The Iorio Bound

I was thirty-one years old, the first Reyes to finish college, and at two o’clock in the morning I was trying to figure out if I had just thrown away the last six years of my life over a centimeter.

The dome at Table Mountain was always too cold at two a.m. No matter how many space heaters you smuggled in, the concrete sucked the heat out of your shins. I had a stack of ungraded E&M problem sets in my bag, my advisor had reminded me twice this week that I was supposed to be writing my thesis, and all I wanted was to finish the routine Martian ranging reduction so I could go home and sleep.

That was the only reason I ever did routine anything: so it would go away.

Mars ranging is the most boring thing in radio astronomy, which is why it is the most important. Twice a week the Deep Space Network bounces a coded signal off one of the Mars orbiters or landers, times the round trip down to picoseconds, and gives us the Earth to Mars distance with meter precision. Do that for decades and you can watch the orbits breathe. You can see the mass of Phobos jiggle. You can see General Relativity hold up to absurd precision.

I had been reducing the last six months of passes for MRO, and the post fit residuals should have been noise. They were not.

It was a small sinusoid. About a centimeter peak to peak, creeping in slowly against the orbital phase. I stared at it for a full minute, hoping my eyes were crossing. Then I pulled up the previous year’s residuals and overlaid them.

It had been there the whole time. I had just been subtracting it with everything else.

I spent the next three hours trying to kill it.

First, tropospheric delay. I re ran the Niell mapping functions with tighter constraints. The fit got worse. Then solar plasma: Mars had been on the opposite side of the sun last autumn, but the anomaly peaked later. I checked Earth station clocks against the GPS ensemble. I checked transponder turnaround times. I modeled the thermal expansion of the DSN antennas down to the screw. I even went back to the raw baseband and looked for a cycle slip in the phase.

Nothing ate it. Every test I could think of either did nothing, or made the centimeter sit up straighter.

By five a.m. I sat in the dark, laptop glow burning my eyes. The only thing left that could move an orbit this gradually, without tearing it apart, was the spacetime it moved through.

Because if it was what I was starting to think it was, it was not a problem I could solve for my thesis. It was a problem that would solve me right out of a normal life.

Phantom dark energy.

Not the cosmological constant. The constant is w equals minus one, and it pushes gently. Phantom means the equation of state parameter w is less than minus one. The more space expands, the stronger it pushes. Push hard enough for long enough, and in the far future the expansion rate becomes infinite. The big rip. Everything bound by gravity, from galaxy clusters down to atoms, is pulled apart at a finite time.

The thing that keeps phantom from being a free lunch is that it also affects the orbits of planets in the solar system. A smooth, repulsive energy density changes the effective potential. The orbits should expand, ever so slightly, faster than General Relativity alone predicts.

There is a bound for that. The Iorio bound. It comes from decades of laser ranging to the Moon and radar ranging to the inner planets. It constrains how strong phantom can be, which constrains how soon the big rip could happen.

I pulled up the numbers. If my centimeter was real and due to phantom energy with a constant w less than minus one, the time to the rip scaled inversely with how far above the bound it sat.

My hands went cold, and it had nothing to do with the concrete.

The only way to know for sure was to see if the effect matched on both ends.

So I looked the other way.

Two hundred and fifty million kilometers the other way, in the Mars Navigation Operations Room at JPL, it was 11 p.m. Pacific. The room never looked as tired as grad students do. The overheads were harsh, the displays were color coded, and everybody spoke quietly because a comma in the wrong place can cost you propellant.

I had a backdoor into the nav logs I had earned by fixing their clock correlation code two years prior. I wasn’t supposed to browse at two a.m. my time, but nobody was going to look.

I filtered for MRO and Odyssey over the same window as my residuals.

It took me ten minutes to see it.

The OD team had been quietly chasing a tiny, persistent trend in the areocentric range residuals. For months, the observed range to the orbiters had been coming in just slightly shorter than their predicted trajectories allowed. Not enough to trigger an alarm. Just enough that the orbit determination solutions had been nudging the fitted semi major axes down a hair with each new arc, trying to soak it up.

They were seeing the opposite sign of what I was seeing from Earth.

That is the tell.

From Earth, the light time residuals looked like the distance was increasing a touch faster than Newton and Einstein said. From Mars, the spacecraft were arriving a touch early, as if the orbits were shrinking into the Sun relative to the best fit.

Except neither was shrinking. If the underlying space between orbits was stretching due to a repulsive component, the round trip light time grows, and a local observer in orbit sees their trajectory drift against an old ephemeris. Both threads point the same way. You only see it if you stop assuming one of the measurements is broken.

My stomach flipped. I had both threads on the same screen.

At seven a.m. I walked into my advisor’s office without knocking. He had his glasses pushed up onto his forehead, grading the same E&M sets I still had in my bag.

He took one look at the overplot and sighed. “It’s systematics, Reyes. You have a correlated error between stations. Pull the Goldstone passes and try the closure.”

“I did. It stays.”

“Solar plasma then. The Sun is active.”

“Mars is past quadrature. The plasma contribution is modeled to sub millimeter.”

He rubbed the bridge of his nose. “Then it’s the transponder. Or local oscillator drift. You do not go from a centimeter of residual to rewriting cosmology on six months of ranging.”

“I already checked transponder. I checked every station clock. I checked tropo, thermal, iono. It matches the Mars OD drift.” I slid the nav extract across his desk. “They are chasing the same thing from the other side and they do not know why yet.”

He stared at it for a long time. Then he closed his laptop halfway over the plots, gentle, like he was putting a hand over mine. “I know you want this to be real. But you need to stop chasing noise. Finish the thesis you actually have. That is what is going to pay your rent.”

I walked out before I said anything I would regret. He was not cruel. He was trying to save the first gen kid in front of him from becoming the crank that chased one centimeter for ten years.

I went back to the dome by myself.

The Sun was up now, slanting through the slit and striping the concrete. I locked the door. I did not turn the heaters back on.

I reran the full fit with only the relativistic model, then with the minimal phantom extension. I fixed w to just below minus one and let the amplitude float. The chi squared dropped by a factor of seven.

Then I computed the constraint.

The amplitude I was seeing sat below the old Iorio bound. By a hair. But it was not zero.

With that amplitude, and assuming the simplest constant w, the time to the big rip came out to twenty seven years.

I sat there with that number on the screen long enough for the sun to move across the slit and hit the far wall.

Twenty seven years.

Not enough for me to retire. Maybe enough for me to see the first signs hit the edges of the observable universe if I lived a very long life. Definitely enough that nobody sitting in this building right now was going to rearrange their life over it.

I could email the nav lead. I could write up a careful memo. I could spend the next five years fighting referees who would say systematics, who would ask for ten more years of data, who would point to the bound and say it was still consistent with noise.

And the world would keep getting up tomorrow. The TA assignments would go out. The problem sets would still be turned in late. The coffee cart outside JPL would still open at 6:45.

Somebody would have to decide to believe a centimeter.

I saved the plot. I saved the numbers. I archived both data sets with timestamps.

Then I stood up and walked over to the narrow slit that looked east, toward the rising Sun.

Through the gap in the dome I could see Mars, a tiny orange fleck against the pale blue, already pulling away for the next month.

I knew the date.

I was the only person in that building who knew it to the day.

And there was nobody I was going to call about it tonight.