6 Why Do You Ask
JPL, Pasadena · Monday, 2 October – Tuesday, 5 December 2028.
The program came through on the first Monday in October, from a program coordinator in Baltimore who had plainly been told to move fast and not to wonder why.
Your Director’s Discretionary program has been approved for 40 orbits. Please submit Phase II by Friday so we can begin scheduling as the solar constraint allows.
Forty orbits, no panel, no ranking, no year’s wait behind the people studying galaxy clusters. Twelve months proprietary, which Director’s time almost never got and which somebody had asked for on her behalf. Nobody on Earth could pull a frame of it out of the archive until October 2029 unless she handed it to them. She opened the Phase II form. Someone at headquarters had already filled in the cover page from her request, and filled it in the way she would have if she had been being careful: Principal Investigator, D. Okafor, Jet Propulsion Laboratory. Below that, the co-investigator table, one blank row and a plus sign.
In March she would have typed Kessler into that row before she finished reading the email, Darmstadt after it, because the light curve was his and the light curve was half the reason anyone would believe the orbit. Glen would have gone in under him, and Glen would have asked to go above him, and they’d have had a short stupid argument about it at the coffee cart.
She tabbed past the row to the target page, where the trouble was. Hubble had been flying on one gyro since 2024 and wouldn’t point closer than 50° to the Sun. The object sat at 60° and always would, so every visit had to fit in the slice of each orbit where the planner would allow it, with the roll pinned and no slack, and on one gyro it could barely follow anything moving a degree a day; the program ran on a tracking waiver, short exposures stacked on the motion. She laid out the visits two a week and the planning tool turned half of them yellow.
Glen came by at four with his reading glasses pushed up.
“Forty,” he said.
“Forty.”
“You’re PI.”
“It’s my request.”
“Sure.” He looked at the screen, at the cover page, and if he saw the one-row table he didn’t mention that his name wasn’t in it either. “Good. That’s good, Dana.”
The STIS spectrum came down on the seventeenth, two low-resolution gratings stitched at 0.55 µm and divided by a solar analog she’d had them take in the same visit. She had prepared herself for something that would make her think about instrument artifacts for a week, and she got a line instead.
Flat from 0.4 µm to 1.0 µm, tilting a few percent red toward the long end. No 0.9 µm dip, no 1 µm band, none of the pyroxene and olivine bands that every stony asteroid in the catalog showed somewhere. She put Psyche’s spectrum over it from the literature, scaled at 0.55, and the two curves sat on each other inside the error bars, except at 0.9 µm, where Psyche had its faint dip and this had nothing. An M-type. Iron and nickel, give or take whatever hid under a featureless spectrum, which was a lot, and which every iron-meteorite person in the field would remind her of at length.
Glen had it printed and on his door by lunch.
“It’s a little Psyche after all,” he said, leaning in her doorway with his coffee. “I said it in April. I’ll take April.”
“You can have April.”
“Metal at L5. That’s a paper by itself, never mind the rest.” He was already doing it in his head, she could see, the short version, the one that might get out from under the hold if it had nothing in it about heat. “Iron’s dull. Iron’s allowed.”
He was pleased with it through Wednesday. Thursday morning the printout was gone from his door, and when she passed he was at his desk with the spring light curves open, Kessler’s, not hers, and he closed the window when he saw her. He didn’t say anything about it until Friday, in the hall, when he said only, “Psyche doesn’t stop,” and kept walking.
The routine settled in through the back half of October: a visit, the frames into the pipeline overnight, photometry in the morning, astrometry into the orbit fit by noon. Most days nothing in it was worth a sentence.
The glints were worth one. Kessler had seen them in March at a rate of one every few hours of tracking. Hubble saw one most visits now. She had nine visits by the end of the month and seven of them had a spike, some of them two, a few of them close to half a magnitude for a single exposure. She folded them on 5.81 hours because she had to and got a flat smear of points with no structure in it, which she’d expected, since the cycle had died in August. She folded them on everything else she could think of, half of 5.81, double, the Hubble orbit, the visit start times, and got the same smear.
Glen stopped in her doorway without tapping the frame.
“Got a minute?”
“Come in.”
He didn’t. He stood there and looked at the plot on her second screen, the spikes marked in red against the baseline.
“It’s cracking,” he said.
She waited for the rest of it, the line about thermal stress that would turn it into something they could laugh at over lunch. He didn’t have one.
“Iron that was near melting three years ago,” he said. “Cooling at a percent and a half a week, from the outside in. It’s going to split. Every time it opens a fresh face you get a flat new surface for a while, until it rotates out of the geometry.” He shrugged. “That’s what I’d write.”
“It’s holding attitude.”
“Then it’s holding attitude and cracking.” He looked at the plot a moment more. “I’m just telling you what I’d write.”
Later she heard his door close, which it usually didn’t.
In November she masked the glints out and fit what was left.
The masking was simple. Anything more than three sigma above the running median of its own visit, plus one exposure either side of it, went. What remained was a baseline that she’d been treating since October as the one boring number in the program, and when she fit a line to it the slope came back at −2.0 ± 0.3% per month.
V 17.91 on the first visit in October. V 17.95 on the visit before Thanksgiving.
She went through the comparison stars first. Six field stars in the same WFC3 frames, through the same filter, on the same chip, and every one of them flat to 0.3% across the two months. She checked the geometry next and didn’t need long for it. Its loop around L5 changed the distance and the phase angle a little over the year, so she reduced every point to unit distance and a fixed phase before she fit anything. Nothing left in the viewing could take 2% a month out of it.
The detector’s charge-transfer losses, corrected two ways, moved it to −1.9. The one-gyro jitter showed no trend in the field stars. The mask was hers, and she knew how a mask could make a trend somebody already half believed in, so she let the pipeline pick the spikes at its own thresholds and kept none of her choices. −2.0 ± 0.4.
On Friday Glen asked about it in the hall, on his way to the stairs.
“Comparison stars?”
“Flat.”
“How flat?”
“Three tenths of a percent.”
He didn’t ask what the slope was. He stood there long enough that she thought he would, and then he said he’d see her Monday.
She wrote to Kessler on the Tuesday after Thanksgiving, in the morning, so it would land in his evening and not in the middle of his night.
It was the first thing she’d sent him since Not yet. She didn’t put a greeting on it or a subject line he could forward with any sense.
Does a cooling metal surface darken in the visible? D.
His answer was waiting when she came in on Wednesday, sent at 06:40 his time, which meant he’d read hers before he went to bed and slept on it.
No. Iron in vacuum does not oxidize. There is nothing there to oxidize it. Cooling changes the thermal infrared, not the visible; the visible reflectance of iron-nickel is almost flat with temperature over the range we are talking about. If anything, a cleaner and cooler metal surface reflects slightly better. You would expect it to brighten by a small amount, if it changes at all.
Why do you ask?
M.
He hadn’t asked what she’d done with his solutions, or whether Washington had read his cooling curve, or why the line from Pasadena had gone quiet for ten weeks. He’d answered the physics completely and correctly, put the one question at the bottom where she couldn’t miss it, and signed it with one letter, the way she had.
She wrote the answer that afternoon, the whole of it, and the one sentence she hadn’t let herself say out loud. It took most of an hour. She read it once and saved it.
At five Glen asked if she wanted to get a drink, and she said she had a call. When she closed the laptop the folder in the left margin said Drafts (1).
On the first Tuesday in December she built the deep stack.
Sixty-eight WFC3 frames from every visit since October, every glint exposure thrown out, each frame shifted on the object’s predicted motion and median-combined so that the field stars smeared into faint dashes and the object stood up alone in the middle. It took the cluster most of the morning. What came back was a point. Sharp to the edge of the detector’s resolution, the background flat and gray out to the corners, no coma, no fan, no tail along the antisolar direction or the orbit or anywhere else, down to a surface brightness she’d have needed the full forty orbits again to beat. She subtracted a model star from it, and what was left in the middle was noise.
So nothing was coming off it as dust or gas. She went to the orbit to check the other way it could be leaving.
She’d been carrying the astrometry since the crossing, every Hubble position and every ground position she could get, with the non-gravitational terms left free. Along-track, the term that had braked it all spring, came out at 0.02 ± 0.06, in the spring’s units, which was zero. Radial was zero. Normal was zero. If it had been throwing material off at any speed in any preferred direction, at the rate the photometry wanted, she’d have seen the orbit flinch the other way inside a week. She ran the number on the back of the printout: a 2 km iron body losing 4% of its cross-section in two months was something like 10¹² kg that wasn’t where it had been in October. None of that mass had left with dust, and none of it had pushed.