Behind one concrete wall on a Chinese river sits enough water to fill a small sea.
Ten trillion gallons of it, lifted and held about 575 feet above sea level.
The line that travels is that this moved the planet, and the line is true.
What almost never travels with it is the part that matters, which is that nobody measured it.
The four fifths of an inch is arithmetic, and the thing that got counted was happening in the water below the wall.
Why holding water still changes a spin
The physics is the one everybody already knows from a skating rink.
A spinning body with its mass tucked close to the axis turns faster, and the same body with its mass pushed outward turns slower, which is why a skater pulling in her arms accelerates without pushing on anything.
A reservoir does the outward version. Water that was distributed along a river valley gets gathered into one place and lifted, which moves mass away from the axis and raises what physicists call the moment of inertia.
More inertia, slower rotation, a fractionally longer day. Longer, not shorter, and the viral version of this story routinely gets that backward.
The pole follows the same logic. Pile mass off to one side of a spinning planet and the axis shifts a little relative to the crust underneath it, which is what the polar shift figure describes.
The number, and the numbers it sits between
The published result is small and specific. Filling the reservoir lengthens every day by six hundredths of a microsecond and moves the pole roughly four fifths of an inch.
Set that against what geology does casually. The great Sumatran earthquake rearranged enough mass to shorten the day by 2.68 microseconds, roughly 45 times as much as the dam does, in the opposite direction, over a few minutes.
The Japanese earthquake of the following decade took another 1.8 microseconds off the day, again in minutes rather than years. So the wall is real and it also sits near the bottom of the list of things that move planets, well below what a single large fault does in an afternoon.
The distinction the headlines drop
Here is the part that changes how the story should be read. The figure comes from a calculation published by a government geophysicist in 2005, and it is a model result.
Nobody observed the pole move four fifths of an inch and traced it back to a reservoir.
The exercise was a demonstration of physics, using a conveniently enormous example that happened to be under construction.
The arithmetic is not in dispute. Move enough mass far enough and the axis obliges, and the same equations underpin far larger conclusions about melting ice and rising oceans.
But calculated and observed are different words, and a story that quietly swaps them has stopped being about physics.
The ledger this is one line in
A study of 6,862 dams built between 1835 and 2011 found the pole had wandered about 3.7 feet in total across that period. The path is the interesting part, because it did not travel in one direction.
Until the middle of the last century, reservoirs going up in North America and Europe pulled the north pole roughly eight inches one way. After that, dam building across Asia and East Africa dragged it about 22 inches the other.
The same impounded water took roughly eight tenths of an inch off global sea level, which is very nearly the figure the single reservoir gets credited with for the pole.
The researcher who led the work said flatly that nobody is dropping into an ice age over a shift of roughly three feet.
One dam is a single line in that ledger, and the ledger is a long one. Nobody has finished adding it up.
The number that somebody actually counted
Meanwhile, down in the river itself, an effect was measured rather than modeled, and it was measured by people standing in boats.
The Chinese sturgeon comes in from the sea in early summer, then spends up to 18 months in the middle reaches while its gonads mature, spawning in the autumn of the following year. Its spawning grounds lay far upstream.
A different dam closed that route two decades before the famous one existed, leaving a single small site in the tailwater below it. The reservoir upstream then altered the flow and the temperature that tell the fish when to breed.
Adults numbered around 32,260 before that first wall went up. By 2010 the count was 6,000, and by 2015 it stood at 2,569 animals.
Natural spawning failed outright in several years after 2013, and the species now leans on hatchery releases into a shortened stretch of river.
Fish blocked by concrete is not a new story, as a century of it on a Michigan river shows.
The pole shift is a sum somebody did. The sturgeon is a number somebody went out and counted.
Read the whole thing?
Get the week's signal, not the noise
Our sharpest reporting on energy, climate and nature — free, once a week.
