Empirical Earth · aviation, plain version
Do pilots have to compensate for terrestrial curvaturepoint the nose down because the Earth is round?
Below is the real answer, written the way it usually gets written. Then every word that exists to sound impressive is crossed out and replaced with what it means. Not one number changes. Only the vocabulary does.
How to read this. The jargon looks like thisthe big words get crossed out, and the plain version follows in green. Both are correct, and that is the whole joke. If an argument stops working once you remove the big words, the words were doing the work.
01What the claim gets right
Start with the concession, because it is real. Pilots do not push the nose down to chase a curve. There is no curvature knob on any panel. Nobody is taught to correct for the shape of the Earth during level cruisethe long boring part of the flight. All of that is true, and an answer that denies it is not worth reading.
The question is what follows from it. The answer turns out to be: nothing.
02The number everything hangs on
As an aircraft travels over a sphere, the direction of “down” beneath it rotates. The rate is called the transport ratethe how-fast-does-down-tip number, and it is ground speed divided by the radius of the Earthhow fast you are going, divided by how big the ball is.
ω = v ÷ R = 250 m/s ÷ 6,371,000 m = 0.00225° per second
That is about 8 degrees per hour. Take the smallest slice any recorder stores, one eighth of a second, and the rotation inside it is 0.00028 degrees.
In plain words. The plane tips down a tiny bit as it flies, because the ground curves away underneath it. How much? About eight degrees in an entire hour — slower than the hour hand on a clock. Nobody feels that. Nothing on the dashboard notices it either.
What the claim is really picturing
Behind the argument is an image worth taking seriously: a plane flying a rectilinear path through inertial spacea dead straight line, like a laser while the ground curves away below. If that were happening, the effect would be gigantic.
| Flying dead straight for | The ground would drop away by |
|---|
| 100 km | 785 meters |
| 1,000 km | 78 km |
| 5,000 km | 1,728 km — congratulations, you are in space |
So the claim is right that something dramatic would show. The arithmetic is fine. The premise is wrong. A plane does not fly a straight line through space, and no pilot has ever tried to.
03The instruments
Three kinds of attitude referencetiny-plane-picture gadget sit in cockpits today, and the claim usually treats them as one thing.
The vacuum-driven artificial horizonthe old spinny one
A gyroscopea spinning top held upright by a pendulous erection mechanismlittle weights that hang down and keep it straight. Its face has pitch bars every five degrees and a needle about a degree thick. Rugged, reliable, coarse. Nobody has ever asked it to see thousandths of a degree.
The glass primary flight displaythe TV screen one
A smoother ladder drawn on a screen, marked in whole degrees, with the underlying value rounded prior to renderingchopped off before it gets drawn. It looks more precise. For this purpose, it is not.
The air data inertial reference unitthe box that knows which way is up
Underneath both sits the real source: ring-laser gyroscopeslasers chasing each other in circles working out the attitude and pushing it onto the data busthe wire that everything talks on. Genuinely precise hardware. It still cannot show the curve, and the reason has nothing to do with how many bits of resolutiontiny counting steps it has.
| Instrument | Resolutionsmallest wiggle it can notice | Curve alone needs | Sees it? |
|---|
| Old spinny one, painted bars | 5° | 37 minutes to move one bar | No |
| Old spinny one, as a human reads it | ~1° | 7.4 minutes | No |
| TV screen one | ~0.5° | 3.7 minutes | No |
| Black box, most planes | 0.176° per step | 1.3 minutes | No |
| Black box, big Airbus | 0.352° per step | 2.6 minutes | No |
04Yes, the plane really does tip. So why does no dial show it?
It is tempting to say there is nothing to see. That is overstating it, and the true answer is better.
The plane really does turn. Measured against the stars, it turns 8 degrees every hour — and about 81 degrees on a long flight. That is nearly a quarter of the way around. It is not a tiny rounding error, and nobody is hiding it. So why does the dial say nothing?
Because there are two different angles, and the claim mixes them up
| The angle | How big? | Does anything measure it? |
|---|
| How far the plane has turned in space | 8° an hour. 81° on a long trip. | Yes, all the time |
| The nose, compared to “level” | Zero. Always. | There is nothing to measure |
Those are not the same thing. The first one is real and big. The second one is zero because of what “level” means — and “level” turns along with the plane.
Try this at your desk. Stand a pencil upright on an orange, then roll the orange. The pencil turns all the way over compared to your room. That is real, and you can watch it happen. But compared to the orange, the pencil never leaves straight-up. Now ask the pencil to tell you it tipped. It cannot. The pencil does not know about your room. It only knows the orange.
So a sensor with arbitrarily fine quantizationgadget that counts in the tiniest steps you can imagine would still show the same steady nose-up. The problem is not that the gadget is not fussy enough. The gadget is looking at the wrong angle.
And the real turn is measured, on every single flight
Here is where the whole claim flips over. A ring-laser gyroscopethe lasers-chasing-each-other box does not measure turning compared to level. It measures turning compared to the stars — which is where the turn really is. So the plane following the curve shows up in the gyro, plain as day.
| What the laser box sees | How big |
|---|
| The Earth itself spinning | about 15° an hour |
| The plane going round the curve | 8° an hour |
The curve is more than half as big as the Earth’s own spin, sitting right there in the signal. This is not a whisper. It is one of the loudest things the box hears.
And if you switch it off, the plane gets lost. The navigation computer has to subtract that turn to keep its idea of “down” aimed at the middle of the Earth. Skip it for one hour and “down” is wrong by 8 degrees — and the plane no longer knows where it is.
So the real answer is not “there is nothing to see.” The real answer is: the curve is measured by every airliner, on every flight, and the plane could not find its way without it. The only place it does not turn up is the nose-angle dial — and that is because the nose-angle dial was never measuring it.
05“Then show us the pitch data”
The strongest version of the challenge, and it deserves a straight answer.
One. The curve never gets into the number being recorded, for the reason above.
Two. Even if it did, the flight data recorderthe black box cannot hold a number that small. It writes pitch down in steps of 0.176°. The curve moves it 0.00028°. That is a factor of about six hundred — a number nobody can picture, so here it is in things you can hold.
The ruler. If the black box were a ruler marked in millimeters, the curve would be 1.6 micrometers. That is about forty times thinner than one of your hairs. You are not being asked to read between the lines. You are being asked to read something that is not there to see.
The needle. This one is literal. One degree of pitch swings the horizon bar about 1.6 mm across the instrument face. So the curve, in one recorded sample, would move that bar 0.45 micrometers.
Green light has a wavelengthsize, as a wave of about 0.55 micrometers. The needle would move less than one wave of light.
That is not the instrument failing. Not the recorder, not the rulebook, not the manufacturer. It is light. You cannot see a thing smaller than the wave you are looking at it with.
Three, and this settles it. The pitch number is already being shoved around by everything else on the plane:
| What moves the needle | How much | Compared to the curve |
|---|
Center-of-gravity migration from fuel burnthe plane’s balance shifts as it drinks its gas | 1–2° | ~7,000× bigger |
Light turbulencebumpy air | ~0.5° | ~1,800× bigger |
Autopilot altitude-hold oscillationthe robot wobbling a little | ~0.2° | ~700× bigger |
Asking to see the curve in the pitch data is asking to spot a grain of sand underneath a landslide — in a column that was never measuring sand.
06What the flight trackers actually show you
Push the claim and it usually turns into a different one: the plane should be sinking, and the altitude should show it. That is a claim about height, not the nose, and it gets its own answer.
The nose is not in the feed at all
The altitude on your screen comes from ADS-Bthe radio that shouts “HERE I AM” twice a second. It shouts where it is, how high, how fast, which way, and its name. It never shouts the nose angle. Not in any version of the DO-260B standardthe rulebook for the shouting radio. A tracking site cannot show you something the plane never says.
The altitude, since it is a fair question
The height rides in a twelve-bit fieldtwelve little on-off switches, and one of them, the Q bita switch that says which ruler we are using today, sets the step: 25 feet or 100 feet. That is why the number on your screen jumps instead of sliding.
But the fineness of the ruler does not matter. There is nothing to measure. The plane holds one height for hours, so the line on your screen is flat. The flat line is the curve. A flight levelthe height everyone agrees to fly at is an isobaric surfacean invisible floor made of squished air, and that floor is already wrapped around the ball. Riding it is following the curve.
Being honest about it
A flat Earth with a level-flying plane would also give a flat line. This one number does not settle the question, and pretending it does would be dishonest. What settles it is underneath the floor.
06aThe knob that gives the whole thing away
Look at any barometric altimeterair-squeeze-o-meter. On the right of the face is a little window showing a number like 29.92, with a knob beside it. That is the Kollsman windowthe little number window with the twisty knob. Paul Kollsman built the first accurate one in his attic in 1928, and Jimmy Doolittle flew the first blind flight in history with it the next year.
Here is what the knob does. Turning it measures nothing. It rotates the entire mechanism inside, moving the needle while the plane sits parked on the tarmac. Turn it by one inch of mercury and you have “climbed” 1,000 feet without moving an inch. It is not a sensor. It is a setting, and a human has to type it in.
So what is an altimeter, really? A squeeze-o-meter. It feels how hard the air is squeezing it, and guesses your height from that, because air squeezes hard near the ground and gently up high. But the weather keeps changing how hard the air squeezes down at the bottom — so a person has to keep telling the gadget what “the bottom” feels like today. That is what the knob is for.
So when the needle reads 35,000 feet, it is not saying “I am 35,000 feet above the ground.” It is saying “the squeeze out here matches what 35,000 feet would feel like, if the squeeze at the bottom were the number somebody dialed into my window.” That is a completely different sentence.
Pilots reset it roughly every hundred miles. Get it wrong and you fly lower than you think. There is an old warning for it: “high to low, look out below.”
The argument dies on its own instrument. “The altimeter reads level, therefore the Earth is flat” needs the altimeter to be measuring height above the ground. It is not, and every pilot knows it is not — because they spend the whole flight correcting the thing by hand.
07Where the curve is written down
The plane is not hiding the curve. It is running on it.
The correction inside the navigation box
The inertial reference unitbox that knows which way is up has to keep its idea of “down” aimed at the middle of the Earth as the plane travels. To do that, it continuously subtracts a rotation equal to velocity over radiusspeed divided by the size of the ball. That is the same v ÷ R the claim says is not real, written into the working math of the navigation system.
Leave that correction out for one hour and the box’s idea of “down” is wrong by 8 degrees. The plane then does not know where it is. Every airliner does this math, continuously, on every flight. The roundness of the Earth is not a footnote in the avionics. It is holding the roof up.
The route it flies
The flight computer plans a great-circle route on the WGS-84 ellipsoidthe shortest path across the official ball-shape of Earth, and the plane flies it. Between two cities at the same latitude, the compass heading changes the entire way. A flat map has no reason for that. A ball insists on it.
The same shouting radio
That ADS-B (Automatic Dependent Surveillance – Broadcast, the position report every airliner sends out about itself) message also carries a height from GNSSGPS, measured against the WGS-84 reference ellipsoidthe official ball-shape, a bit squashed at the poles — a ball whose waist is about 21 km wider than its poles. Every receiver on the planet, including the one feeding the app on your phone, is decoding a height above a ball. If the Earth were flat, that number would be meaningless.
08What a pilot can see
The curve is not invisible from an aircraft. It turns up in the one place where it is big enough: the depression of the visible horizon below the astronomical horizonthe edge of the world sits lower than straight ahead.
| Height | How far below level the edge sits | How far away it is |
|---|
| Standing on a beach | 0° | ~5 km |
| 35,000 ft (a normal flight) | 3.3° | 369 km |
| 60,000 ft (Concorde) | 4.3° | 483 km |
| 128,000 ft (the balloon jump) | 6.3° | 706 km |
Being straight with you: those numbers assume there is no air. They are the pure
geometry of a ball. But there
is air, and air bends light downward a little, which lifts the horizon
and makes the dip slightly smaller. So if you actually took a level up there and measured it, here is what
you would really get:
| Altitude | Pure geometry says | What you would actually measure | How far you can see |
|---|
| 35,000 ft | 3.3° | 3.1° | ~398 km |
| 60,000 ft | 4.3° | 4.0° | ~521 km |
| 128,000 ft | 6.3° | 5.9° | ~762 km |
We are telling you this
before you go and check, not after. And notice it does not matter to the
argument at all — a flat Earth says the dip is
zero at any height. Not 3.1°. Not 3.3°.
Zero. We would rather hand you the exact right number than have you catch us rounding.
On a flat Earth the edge would sit at eye level from any height, and that middle column would be zero all the way down. It is not zero. It grows as you climb, the way a ball 6,371 km across requires, and you can measure it out of a passenger window with the level app on your phone.
The entire page, for a nine-year-old. The plane never tips down, because it is not flying in a straight line. It is riding an invisible floor made of squished air, and that floor is already curved around the ball. The gadget in the cockpit measures the nose against “down,” and “down” swings around with the plane, so there is nothing for it to show. The black box counts in steps far too big to notice. The shouting radio never mentions the nose at all. But the box that steers the plane subtracts “we are on a ball” math every second of every flight — and if you switch that off, the plane gets lost.