There are two accounts of what happened at the Aleshka Deep Ice Station in the winter of 2024. Both accounts are complete. Both account for every piece of evidence, every timestamp in every log, every witness statement from every member of the four-person research team. Both are internally consistent. Both satisfy Occam's razor, though they cut in different directions. Both are believed, with equal conviction, by the person who was there.

The problem is that they contradict each other at the most fundamental level. And the person who was there, Dr. Linnea Isaksen, a forty-one-year-old glaciologist from the University of Tromsø who had spent seventeen field seasons on the ice, has refused to choose between them. She has stated, in her official report to the Norwegian Polar Institute, that both accounts are true. The institute has not known what to do with this. Neither has the insurance investigator. Neither have the three journalists who have called her satellite phone in the six months since.

Here is the first account. Read it carefully. It will make sense. It will seem complete.

The Aleshka Deep Ice Station was a cluster of three prefabricated modules sitting on the flank of the Brooks Range in northern Alaska, sixty-eight degrees north, two hundred and forty kilometers from the nearest road. The modules were connected by insulated walkways. The generator ran on diesel flown in by bush plane every six weeks. The internet came through a Starlink terminal bolted to the roof of Module B, and the bandwidth was sufficient for Zoom calls and Linux kernel updates and not much else.

Dr. Isaksen had come to Aleshka to retrieve the ice cores. Seventeen years of work, forty-one cores extracted from a depth of eighteen hundred meters, each core a vertical archive of atmospheric chemistry stretching back one hundred and twenty thousand years. The cores were stored in a refrigerated vault beneath Module C, suspended in aluminum racks, wrapped in polyethylene, catalogued by a system that Isaksen had designed herself during her postdoc in 2007. The cores were everything. They were her dissertation, her tenure case, her three dozen published papers, her keynote at the 2022 European Geosciences Union. They were the evidence she had presented to the Norwegian parliament in 2019, the data that had persuaded six legislators to support a carbon tax that was still, precariously, in effect.

The cores were also, she understood, a kind of elegy. Trapped in each meter of ancient ice was a bubble of atmosphere from a world that no longer existed. Carbon dioxide at two hundred and eighty parts per million. Methane at seven hundred parts per billion. A fingerprint of a planet that had not yet been burned. Reading the cores was like reading a letter from a dead relative. Every measurement was a reminder of what had been lost, and what was still being lost, and what could not be recovered even if every coal plant on earth shut down tomorrow.

On the night of November 17, 2024, the power failed.

The first account says it was a solar flare. A coronal mass ejection of class X2.8, detected by NOAA's Space Weather Prediction Center at 19:42 UTC, striking the magnetosphere at 21:15, inducing geomagnetic currents that surged through the generator's voltage regulator at 21:17 local time. The regulator, a fifteen-year-old component that had been flagged for replacement in three consecutive maintenance reports, failed catastrophically. The surge propagated through the station's electrical system, tripping every breaker that was still functional and bypassing the ones that were not.

The backup generator, a propane-fired Cummins that had been serviced in September, attempted to start. Its starter motor engaged. The ignition fired. And then the second thing happened, the thing that the first account attributes to pure, indifferent chance: the propane line to the backup generator had developed a pinhole leak sometime in the preceding two weeks, a leak so small that the pressure sensor had not registered it, and when the ignition fired, the accumulated propane in the generator housing ignited with a force sufficient to shear the mounting bolts and sever the fuel line entirely.

The station went dark at 21:19. The temperature in Module C, where the cores were stored, was maintained at negative twenty-five Celsius by an electrically powered refrigeration unit. With no generator, no backup, and no solar because the sun had set in October and would not rise again until February, the refrigeration failed.

The first account says that Dr. Isaksen did everything correctly. She radioed the University of Alaska Fairbanks field operations center at 21:22. She instructed her three team members to gather at the emergency shelter in Module A and inventory their supplies. She calculated, correctly, that the station's passive insulation would keep Module C below negative fifteen for approximately forty-eight hours, and that a rescue flight from Fairbanks could reach the station in thirty-six.

What she did not know, what no one could have known, was that the storm that had been building over the Chukchi Sea for forty-eight hours was accelerating faster than any model had predicted. The rescue flight was grounded at Fairbanks. The backup rescue flight from Anchorage was grounded at Anchorage. The satellite phone connection to the National Weather Service produced a grim, static-laced prognosis: the station would be unreachable for five days minimum, possibly seven.

The first account says that the cores melted. The temperature in Module C rose above negative fifteen at 14:00 on November 19. Above negative five at 02:00 on November 20. Above zero at 11:00 on November 20. The ancient ice became ancient water. The gas bubbles trapped for a hundred millennia escaped into the Alaskan air, a final exhalation of Pleistocene atmosphere into an Anthropocene sky.

The first account ends with Dr. Isaksen standing in Module C on the morning of November 23, when the rescue team finally arrived, looking at the empty aluminum racks, the puddled polyethylene wrappings, the drip-drip-drip of seventeen years of work reduced to a wet stain on the floor. The first account says she stood there for eleven minutes without speaking. The first account says she did not cry. The first account says she filed her report with the institute in January and took an indefinite leave of absence and has not returned to the ice.

The first account is complete. It explains everything. It is probably correct.

Now here is the second account. It also explains everything. It is also probably correct. Do not choose between them. The choice is not yours to make.

The second account agrees with the first on every factual detail. The solar flare. The voltage regulator. The pinhole leak in the propane line. The grounded rescue flights. The melting of the cores. The empty aluminum racks. The eleven minutes of silence.

Where the second account diverges is here: Dr. Linnea Isaksen let it happen.

This is not to say she caused it. The solar flare was real. The generator failure was real. But the second account insists that Linnea Isaksen, in some recess of her consciousness that she has never fully illuminated, did not want the cores to survive. The second account points to evidence that the first account ignores. Not contradictory evidence, not evidence that disproves the first account, but evidence that the first account simply does not address, because the first account is a narrative of physics and engineering and weather, and physics and engineering and weather do not have unconscious minds.

The evidence, according to the second account:

Item one. On the morning of November 17, eight hours before the solar flare, Dr. Isaksen wrote an email to her ex-husband, a geochemist at the University of Bergen named Henrik who had left her in 2021 for a postdoc he had met at a conference in Oslo. The email was long, digressive, and not sent. It was saved to her Drafts folder on the station's Mac Studio, where it was recovered by the data forensics team that investigated the incident. The email described, at length, the feeling of standing in Module C and knowing that each core contained a precise, irrefutable record of a crime. Not a crime against the ice. A crime against the future. The email asked Henrik, rhetorically, what the point was of measuring a wound that everyone had already agreed to ignore.

Item two. In her personal notebook, the leather-bound Moleskine she had carried since graduate school, an entry dated November 12, five days before the incident, reads: "The data is perfect. The data will be cited for fifty years. The data will change nothing. If the data burned tonight, the world would wake up tomorrow and emit exactly the same amount of carbon as yesterday. The fossil fuel lobby would issue exactly the same press release. The stock market would open at exactly the same level. What is the word for something that is simultaneously priceless and worthless?"

Item three. During the seventy-two hours of darkness, when the station was running on battery lanterns and the emergency propane heater, Dr. Isaksen was observed by her research assistant, a twenty-seven-year-old postdoc named Sanne, to exhibit what Sanne later described in her deposition as "an unusual degree of calm." When Sanne asked whether they should attempt to move the cores to the emergency shelter, where the propane heater might preserve them a few hours longer, Dr. Isaksen replied that the emergency shelter protocol did not permit contamination of the shelter environment with non-essential equipment. The cores, she said, were non-essential equipment.

Sanne, in her deposition, paused here. The court reporter noted the pause. Sanne said, "The cores were literally her life's work. And she called them non-essential equipment. And the thing is, her tone was completely neutral. Like she was reading a weather report."

Item four. The email to Henrik. After the forensics team recovered it, the institute's HR department sent a counselor to speak with Dr. Isaksen. The counselor asked why she had not sent the email. Dr. Isaksen replied, "I didn't want Henrik to think I was asking for something I wasn't asking for." The counselor asked what she wasn't asking for. Dr. Isaksen said, "Permission to stop."

The second account does not claim that Dr. Isaksen sabotaged the generator. The second account does not claim that she delayed the rescue call or misled the team about the temperature projections or did anything at all that could be characterized as negligence, malfeasance, or professional misconduct. The second account claims something more subtle and more disturbing: that somewhere in the space between the woman who had spent seventeen years extracting truth from ice and the woman who had come to understand that truth was insufficient, a door had opened. And when the door opened, the woman who had come to understand walked through it, and the woman who had spent seventeen years did not stop her, because both women were the same woman, and both women were right.

There was a moment, according to the second account, on the night of November 19, when the temperature in Module C was falling past the point of no return. Dr. Isaksen was alone in the walkway between Module A and Module C. She could see the vault door at the end of the corridor, lit by the red glow of the emergency exit sign. She could hear, in the humming silence of the dead station, the faint crackle of ice adjusting to a world it was not designed for.

The second account says she did not open the vault door. The second account says she stood in the walkway for eleven minutes, the same eleven minutes she would later stand in the flooded vault, and she listened to the ice melting, and she felt something that was neither relief nor grief but a third emotion for which no word exists in Norwegian or English or any language she had ever encountered in any ice core from any depth in any century.

The second account says that when the rescue team arrived and found her standing before the empty racks, she was not grieving seventeen years of work. She was grieving seventeen years of believing that measurement was the same thing as meaning. She was grieving the version of herself who had thought that if you wrote the truth down clearly enough, the world would have no choice but to read it. She was grieving the naivety of a woman who had not yet understood that the fossil fuel industry had offices in every capital on earth and that six Norwegian legislators were a rounding error in the global carbon budget.

The second account ends, like the first, with Dr. Isaksen filing her report and taking leave and not returning to the ice. But the second account adds a detail that the first cannot accommodate: on the flight back to Tromsø, somewhere over the Greenland Sea, she took out her Moleskine notebook and wrote a single sentence on the last page.

The sentence reads: "I let it all go, and I would let it go again."

Both accounts are true. Both accounts explain everything. Both accounts are believed by the person who was there. Do not ask which account is correct. The question is not answerable, not because the evidence is insufficient, but because reality, at the Aleshka Deep Ice Station in the winter of 2024, was a superposition of two states that could not and cannot and will not ever collapse into one.

Dr. Linnea Isaksen is currently living in Oslo. She has not applied for another research grant. She has not published another paper. She has not returned any of the three phone calls from the insurance investigator, who is trying to determine whether the loss of the cores constitutes an insurable accident or an uninsurable act of God or something else entirely, something that the policy's language did not anticipate and cannot name.

She has, however, kept the Moleskine notebook. It sits on her nightstand, next to the battery lantern and the weather radio she brought back from the station. Sometimes, before she sleeps, she reads the sentence on the last page. And sometimes, on nights when the wind blows from the north and the temperature drops below freezing, she adds another sentence beneath it. She has never shown anyone what these additional sentences say.