The anomaly appeared on the morning of June third, in the third column of the third data file, at a frequency that Dr. Rowan Kier would later note was either a coincidence or a warning, depending on which interpretation you chose to believe. The numbers on the screen did not make sense. The methane flux readings from Borehole Seven, located at the northern edge of the permafrost monitoring grid, had jumped by a factor of four in a single twelve-hour sampling window. Rowan had been measuring permafrost methane at the Toolik Lake Field Station for seven years, through three complete freeze-thaw cycles, and he had never seen a spike of this magnitude outside of a controlled burn experiment. He scrolled back through the raw telemetry. The instruments were functioning within normal parameters. The weather station reported nothing unusual. The satellite feed from NOAA's Sentinel constellation showed the expected thermal profile for early June: surface temperatures just above freezing, snowmelt accelerating, the Arctic tundra beginning its brief and violent summer respiration.

Interpretation One was obvious and terrifying. The permafrost was destabilizing faster than any model had predicted. The methane hydrates locked in the frozen soil for thirty thousand years were releasing at a rate that made the IPCC's worst-case scenario look optimistic. If the readings were real, the feedback loop had already begun, the carbon bomb was already detonating, and every projection Rowan had published in the Journal of Climate Dynamics was not just wrong but catastrophically wrong, wrong in the way a fire alarm is wrong when it fails to ring while the building burns.

Interpretation Two was equally plausible and far more manageable. The instruments were lying. The gas chromatograph in Borehole Seven had a known tendency to drift in high humidity, and the humidity that morning had been ninety-four percent. The soil moisture sensors had been recalibrated only two days earlier by a technician who had been at the station for less than a month. The data transmission had passed through three relay nodes before reaching Rowan's workstation, and each node introduced a potential point of corruption. A single byte flipped in transit, a single voltage fluctuation in the analog-to-digital converter, a single condensation droplet on a circuit board, and the numbers on the screen would transform from a climate catastrophe into an equipment failure.

Both interpretations were true. Rowan knew this in the way a quantum physicist knows that a particle can be in two places at once until someone measures it. The data was both a warning and a mistake. The permafrost was both collapsing and holding steady. The world was both ending and continuing. And he, Rowan Kier, forty-one years old, divorced, living in a prefabricated metal box four hundred miles north of Fairbanks with eight other scientists and a rotating cast of graduate students, was the observer who would have to choose which reality to live in.

He did not choose. He could not choose. He sat at his workstation in the common room of the Toolik Lake Field Station, the fluorescent lights humming at exactly sixty hertz like the heartbeat of the building, and he stared at the numbers on the screen. Outside the window, the Arctic summer was doing what Arctic summers do: the sun was circling the sky without setting, the caribou were migrating across the tundra in long brown lines, the mosquitoes were rising from the meltwater in clouds so dense they blurred the horizon. It was June third, 2024. The world was the warmest it had been in one hundred twenty-five thousand years. And the numbers on Rowan's screen said that it was about to get much, much warmer.

Dr. Mira Voss had arrived at the station in April, three months into her postdoctoral fellowship, with a duffel bag full of cold-weather gear and a reputation that preceded her like a weather front. She had published six papers in Nature Climate Change before the age of thirty. She had developed the algorithm that the NOAA Arctic Program used to calibrate its methane flux models. She was, by every measure Rowan could find, the most qualified person ever to set foot in the Toolik Lake Field Station. She was also, by every instinct Rowan could not suppress, the most likely explanation for why the numbers did not add up.

The evidence for Mira's sabotage was as compelling as the evidence for the methane spike. She had access to every instrument in the monitoring grid. She had written the calibration protocols. She knew exactly which variables would need to be altered to produce a fake signal that would pass all automated validation checks. And she had a motive, a motive that Rowan had pieced together from late-night conversations and casual remarks and the careful way she looked at the permafrost cores as if she were looking at a prophecy. Mira believed that the world needed to be scared. She had said as much at a conference in Reykjavik the previous year, a conference Rowan had attended and remembered because Mira's keynote had been the only one that did not end with the usual qualifiers and caveats. The models are too conservative, she had said. The policymakers need to see the worst case. If we keep hedging, we keep giving them permission to do nothing.

It was not hard to imagine Mira adjusting a calibration constant or introducing a systematic bias or simply selecting the most alarming subset of data from a noisy signal. It was not hard to imagine her doing it for the right reasons, for the reasons Rowan himself had sometimes entertained in the dark hours of the Arctic winter when the sun did not rise for sixty-seven days and the only thing that kept him going was the conviction that his work mattered. But imagining was not knowing. Suspecting was not proving. And every time Rowan looked at Mira across the common room table, every time he watched her calibrate the gas chromatograph with the steady hands of a surgeon, every time he asked her a question about the data and she answered with the calm precision of someone who had nothing to hide, the superposition held. She was both a saboteur and a savior. The data was both falsified and real. The world was both burning and not burning.

The other interpretation of Mira was simpler and more difficult to hold. Mira was exactly what she appeared to be: a brilliant scientist doing her job correctly, finding a signal that was real and terrifying, and trying to get her colleagues to take it seriously. The late nights at the workstation were not evidence of data manipulation but of diligence. The careful way she looked at the permafrost cores was not the look of someone looking at a prophecy but the look of someone who understood, more deeply than anyone else at the station, what the cores were saying. She was not manufacturing a crisis. She was discovering one.

But if that were true, Rowan thought, then why did she delete the raw telemetry files from Borehole Seven on the night of June second, twelve hours before the anomaly appeared? Why did the server logs show her account accessing the calibration interface at 3:17 AM, a time when no legitimate calibration would ever be performed? Why had she refused to share her field notes from the sampling trip in May, citing data integrity protocols that Rowan had never heard of and could not find in any NOAA manual?

And if the first interpretation were true, the interpretation in which Mira was sabotaging the data to create a false alarm, then why had she been the one to flag the anomaly in the first place? Why had she walked into Rowan's office at 7:15 AM, two hours before his shift started, with a printout of the methane readings and a face that looked genuinely afraid? Why had she spent the next three days rerunning every diagnostic, checking every sensor, recalibrating every instrument, doing everything a scientist would do if she believed the data and wanted to make absolutely certain it was correct?

The questions looped back on themselves like the caribou trails on the tundra, returning always to the same starting point. Rowan sat at his workstation and watched the numbers and thought about Mira and thought about the permafrost and thought about the thirty thousand years of methane locked in the frozen soil beneath his feet. He was standing on a bomb. He was standing on solid ground. Both statements were true.

The station's satellite link went down on June seventh. The official explanation was solar flare activity, a coronal mass ejection that had been predicted by NOAA's Space Weather Prediction Center two days earlier. The unofficial explanation, the explanation that Rowan wrote in his personal journal and then deleted and then wrote again, was that someone had disconnected the satellite dish. He had walked past the equipment shed at midnight on June sixth and he had seen a figure standing in the shadow of the dish, a figure in a red parka that could have been Mira or could have been any of the eight other scientists at the station. When he walked back ten minutes later, the figure was gone. When he checked the server logs the next morning, the satellite connection had been interrupted at 12:08 AM and restored at 12:23 AM, the exact window during which the figure had been standing by the dish.

He asked Mira about it over breakfast. She looked at him with an expression that was either innocence or the performance of innocence, and she said she had been in her bunk at midnight, asleep, and if the satellite had gone down it was probably the solar flare, and if he was worried about data integrity he should check the local backups. He checked the local backups. They were intact. Every data point from Borehole Seven was stored on the station's redundant drives, timestamped and checksummed and verified. The methane spike was still there. The anomaly was still real. Or the anomaly was still fabricated, so perfectly fabricated that it had survived every verification protocol on the station.

The isolation was making things worse. Rowan knew this. He had studied the psychology of remote research stations, the way the constant daylight of Arctic summer disrupted circadian rhythms, the way the absence of normal social cues amplified suspicion and paranoia. He knew that the Toolik Lake Field Station had a documented history of interpersonal conflicts, that three researchers had been medically evacuated for psychological reasons in the past decade, that the combination of extreme cold and extreme isolation and extreme intellectual pressure was a recipe for the kind of thinking that could not be trusted. He knew all of this and he could not stop thinking it and the knowing did not make the thinking stop.

On June tenth, he walked out onto the tundra alone. He walked past the monitoring grid, past the boreholes with their sensor arrays and their data cables and their silent testimony to the changing climate. He walked until he reached the edge of the permafrost, where the frozen soil gave way to the meltwater ponds that had been growing larger every summer for the past thirty years. He stood at the edge of a pond that had not existed when he first came to Toolik Lake in 2017. The pond was a hundred meters across and three meters deep and it was releasing methane in visible bubbles, bubbles that rose to the surface and popped and released their contents into the atmosphere. He could see them with his naked eye. He could count them if he wanted to. He could measure them, quantify them, reduce them to a number on a screen, and that number would confirm Interpretation One and refute Interpretation Two and end the superposition forever.

He did not count them. He did not measure them. He stood at the edge of the pond and watched the bubbles rise and burst and he thought about Mira and the satellite dish and the deleted telemetry files. He thought about the figure in the red parka and the access logs at 3:17 AM. He thought about the calibration constants and the systematic bias and the possibility that everything he was seeing was a construction, a carefully crafted illusion designed to make him believe something that was not true. He thought about the possibility that everything he was seeing was real, that the permafrost was collapsing and the methane was releasing and the world was ending and he was the only person in a position to sound the alarm.

Both possibilities were equally terrifying. Both possibilities were equally supported by the available evidence. Both possibilities demanded a response that would be catastrophic if the other possibility turned out to be correct. If he reported the data to NOAA and it turned out to be a fabrication, he would destroy his career and Mira's career and the credibility of the entire Arctic monitoring program. If he dismissed the data and it turned out to be real, he would be complicit in the greatest failure of scientific communication in human history.

He walked back to the station. The sun was still in the sky, where it had been for the past seventeen hours and where it would remain for the next seventeen hours. The caribou were still migrating. The mosquitoes were still rising. The permafrost was still frozen or still melting or both, depending on which interpretation you chose to believe.

Mira was waiting for him at the door of the station. She was holding a printout, the same printout she had shown him on June third, the same numbers, the same anomaly, the same impossible methane spike. Her face was the face of a person who was telling the truth or the face of a person who was lying. Rowan could not tell the difference. He had spent his entire career learning to distinguish signal from noise, and he could not distinguish truth from deception in the face of the person standing in front of him.

I ran the diagnostics again, she said. Everything checks out. We need to report this.

Or, he thought, we need to investigate further. Or, he thought, we need to wait for independent verification. Or, he thought, you need to tell me what you did to the calibration interface at 3:17 AM on June second.

He said none of these things. He took the printout from her hand and he looked at the numbers and he nodded. The numbers were either the most important climate data ever recorded north of the Arctic Circle or the most elaborate fraud ever perpetrated at the Toolik Lake Field Station. The superposition remained. The truth existed in both states simultaneously, waiting for a measurement that Rowan did not know how to make.

He walked past Mira into the station. He sat at his workstation. He opened the draft of his report to NOAA, the report he had been writing and deleting and rewriting for seven days. The cursor blinked on the empty page. Outside, the sun circled the sky without setting. The permafrost held its thirty thousand years of methane in its frozen grasp, or it released them into the atmosphere, or it did both at once. And Rowan Kier, alone at his desk in the strangest and most beautiful place he had ever known, made the only decision that both interpretations would allow.

He closed the report without saving it. He opened a new document. He wrote a single sentence and left it on the screen for Mira to find, or for himself to find, or for no one to find.

I do not know what is true.

He walked out of the station and onto the tundra and stood in the light of the sun that would not set. The world was ending. The world was continuing. Both were true. Both were false. Both were waiting for someone to measure them, and Rowan Kier, climate scientist, forty-one years old, the only person in the world who could resolve the superposition, had decided that some measurements were not his to make.