine blades. The turbine and compressor are connected, and both are forced to turn by the exhaust, thus maintaining airflow through the engine.

The melting ash began forming glassy deposits on the turbine blades, quickly choking the airflow and causing a buildup of pressure in the compressors. In seconds, the compressors in all four engines stalled and the engines functionally shut down.

Nobody on the jetliner understood exactly what had happened, though the pilots quickly realized they had made a wrong choice. As the engines failed, the jetliner began a steep glide down towards the Talkeetna Mountains. For eight harrowing minutes flight 867 fell, dropping over two miles and coming within 6000 feet of the Talkeetnas. The pilots repeatedly tried to restart the engines. Eventually their efforts had combined with the cold Alaskan air to partially break up the glass deposits. They managed to restart two engines, and five minutes after that, all four were back in operation. In 25 minutes the 747 landed safely in Anchorage. The estimated repairs: $80 million, which included replacing all four engines, the electrical and avionics systems, and a sandblasted cockpit windshield.

Within days of the initial eruption, Tom had called Bernard to find out how credibly he had predicted Redoubt’s eruption. Bernard explained his work in modeling LPs, long period events.  He walked Tom through the data and explained how on the afternoon before the eruption, some of his coworkers in Menlo Park, who knew little of Bernard’s work, had called him in to examine the data. Bernard took a methodical look and asked them if it had erupted yet. They were nonplussed and understandably skeptical.

It didn’t take Tom long to see that Bernard knew what he was talking about. He understood the caution that Bernard’s coworkers had exhibited. How could they know on such short notice that Bernard had a well-supported empirical model underlying his prediction? The next morning, Bernard had gone back to his coworkers and asked if Redoubt had erupted yet. Rather than calling Alaska themselves, they urged Bernard to call. After all, he was the one who was interpreting the data. Bernard called Alaska. By then of course, Redoubt had erupted.

Bernard had flown up to Anchorage that December and both he and Tom had spent time together circling Redoubt in a helicopter, observing the formation of a new lava dome. After Bernard had returned to Menlo Park, they stayed in occasional contact. By January 2nd, today, they had been speaking to each other several times a day. Tom had begun appreciating more fully what a volcano seismologist could do.

Tom enjoyed the fact that Bernard confidently interpreted data. He often mused at how difficult it seemed at times to get seismologists to give him interpretations, even when he assured them that he would take full responsibility for any final decisions. Bernard was different. He loved his work and passionately gave his opinions, and Tom respected him for that. They were like souls.

When they had flown around Redoubt, Bernard pointed out that the winter snowpack and glaciers around the volcano were likely to generate more lahars, highly dangerous and fast-moving slurries of water, mud, rock, and sand that result when hot volcanic debris melts snow and ice. It was a lahar that killed over 22,000 people in Armero, Colombia, in 1985 when Nevado del Ruiz erupted. Tom already understood the danger Redoubt posed because the first eruption had caused lahars to cover a portion of the upper Drift River valley that extended northeast of Redoubt. The new lava dome concerned them because it could plug the underlying magma and gases, begin pressurizing, and eventually explode catastrophically. Such an eruption would cause a huge lahar to inundate the lower Drift River valley, and more critically, the Drift River Oil Terminal at the mouth of the Drift River about 21 miles downstream. The oil terminal had already experienced a minor lahar during the initial eruptions.

The Drift River Oil Terminal was built in 1967 by the Cook Inlet Pipeline Company to collect and store oil from offshore platforms through a 40-mile-long pipeline. The oil was then pumped to tankers waiting at a loading platform. Over two dozen oil workers at any one time spent their days and nights at the facility on a rotating schedule.  The facility contained seven storage tanks that held almost as much oil as the Exxon Valdez, which had experienced its well-publicized disaster the previous March and was still very much on everyone’s mind. The terminal also had its own small airport and heliport. Cook Inlet executives had spent time flying around Redoubt, keeping tabs on its activity. Closing the terminal would be extremely costly, even for a few days, so they would only consider such a step when they had clear evidence of danger. They saw the lava dome forming, but they saw little activity otherwise and assumed the main activity had already passed.

But Tom and Bernard knew differently. Tom had flown over the lava dome every day for the last ten days. He observed the steady deformation and over-steepening of the north side of the dome. Instinctively, he knew it would fail catastrophically.

Seismic data began building up on December 30. The five Real-time Seismic Amplitude Monitors (RSAM) on Redoubt showed a series of LPs, a cyclic pattern of harmonic resonances caused by increasingly pressurized magmatic gases. On the morning of January 2nd, Tom, Bernard, and Dr. John Davies, a seismologist at the Geophysical Institute, all agreed that a moderate eruption was likely. AVO issued an eruption alert predicting a moderate eruption within 24 hours. But by that afternoon, the buildup of LPs had gone from linear to exponential. Tom had called Bernard and both had concluded that the LPs meant a much more spectacular eruption, one that would come sooner than later, resulting in potentially disastrous lahars. The Drift River Oil Terminal had to be evacuated.

Tom was 52 years old. He grew up collecting rocks along the shores of Lake Superior. He loved the outdoors, and his exposure to course work in college fueled his interest in every phase of geology. For over 30 years his paid work had also been his favorite hobby. His passion for science still, on occasion, kept him up nights thinking about a breakthrough he had made during the day on some volcanological problem.

Now he faced the mother of all volcanological problems: Convincing people that the time to get out was now.

Staring at the RSAM data on his desk, he picked up the phone and dialed the main office of the Cook Inlet Pipeline Company in Anchorage. The secretary connected him to an executive he had spoken to several times before.

“So what’s the word on our volcano?”

“Bad news. I told you this morning that our seismic data was building up. Well, today’s data shows it increasing exponentially. I’m about to call Emergency Services and the media, but I thought I better call you first in case you want to consider moving your people out of the Drift River Terminal.”

“Well, thanks for the heads up, Tom, but I gotta tell you, we flew over the crater today and we didn’t see any changes in the dome. There’s no fresh lava. It doesn’t look that bad.”

“That’s understandable. What we have here is data indicating that the dome is pressurizing. There’s gas pressure building up under that dome. I just got off the phone with my volcano seismologist in California. He’s looking at the same data and we both agree that Redoubt is going to go and go big. He says it will happen in hours.”

“This is really a huge decision, Tom. If we shut down that facility, think of the cost. The oil in the pipeline will freeze. Starting up again would take weeks. You would have to be absolutely right about this. You were calling for a moderate eruption this morning, one that wouldn’t affect us. Now you’re saying otherwise. How can you be so certain? How do you know?”

Tom tried not to let his frustration show. He couldn’t order an evacuation. How could he convince them? Just then his wife, Shirla, walked into his office. Tom had an idea.

“Tell you what. I’ll fax over this data so you can see for yourself. I don’t care if you understand what the units are, but you have to see the dramatic increase in the seismic parameter, and regardless of what it is, this is what we’re basing our interpretation on.”

Maybe that would get their attention. Tom handed the RSAM plot and the executive’s fax number to his wife and waited. Somehow seeing data with your own eyes carried greater impact. He didn’t have to wait long.

“Jesus.”

“Yeah. I hope it doesn’t take long to close up shop.”

“Yes . . . right. Well, I think we both have calls to make.”

Tom hung up and looked at the clock. It was 1:45 p.m.

By 3:50 p.m. the pumps were shut down, the facilities secured, and the last of the oil terminal employees evacuated by helicopter.

Two hours later at 5:48 p.m., Redoubt exploded. A pilot flying 35 miles south of the volcano reported seeing an orange flame shoot straight up from the summit like a cannon. At 7:27 p.m. a massive second explosion rocked the volcano, destroying 80 percent of the lava dome. A pyroclastic flow of hot ash and avalanches of hot lava blocks roared down the north flank, across the Drift Glacier and up the other side of the valley at almost 100 miles per hour. The hot volcanic ash and 25-foot blocks from the lava dome scoured the glacier, melting ice and snow, uprooting trees three feet in diameter, and creating a lahar that began to fill the Drift River Valley and flow down toward the east.

Picking up momentum and mass as it moved, this monstrous, mud-filled, debris-laden, 1.5-mile-wide wall of volcanic debris thundered down toward the Drift River Oil Terminal.