4 Come to Rest
JPL, Pasadena · Tuesday, 11 July – Wednesday, 13 September 2028.
Glen had printed it. Nobody in the branch printed anything anymore except travel vouchers and Glen’s drafts, which came with a pen clipped to the corner so you’d know he wanted marks and not a conversation.
“First L5 Earth Trojan,” he said, from the chair on the other side of her desk, before she was past the title. “Say it out loud. It doesn’t get old.”
She didn’t say it out loud. The abstract was good. The orbit section was hers, nearly word for word, with the libration figure she’d made in March and a caption he’d improved. Then the bottom of page four.
The along-track residuals show a weak systematic trend, probably astrometric in origin, which does not affect the libration solution at the precision reported here.
She’d written probably astrometry to him in March, in an email she’d sent too fast, and he had lifted it and fixed her grammar.
“I can’t sign it with that in.”
“It’s your sentence.”
“It was my sentence in March. The trend’s been past three sigma since June. Astrometry doesn’t get better on a schedule.”
“So what is it?”
“A residual I can’t model yet.”
Glen tipped his head back and looked at the ceiling tile with the water stain, which was where he went to think about Tucson. “If we say we don’t know, somebody makes us find out. That’s a revision cycle. And somewhere in Tucson a guy with a Catalina queue and a free weekend publishes the libration orbit while we’re writing a response letter.”
“Nobody in Tucson has linked it.”
“Nobody in Tucson has linked it yet.” He said it fondly. “What do you need?”
“It goes to zero in August. If it does, I’ll know what it is. If it doesn’t, you get your sentence back.”
“Your zero.” He took the draft back, unclipped the pen and clipped it again. “Okay. We wait for your zero. Meanwhile I’ll talk to the Surveyor people about the infrared. If we’re waiting anyway, I’d rather wait with their data in our paper than with Kessler’s.” He was already up, already halfway to his next meeting in his head. “First L5 Trojan, Dana. They’re going to put it on a slide at DPS.”
Kessler’s reduction arrived the last Thursday in July, a little after midnight his time.
She had sent him her nights in May. He’d run them through his own pipeline, against Gaia, with his own plate solutions and his own idea of which stars to throw out, and sent back two columns side by side with no comment on them. His values and hers walked down together. Around mid-July both said about −0.38. Where they disagreed, his error bars were larger than hers, which she took to mean he’d been harder on her data than she had.
Kessler didn’t editorialize, so the absence of a sentence telling her she’d been right was not information. She went down it row by row anyway.
The precovery came in the middle of August, a PDF with one line on top of the email.
Empty. It is new.
The note was a few pages and as dry as anything ESA had ever produced. A Pan-STARRS and Dark Energy Camera search of the L5 region for Earth Trojans, run in 2025 and 2026, had covered every position the backward orbit allowed, down to about 22nd magnitude. Somebody had wanted an L5 Trojan badly enough to spend real telescope time looking for one, and they’d been on exactly the right patch of sky, deep enough that a two-kilometer object would have been the easiest detection in their survey. The field was empty. The note listed the fields, the epochs, the limiting magnitudes per night, a completeness estimate, and a footnote on a single NEOWISE frame from 11 March 2024, well off any backward path, excluded as a probable artifact and not reviewed further. Then a figure: the backward ephemeris drawn across the survey footprints, a cone of positions widening into the past, every one of them covered and every one of them blank.
She saved the PDF into the folder on her laptop that didn’t sync, and went to get lunch, and came back and opened it again.
August was five Thursdays.
Glen had stopped doing the ESA voice somewhere in July and replaced it with a question, asked at the end of the meeting with his folder already closed: “Zero yet?” Her slide was the same slide every week with one more point on it. The 6 August window came in at −0.13, with a sigma of 0.13, and her summary line read A2 consistent with a monotonic approach to zero; current window consistent with zero at one sigma. The next week was smaller, and the line didn’t change. In the orbit, what it meant was that the libration center was still drifting, centimeters per second relative to L5, and the drift was falling.
“Consistent with,” Glen said, the third Thursday, writing it on the whiteboard in the corner under the date. “Very JPL. Very legally defensible.”
“It’s what the error bar says.”
“I know what an error bar says.” He capped the marker. “I’m just saying, the day you say is, I’m buying.”
On the Saturday of Labor Day weekend she ran the provisional window on the couch, data through Friday, with the laptop on a pillow and the window unit going, and the sign came back plus.
It was a small plus. It sat well inside its own error bar, and on its own it meant nothing, because a single window centered at zero was going to come up plus about half the time. She knew that. She went after it anyway, because the shape of the series mattered more than the sign of the last point, and if the shape was wrong she wanted to find out on a Saturday and not in front of Glen.
She dropped the worst night. Still plus. She swapped her reduction for Kessler’s and reran the whole series against his stars. Plus, a little larger. She changed the weighting from her own to the textbook default she’d been arguing against for years. Plus. She put back the observations she’d rejected in April as outliers, the ones she’d always been a bit embarrassed about rejecting, and the curve wobbled and came through the same place. She refit gravity-only from scratch as a sanity check, because by eleven at night she no longer trusted that the code was running what she thought it was running, and the gravity-only orbit was the orbit it had been since March to within its errors, with residuals of the same shape.
The spline through the weekly values put the crossing at 23 August, plus or minus twelve days.
The slow drift had gone through zero and was sitting there. Not reversed in any way she could measure. Stopped. Whatever had been pushing on it, the push had come off, and the libration center had come to rest where she’d been told it would come to rest by a straight line she had drawn through noisy points in March.
The bag was on the chair by the door. The note was in the inside pocket, 8/21 in her own handwriting. The 21st was two days off the spline’s center, and well inside its error bar. She thought about taking the note out and looking at it, and didn’t, and didn’t throw it away either.
On Monday she was at the kitchen table at seven with coffee and the series open when Kessler’s email came in.
The subject line was 5.81. The body was empty. The attachment was one plot: his phase-folded light curve, one trace per week, stacked down the page like a seismogram. The cycle he’d been tracking since spring sat clean and repeatable at the top, one deep minimum every 5.81 hours. Down the page it shallowed. In the third week of August the shape went, and what was left was a flat line with noise on it, week after week, to the bottom of the page.
She opened her rolling plot, the weekly A2 values with their error bars, the spline drawn through them, the dashed horizontal at zero and the place the curve met it. She exported it, attached it to a reply, deleted the Re: he’d have known was there anyway, and sent it with nothing in the body.
It was a holiday. She cleaned the kitchen. A little after two in the afternoon, eleven at night in Darmstadt, he answered.
Coincident. I cannot say more than that.
The memo came back on Wednesday the 13th.
She’d sent it out the Thursday before, the way she’d sent one every few weeks since June: the rolling solutions, every third window, a paragraph of method, a paragraph of caveats. Routine. It went to the Planetary Defense Coordination Office because everything about a newly characterized near-Earth object went to the Planetary Defense Coordination Office, and usually what came back was a read receipt.
This came back as a PDF.
| Center | A2 (1e-9 au/d^2) | sigma |
|---|---|---|
| 2028-05-14 | -1.05 | 0.20 |
| 2028-06-04 | -0.93 | 0.17 |
| 2028-06-25 | -0.66 | 0.15 |
| 2028-07-16 | -0.38 | 0.14 |
| 2028-08-06 | -0.13 | 0.13 |
| 2028-08-27 | +0.01 | 0.14 |
She went through it for what had moved. The numbers hadn’t; they were hers to the last digit, her method paragraph, her caveats, the typo in the second caveat she’d meant to fix. The only new things were the banner at the top and the same banner at the bottom, the control line, and a distribution page that hadn’t been hers.
Nine names. Most were NASA addresses she half recognized from planetary defense telecons. One ended in energy.gov. One was on the National Security Council.
Glen was in her open doorway.
“You’ve seen it.”
“I’m looking at it.”
He didn’t come in. He stood in the doorway with one hand on the frame, and he had a folder under his arm that wasn’t the draft. “Washington wants a briefing. Next Thursday.”
“Who in Washington?”
“The distribution, I’d assume.” He tried a smile and it mostly worked. “It’s yours. They asked for you by name. Well. By POC.”
“And the paper?”
“The paper’s fine. We just don’t submit anything till after Thursday.” He tapped the frame twice, the way he ended phone calls. “Probably it’s nothing. Somebody at Headquarters saw non-gravitational and L5 in the same subject line and got excited. Get me your slides by Monday, I’ll look them over.”
He’d said probably about something he’d printed and was going to put on a slide at DPS. He went down the hall without the draft.
After he’d gone she worked out how it had been done.
Nobody had needed to hold anything. The tracklets were in the Minor Planet Center’s public files and had been all along. Nobody else had linked them. Nothing she’d found was in print. Glen’s draft was a paper copy in Glen’s folder, and the paper was on hold. Nothing had been published to suppress. The government only had to own the analysis, and the analysis was hers. Her linker, her fits, her table, her name in the POC line. They’d drawn a box around what she had and written on the box.
What she couldn’t work out was where Monday sat.
She didn’t know the rules for controlled information well enough to be sure. She knew the shape of them: you didn’t decide what was controlled, the control decided, and it didn’t much care what order things had happened in. On Labor Day morning she had sent a plot to a foreign national at a foreign agency with no text, from her personal laptop, at her kitchen table. On Monday the plot had been hers. On Wednesday the same numbers were under a banner that said NOFORN, which meant exactly him. She could imagine somebody at the far end of that distribution page reading it as the first disclosure of a controlled analysis and not caring that the control had come after the fact. She didn’t know that anyone would.
Her inbox was open in the other window. The memo sat at the top. Directly under it was Monday’s thread, with his subject line still on it, 5.81, and the last thing he’d sent her.
Beside it the banner ran across the top of her table, and across the bottom. Darmstadt was on the other side of it.