2 Out of Phase
Izaña, Tenerife, then ESOC, Darmstadt · 9 March – 6 May 2028.
A fortnight before anyone in Pasadena had it, the object was a line in Matthias Kessler’s notebook: K-L5-07, the seventh thing in two years of looking that he had not been able to throw away.
The search was his own. The Planetary Defence Office let him have the Optical Ground Station on nights when the telescope was not needed for laser tests, and tolerated the project the way it tolerated his paper notebooks, because neither cost anything that appeared in a budget. Every Earth Trojan anyone had found sat at L4 and had been found by accident. Kessler had wanted to know whether the trailing side was empty, or only unlooked-at. This was the last run. After March the Optical Ground Station went back to the people who talked to satellites, and his two years would become a short paper with a null result in it, which he had already half written.
K-L5-07 stood in the west after dusk, above the sea of cloud that lay on the north of the island every evening and hid La Orotava under it. The 1 m dome turned toward it. The frames came in one after another, a faint dot against a field of stars that moved together and left it behind.
On the first night it brightened. About a tenth of a magnitude, for a few minutes, and then it was itself again. Kessler did not like single events, so he waited, and it did it again before the object set. By 19 March he had eleven of them across five nights, each carried through consecutive frames, each the same height and the same few minutes wide.
He checked the airmass, the guiding log, the dome position and the flat field under the object’s track, in that order, because that was the order in which such things had betrayed him before. The comparison stars stayed flat through every event. A satellite would have been a streak. A cosmic ray would have been one frame. These were in the object, or in the light coming from it.
He tried rotation next, because everything that tumbles in sunlight has a rotation and every rotation leaves a period. He ran periods from two hours to two days. None of them put the eleven events in phase. He tried non-principal-axis models, the slow tumblers the Pravec group fits, and they did no better; they did worse, because they had more freedom and still could not use it.
The orbit was no better. A gravity-only fit drifted off its own prediction a little further each night. His own frames from a February run, taken before he had known to look, stretched the arc back past the March run, and when he let a non-gravitational term float, the along-track part came out negative at about two sigma. He filed it under his own astrometry, the first place a careful man looked, and did not send the positions to the Minor Planet Center. They were not a result.
Iron would glint. Flat faces on a metal body would throw sunlight back in brief hard flashes, and that explained everything he had, which was the reason he did not trust it.
So he did the arithmetic that the explanation required. A tenth of a magnitude on a body of that brightness, held for a few minutes, from a mirror-like face reflecting the Sun into the dome at Izaña. It came out at a few square meters of flat surface, turned the right way, and more if the surface was anything less than a mirror.
He looked at the number on the page and did not work it out a second time.
Under it he wrote: Not a result. Do not describe it as one.
On Sunday 2 April Rita Albrecht wrote to JPL from her kitchen in Darmstadt, and copied him.
She had heard on Friday, through someone at the Minor Planet Center, that Pasadena had asked Catalina for time at a position near L5. The position was his. Rita had not asked him whether to write. She had asked him for a paragraph on Saturday. He had sent her one. Now he read what she had sent in its place.
It was his paragraph. The epochs, the brightening events, the failed periods, the request for coordination. He was at the end before he found the gap. The sentence that had followed the glints, the one that said how much reflecting surface they required, was not there. The paragraph went from no rotation model accounts for them straight to we would welcome a joint analysis, and the join was clean. She was good at this.
He called her.
“You removed the brightness budget.”
“I removed a sentence that someone would quote.” There was a radio on behind her. “We are asking for their data. We are not asking them to think we are strange.”
He could have put it back in the attachment. The attachment was his.
“No,” he said. “Leave it out. It was premature.”
“Good,” Rita said, and was gone.
The call was on Tuesday 4 April at 16:00 in Darmstadt, which was 07:00 in Pasadena. Glen Farris had a paper cup from a coffee cart and had plainly been up early planning the call.
There were four of them in the frame. Glen in front. A woman beside him who did not look at the camera, whom he introduced as Dana Okafor, who had done the linkage. Two others further back, introduced by first name and nothing else, who did not speak once.
“So our linkage,” Glen said, “and your photometry, if we can put those together, I think there’s something really nice here.”
Kessler wrote our in the margin of the notebook and underlined it once. The tracklets in their linkage had come from surveys Glen’s agency paid for. He did not think that made the object theirs, but he did not think it made it his either, and it was not a question one won at seven in the morning, Pasadena time.
He shared his screen. The lightcurve, five nights, the eleven events marked in red.
Glen leaned in. “Oh, that’s lovely. That’s specular. You’ve got flat facets, you’ve got a metal surface. That’s an M-type.” He was already pleased with it. “That’s a little Psyche, sitting at L5.”
“That is one explanation,” Kessler said. “I have not found a period that supports it.”
“Facets on a slow tumbler, you wouldn’t necessarily —”
“I tried tumbling. The models are crude. They are not that crude.” He took the screen back. “I would like to ask for the Surveyor infrared at these epochs. The thermal data would constrain the albedo, and the albedo would tell us whether it is metal.”
Nobody answered. Glen looked to his left, off camera, at one of the people who did not speak. The other one wrote something down. Kessler watched three people in California work out whose permission it was, whether any of them had it, and whether anyone at the Surveyor project would be awake.
“We’d need to run that past the Surveyor team,” Glen said. “Data rights. I don’t see a problem. I just can’t say yes on this call.”
“I understand.”
“Did you fit non-gravs?”
It was the woman beside him. She was looking into the camera now. Kessler decided he could tell her the truth, if he told all of it.
“Yes. The along-track term is negative. I believe it is my astrometry.”
“Ours was the same.”
Glen glanced at her. She was looking at her laptop again.
He moved on to the Minor Planet Center, what it would release before a joint circular and what it would not. Kessler answered him in complete sentences while he thought about what had just happened. She had not given him a value. He had not given her one. But her tracklets came from Surveyor and Catalina and whatever else her linker had swept up. His orbit rested on ten nights of his own frames at Izaña, stretched past them by a few orphaned detections he had pulled out of the Isolated Tracklet File, which was the only reason a two-sigma along-track term on it was worth saying aloud. If both orbits braked, then two sets of astrometry with almost nothing in common were wrong in the same direction.
He did not hold the silence against her. He had just told her it was his astrometry. He still thought it might be. He had not been sure of it since the second week of March.
The Surveyor infrared arrived on 18 April, about a thousand cutouts, most of them commissioning frames that had never gone to the MPC and a note from Glen with exclamation marks in it: the Surveyor team had put the object on their opportunistic list, four visits a day when the survey pattern allowed, and nobody had yet asked them why. Kessler put open circles beside the early frames flagged PROVISIONAL CALIBRATION and fitted everything else.
The long band, NC2, 6 to 10 µm, did what it should. Run through the near-Earth asteroid thermal model with any reasonable albedo, it gave a body about two kilometers across, which agreed with the optical. The short band, NC1, 4 to 5.2 µm, was several times too bright. He made the surface darker, then smoother, then let the beaming parameter go lower than anyone had published, and every sunlit model he tried fell short in the same direction. The color temperature between the two bands came out about 200 K above the hottest of them. Sunlight at that distance could not do it. Something on the object was warmer than the Sun could make it.
He had written a word on the pad under the fit. He wrote calibration? over it, pressing hard, until the first word could not be read, and turned the pad face down on the desk.
On 27 April he stopped fitting temperatures and looked at the infrared as a time series.
He had not expected anything. Two kilometers of rock seen from L1 at irregular visits was a sparse, ugly sampling, and he ran a Lomb–Scargle on NC1 mostly to be able to say he had. There was one peak. It stood well clear of the noise.
5.81 hours, plus or minus 0.03. About eight percent in NC1, weaker in NC2, the way a varying hot component would show more strongly in the shorter band. He went to his visible photometry from March and looked for the same period. Less than 0.02 magnitudes. A two-kilometer body spinning in sunlight would have shown at least a tenth, from shape alone. This was not spin. The heat was going up and down and the body was not turning to do it.
He folded his eleven glints on the infrared period. The difference in light-time between L1 and Izaña was about 2.5 seconds, which he wrote down and ignored. Nine of the eleven fell in the third of the cycle where the infrared was lowest. He shuffled the glint times against the phase and counted how often chance did as well. About four times in a thousand.
He sat with the fold on the screen and did not run it again. Then he carried the period forward to the first minimum that would fall after dark at Izaña and inside a Surveyor visit. He wrote it in the notebook in his square hand, with the error he had earned and no less:
06.05.28 21:47 UT ± 1.1 h.
He took it to Rita the same afternoon.
“One more night,” he said. “The sixth of May.”
“The station is with the laser people from April.”
“I know.”
She looked at the notebook, not at him. “What happens if it does what you say?”
“Then it is a result,” he said. “And I will have to describe it as one.”
She found him the night. Someone’s terminal test had slipped, and Rita had been on the phone to Tenerife before the slip was official.
On Saturday 6 May Kessler ran the telescope from his office in Darmstadt, with the night assistant at Izaña on an open line. At the observatory it was an hour earlier. The assistant reported the cloud low and flat over the north coast, the seeing good. After that he spoke only to read off the humidity.
K-L5-07 was already in the frames at dusk, faint, low in the west. The comparison stars were steady. He watched the UT clock in the corner of the screen go past 21:40. That was Surveyor’s visit; somewhere above the Earth’s day side a telescope was looking at the same two kilometers. He did not know yet what it saw.
At 21:58 the object brightened. A tenth of a magnitude, carried through consecutive frames, a few minutes, and then it was itself again.
He did not say anything. He checked the airmass, the guiding log, the dome position, the flat field under the track.
“Still clear here,” the assistant said. “It is getting low.”
He took frames until the object set into the cloud. There were no others.
Before midnight the quick-look from Surveyor came through the pipeline Glen’s people had set up. One visit, 21:40 UT. He put the NC1 point on the fold. It sat at the bottom of the curve.
He opened the notebook to the page with the prediction on it and wrote underneath, without an open circle:
21:58 UT. One. Inside the window.