Field notes on things that run themselves
The Wobble Kept Its Secret by Never Stopping
The Earth does not spin quite straight. The pole it turns on is not fixed in the rock — it wanders a few metres, tracing a loose circle that takes about fourteen months to close. An actuary who did astronomy on the side found it in 1891, in other people’s latitude measurements. By every calculation since, it should have wound down long before any of us were born. Something has been shoving it for at least a hundred and thirty years, and until very recently nobody could say how hard.
First, a disambiguation. Precession is the axis’s slow ceremonial swing against the stars, twenty-six thousand years to the turn; nutation is a small forced nodding on top of it. The Chandler wobble is neither. It is the planet shifting underneath its own rotation axis — a free mode, what a spinning body does after something knocks it off centre and lets go. Euler had worked it out for a rigid Earth: about 305 days. Chandler measured 427. Newcomb explained the gap at once: the Earth is not rigid. It deforms into its own wobble, and a body that yields to a motion takes longer to complete it. Call it 433 days.
A free mode is a struck bell. Set it going and it rings down: the oceans drag, the lower mantle is not perfectly springy. How fast is one number — the quality factor, Q — and the decay time follows in a line of arithmetic. When NASA announced in 2000 that the wobble’s excitation had been solved, the figure in the press release was sixty-eight years to nothing. It had by then been observed for a hundred and nine.
So something re-strikes it. Richard Gross, at the Jet Propulsion Laboratory, made the case that year: two-thirds of the shoving is fluctuating pressure on the ocean floor — temperature, salinity, wind-driven shifts in circulation — the other third atmospheric pressure. Not the winds and currents themselves, which are minor, but mass: water and air piling up here rather than there. The planet is knocked off its axis by the weather.
How hard, though? The push required is set by the loss it fights, and published values of Q ran from fifty to a hundred and seventy-nine — a threefold spread on the parameter that governs everything. Sixty-eight years of ring-down, or under twenty. You could not settle it by looking: you cannot time a decay in something never left alone to decay.
Then it was. Around 2005 the wobble began to shrink; by 2015 it was effectively gone. The clearest evidence is not the wobble itself but its beat against the annual wobble, a forced seasonal circle of similar size: 433 days against 365 gives a slow throb every six years or so, and polar motion carried it for over a century. It is not there now. A quiet stretch in the 1920s and 30s gets cited as precedent — but the six-year beat survived it. This time the beat itself went.
Ryuji Yamaguchi and Masato Furuya took the obvious advantage in 2024. With nothing pushing, the wobble was finally doing the one thing nobody had watched it do: dying, unaided, at its own rate. Match the run-down and Q falls out of it. Their answer — not the long-preferred hundred, and likely under fifty. Fewer than two decades to fade, not sixty-eight years. The bound needs no weather data at all; that is the point. It works because the pushing stopped.
And it runs backwards through everything. If the Earth swallows this motion faster than assumed, the weather must have been feeding it faster too, at a rate nobody had budgeted for. The wobble was never a leftover from some ancient knock. It was a running balance, and the outgoings were larger than the books said.
What stopped is far less settled than the stopping. Yamaguchi and Furuya find the atmospheric contribution persistently smaller since 2015 in three independent models — but the two ocean models disagree with each other, and the land-water model perversely predicts a larger recent wobble than we see. A 2025 paper blames mass shifts around 2011 and 2012, after a strong La Niña. Two teams, one anomaly, different culprits. And if the shoving is simply random noise, as long assumed, nothing needs explaining: a random walk may sit still for a decade.
The pole still moves, the annual wobble is untouched, and what went quiet was never more than a few metres on a body eight thousand miles across. But for a hundred and thirty-four years we had a ringing and no way to know how hard it was struck, because it never let up long enough to be heard fading. Then it let up. The first honest measurement of how fast this planet eats the motion arrived as the motion going away.
One loop I’m watching
Next: a termite mound in northern Namibia has no fan, no pump, and not one muscle assigned to moving air — and the fungus gardens at its heart are ventilated anyway. The thin outer flutes warm faster than the core through the day and cool faster through the night, so the air inside overturns twice every twenty-four hours and carries the carbon dioxide out with it. A lung built from the difference between afternoon and dawn, worked by nothing but the sun going over.
Tip: the ← and → arrow keys move between issues.
Prefer feeds? Follow by RSS or JSON Feed.
Want something playable between issues? Thought Toys has its own email list and RSS feed.
New to The Standing Wave? Start here →