today is not exactly twenty-four hours long
What was checkedA day is never exactly twenty-four hours; the measured length of day fluctuates continuously above and below 86,400 seconds.
Sources · Ep. 0003 · science
El Niño makes the day longer — not as a metaphor, as a measurement. The 1982-83 event added about 0.9 milliseconds to Earth's rotation, and the 2015-16 event hit 0.81 milliseconds in January 2016. A patch of warm water in the Pacific is measurably braking the planet.
Compiled during fact-checking, before the script was written. Nothing here was reconstructed afterwards.
Three areas of real exposure. First, two claims carry no retrieved source: C02 (the caesium atomic clock, its 1955 construction and roughly nine billion oscillations per second) and C14 (Earth's mass as six sextillion tons). Both are textbook constants used as colour rather than as load-bearing argument, but neither was individually checked. Second, the recent acceleration of Earth's rotation is genuinely unresolved science: C21 and C22 hedge deliberately, the core-process explanation is a leading suspicion rather than a finding, and the 2029 date for a possible negative leap second is a projection that may not happen at all. Third, C25 rests on a NASA IVS paper consulted at title and abstract level only; VLBI length-of-day measurement is not in doubt, but that specific document was not read in full and the satellite-laser-ranging half of the claim is not sourced from it. Separately, the entire current-conditions passage (C23) is a live forecast and will age: it is accurate as of the 13 August 2026 advisory and must be re-checked if publication slips beyond the 2026-27 winter.
today is not exactly twenty-four hours long
What was checkedA day is never exactly twenty-four hours; the measured length of day fluctuates continuously above and below 86,400 seconds.
counts the vibrations of a cesium atom
What was checkedThe caesium atomic clock, first built in 1955, keeps time by counting a bit over nine billion oscillations of a caesium atom per second and does not drift appreciably.
No source was retrieved for this. The 1955 Essen-Parry caesium standard and the 9,192,631,770 Hz definition of the SI second are textbook metrology, uncontested, and used here as scene-setting rather than as an argumentative step. Recorded as not-verified because it was never individually searched, not because it is doubted.
does not spin at a constant rate
What was checkedEarth's rotation rate is irregular: it speeds up, slows down and wobbles rather than running at a constant rate.
About 1.7 milliseconds added to the length of a day, per century.
What was checkedTidal friction from the Moon lengthens the day by roughly 1.7 milliseconds per century.
a day lasted about twenty-three and a half hours
a day under twenty-two hours long
What was checkedAround 620 million years ago a day lasted approximately 21.9 hours, per tidal sediment records.
Source gives 21.9 plus or minus 0.4 hours. The script rounds to 'under twenty-two hours', which holds across the stated uncertainty range.
There were around four hundred of them in a year.
Arithmetic from C06, not quoted from the source: 8,766 hours per year divided by 21.9 hours gives 400.3 days. Across the source's stated uncertainty of plus or minus 0.4 hours the figure ranges from roughly 393 to 409. The script originally read 'more than four hundred', which held only at the central value; it was changed to 'around four hundred', which holds across the entire uncertainty range.
Bumps and dips of a millisecond or so
What was checkedSuperimposed on the long-term tidal slowing are short-term length-of-day variations on the order of a millisecond, appearing over months, far too fast for tidal braking to explain, and driven by the atmosphere.
The planet gives up a little of its rotation to the wind.
What was checkedBecause total angular momentum is conserved, an increase in atmospheric angular momentum is matched by a slowing of the solid Earth.
the trade winds that normally push warm water westward across the Pacific weaken
What was checkedDuring El Nino the equatorial trade winds weaken or reverse, and a large pool of warm water shifts eastward across the Pacific toward South America.
NOAA confirms the easterly winds weaken or reverse and that warming occurs in the central and eastern tropical Pacific. It does not itself describe a pool of warm water sliding back east; that is a standard but compressed narrative rendering of the same phenomenon. Marked partial because the script's imagery goes slightly beyond the source's wording.
presses against the mountain ranges of the Americas
What was checkedAn east-west pressure dipole across the Pacific during El Nino produces a positive mountain torque against the American cordillera, raising atmospheric angular momentum; friction torque from an enhanced Hadley circulation contributes as well.
the length of a day grew by about nine tenths of a millisecond
What was checkedThe 1982-83 El Nino lengthened the day by approximately 0.9 milliseconds.
it reached 0.81 milliseconds
What was checkedThe 2015-16 El Nino produced a length-of-day excursion reaching 0.81 milliseconds in January 2016.
CORRECTION APPLIED AT FACT-CHECK. The draft originally described this as 'the peak of the last great event'. That is wrong - a strong El Nino followed in 2023-24. The figure was correct; only the framing was changed, to 'the peak of the event that followed'. No comparable published LOD figure was retrieved for the 2023-24 event, which is why 2015-16 remains the most recent one cited.
a rock weighing six sextillion tons
What was checkedEarth has a mass of roughly six sextillion metric tons.
No source retrieved. Earth's mass of about 5.97 x 10^24 kg, i.e. 5.97 x 10^21 metric tons, is a standard constant and the script rounds it up to 'six sextillion'. Used rhetorically, to convey scale. Recorded as not-verified for honesty because it was never individually searched.
Twenty-seven of them since 1972.
What was checkedTwenty-seven leap seconds have been inserted since 1972, the most recent on 31 December 2016.
TIME-SENSITIVE. The count is current as of compilation and would change if a leap second, positive or negative, were announced before publication. None has been scheduled since 2016.
A flaw in the Linux kernel
What was checkedThe leap second inserted on 30 June 2012 triggered a Linux kernel bug that drove processors to full load, taking Reddit offline and disrupting LinkedIn and Mozilla.
dropped offline for around an hour
What was checkedThe Amadeus Altea reservation system used by Qantas and Virgin Australia went offline for about an hour after the 2012 leap second, forcing manual passenger check-in and delaying flights.
voted to abolish the leap second by 2035
What was checkedIn November 2022 the General Conference on Weights and Measures, meeting in Paris, voted to abolish the leap second by 2035.
TIME-SENSITIVE in the sense that the 2035 target is a forward commitment and the implementing details are still under consultation with the ITU.
1.66 milliseconds faster than twenty-four hours
What was checkedEarth has been rotating faster since 2020, and on 5 July 2024 completed a rotation 1.66 milliseconds short of twenty-four hours, the shortest day since atomic records began.
TIME-SENSITIVE. 'Shortest on record' is a standing record that later years may break.
In 2025 it nearly did it again.
What was checkedIn 2025 Earth came close to matching that record, with 9 July 2025 running roughly 1.3 to 1.6 milliseconds short of twenty-four hours.
Something is shifting deep in the molten core
What was checkedThe cause of the post-2020 acceleration is not settled, with processes in the planet's molten core the leading suspicion.
Deliberately hedged in the narration by the adjacent line 'Nobody fully agrees on why'. The script names the core as a suspicion, not a finding, and does not assert a mechanism.
A negative leap second, possibly around 2029.
TIME-SENSITIVE and probabilistic. Reported estimates put the chance at roughly 40 percent before 2035, and the 2029 date is a projection contingent on the current spin trend continuing. The script carries this with 'may' and 'possibly' and asserts no date as fixed.
a greater than ninety percent chance of a very strong event
What was checkedAs of August 2026 NOAA has declared an El Nino Advisory, with a greater than ninety percent chance of a very strong event during the 2026-27 Northern Hemisphere fall and winter, and a 69 percent chance of exceeding +2.5C during October-December 2026.
TIME-SENSITIVE, and the most perishable claim in the video. Taken from the CPC ENSO Diagnostic Discussion dated 13 August 2026, which also reported a July Nino-3.4 anomaly of +1.4C. The script deliberately says 'as of this month' rather than naming the date, so the narration does not contradict itself on re-publication, but the underlying forecast must be re-checked if release slips past the 2026-27 winter.
El Nino does not cause leap seconds.
What was checkedEl Nino does not cause leap seconds; the long-term divergence between atomic time and Earth rotation is driven by lunar tidal braking and core processes, while El Nino is a seasonal wobble that fades with the warm water.
Assembled rather than quoted. It follows from the timescale separation across the sourced premises: tidal braking runs at 1.7 ms per century (S03), ENSO length-of-day excursions are sub-millisecond and decay within months (S01, S02), and the current leap-second question is driven by core-related acceleration (S09). Stated explicitly in the narration as a guard against the video's own framing being over-read.
with radio telescopes staring at quasars
What was checkedLength of day is monitored using atomic clocks, satellite laser ranging, and Very Long Baseline Interferometry observations of distant quasars.
The NASA IVS paper establishes that VLBI is used to measure length of day, and specifically in an El Nino context. It was consulted at title and abstract level, not read in full, and the satellite-laser-ranging half of the claim is not sourced from it. VLBI and SLR as standard Earth-orientation techniques are not in dispute, but this entry is weaker than the geophysics claims around it.
Right now, today is not exactly twenty-four hours long.
It never is. Not once in the entire history of our species has a day been exactly twenty-four hours. And the amount today is off by is changing this year, because of the temperature of a patch of water in the Pacific Ocean.
That is not a metaphor. That is a measurement.
Here is what almost nobody realizes about time. Humanity does not keep one clock. We keep two. And they do not agree with each other.
The first clock is the planet. One rotation, one day. That is how every civilization before ours told time. By the shadow, by the sky, by the sun returning to where it started. It is the oldest measurement we have, and for thousands of years it was the only one that mattered.
The second clock is the atom. In 1955, we built a machine that counts the vibrations of a cesium atom. A bit over nine billion oscillations, every single second, without drifting. It was, by an enormous margin, the most precise instrument ever built.
And the moment we switched it on, we learned something uncomfortable.
The Earth is a terrible clock.
For all of history, we assumed the planet's spin was the fixed thing, and everything else got measured against it. Then we built something steadier than the planet, and the roles quietly reversed. Suddenly we could watch the Earth's rotation the way a doctor watches a heartbeat. And it turns out our planet does not spin at a constant rate. It speeds up. It slows down. It wobbles.
Some of that we already understood.
The Moon has been stealing our rotation for four and a half billion years. Its gravity pulls the oceans into a bulge, that bulge drags against the spinning Earth beneath it, and the friction acts like a brake. An extraordinarily gentle brake. About 1.7 milliseconds added to the length of a day, per century.
But run that backwards far enough and the numbers get strange. When Tyrannosaurus was walking around in the late Cretaceous, roughly seventy million years ago, a day lasted about twenty-three and a half hours. Go back six hundred and twenty million years, and the geological record points to a day under twenty-two hours long. There were around four hundred of them in a year.
The dinosaurs lived on a faster planet. We live on a slower one. And the slowing has not stopped.
So. A long, smooth, predictable deceleration. Case closed.
Except it is not.
Because when scientists laid the atomic clock against the Earth's rotation and looked closely, that long slow curve was not smooth at all. Riding on top of it was a jitter. Bumps and dips of a millisecond or so, appearing and vanishing over a matter of months. Far too fast for the Moon to explain. Something else was pushing on the planet. Something operating on the timescale of a season.
It took a while to find, partly because nobody expected the answer to be so light.
It was the air.
Here is the piece of physics that makes all of this work. In a closed system, angular momentum, the total amount of spin, cannot be created or destroyed. It can only be moved around.
And the Earth's atmosphere is not a separate thing sitting on top of the planet. It is part of the system, locked to us by gravity and friction. So if the atmosphere speeds up, that spin has to come from somewhere. It comes from the ground underneath it. The planet gives up a little of its rotation to the wind.
Picture a playground merry-go-round with someone standing on the edge. If they start running in the direction it is already turning, the platform under their feet slows down. Nobody added anything from outside. The spin simply moved from one part of the system to another.
The atmosphere is the runner. The planet is the platform. And the thing that decides how fast the runner runs is the ocean.
Which brings us to El Nino.
Every few years, the trade winds that normally push warm water westward across the Pacific weaken, and an enormous pool of warm water slides back east, toward South America. That is the entire event, in one sentence. A rearrangement of where the warm water sits.
But warm water dumps heat into the air above it, and that reshapes the pressure pattern across the whole Pacific basin. High on one side, low on the other. That pressure difference presses against the mountain ranges of the Americas and shoves the atmosphere a little faster. The subtropical jet streams strengthen. The air, globally, picks up spin.
And the planet underneath pays for it.
During the monster El Niño of 1982 and 83, the length of a day grew by about nine tenths of a millisecond. In January 2016, at the peak of the event that followed, it reached 0.81 milliseconds. Those numbers sound trivial, and in one sense they are. But sit with what they actually describe.
A patch of warm seawater in the Pacific reached up, took hold of the atmosphere, and applied a brake to a rock weighing six sextillion tons.
Now, the obvious objection. Who cares about a millisecond?
You should. A little.
Because we built a civilization on top of that millisecond.
When atomic time and Earth time drift too far apart, we have historically fixed it by force, inserting an extra second into the calendar. The leap second. Twenty-seven of them since 1972. And every one of those insertions is a moment where the world's computers are informed that this particular minute contains sixty-one seconds.
On June 30th, 2012, we added one. A flaw in the Linux kernel meant a large slice of the internet's servers did not handle the extra second gracefully. Instead, they locked their processors at full load. Reddit went down. LinkedIn and Mozilla stumbled. And the reservation system behind Qantas and Virgin Australia dropped offline for around an hour, forcing airline staff to check passengers in by hand.
Flights were delayed because the Earth was not spinning on schedule.
Humanity's response to this was revealing. In November 2022, the world's timekeeping authorities met in Paris and voted to abolish the leap second by 2035.
Read that slowly. Faced with a mismatch between our clocks and our planet, we did not adjust the clocks. We voted to stop listening to the planet.
And this is where the story turns, because the planet then did something nobody predicted.
Since 2020, the Earth has been speeding up. Not slowing. Speeding up. On July 5th, 2024, our planet completed a rotation 1.66 milliseconds faster than twenty-four hours. The shortest day since atomic records began. In 2025 it nearly did it again. Nobody fully agrees on why. Something is shifting deep in the molten core, and the models we trusted did not see it coming.
Which means that for the first time in history, we may have to remove a second instead of adding one. A negative leap second, possibly around 2029. An event no software on this planet has ever experienced.
And El Niño sits inside all of this as the counterweight. While the core accelerates the planet, a warm Pacific reaches up through the atmosphere and pushes back the other way. Right now, as of this month, forecasters have declared an El Niño advisory, with a greater than ninety percent chance of a very strong event through the coming winter, and roughly a two in three chance it becomes historically large.
The brake is being applied, hard, to a planet that has been accelerating for six years.
Now I want to be honest with you about something, because it matters.
El Niño does not cause leap seconds. The long-term drift between our two clocks is driven by the Moon and by the planet's core. El Niño is a seasonal wobble riding on top of that, and it fades when the warm water does. Anyone telling you the Pacific Ocean is rewriting the calendar is overselling it.
But that is not why this matters.
It matters because of what it proves.
The length of a day is arguably the most carefully measured quantity in human history. We track it with atomic clocks, with lasers bounced off satellites, with radio telescopes staring at quasars billions of light years away. It is our most rigorous number.
And a warm patch of seawater shows up in the data anyway.
That is the real finding. Not the millisecond. The coupling. The ocean is not a separate compartment of the world sitting quietly underneath the weather. It is wired into everything. Including the rotation of the planet itself.
Which reframes every other thing El Niño touches. The coffee harvest that fails. The monsoon that arrives three weeks late. The fishery off the coast of Peru that empties out. The winter that shows up feeling like spring. Those stopped being a list of separate misfortunes the moment you understood that the same ocean is measurably tugging on the length of your day.
It was never a weather event. It is the clearest evidence we have that this planet is a single connected machine, and that we are living inside the mechanism.
We built our entire civilization on the assumption that time is the one thing that holds still. The fixed background. The thing every other thing gets measured against. Schedules, contracts, satellites, markets, the moment your alarm goes off tomorrow morning.
And it turns out the background moves. Slightly, constantly, and partly because of the ocean.
So when we voted to stop inserting leap seconds, we did not fix the drift. We simply decided to stop looking at it.
Which leaves a question worth sitting with. If the length of a day, the most scrutinized number we have, is quietly being edited by seawater, what else are we treating as fixed?
How many other things are we measuring against, right now, that are also drifting?
And how would we ever find out, if we have stopped checking?