← back to the dashboard

What the barometer is telling you Drive a storm across the map and watch the needle do what the dashboard's needle does.

From above  

HYAK WEST EAST weather moves this way → lines are 4 hPa apart

From the side  

 
 
—hPa
 
·—
Real air pressure here—
Low's centre—
Centre arrives in—
Dashboard would say—
—
 
24h ago now
far westapproachingoverheadpassingfar east
a weak troughan ordinary fronta named storm
The big idea 30 seconds

You live at the bottom of an ocean of air about sixty miles deep, and it is heavy. The column of it standing over every square metre of ground weighs about ten tonnes — roughly a tonne pressing down on your own shoulders right now. You don't feel it because the same pressure is pushing out from inside you, and because you have never once been without it.

A barometer weighs that column. That's all it does. And it turns out that the weight of the air above you is one of the best clues there is about what the weather is going to do next, because air that is on its way up weighs less than air that is on its way down.

Standard sea level
1013.25 hPa
The same number as 1013.25 millibars — hPa and mbar are identical units with different names. Also 29.92 inches of mercury, or 14.7 psi.
Highest ever recorded
1083.8 hPa
Agata, Siberia, on the last day of 1968, in a monstrous pool of cold arctic air.
Lowest ever recorded
870 hPa
Inside Typhoon Tip, 1979 — the strongest storm ever measured. The air there weighed 14% less than standard.
Normal swing here
~980–1040 hPa
Everything an ordinary year at the Pass throws at you fits in that 60-point band.

The instrument itself is 1643 vintage. Torricelli stood a sealed glass tube of mercury upside down in a dish and found that the air outside could only hold a column about 760 mm — 30 inches — high, and that the height wandered from day to day with the weather. Mercury, because it's thirteen times denser than water: the same experiment with water needs a tube ten metres tall, which is also, not coincidentally, the deepest a suction pump can ever lift from.

Why falling pressure means rain the actual mechanism

This is the part worth understanding, because once you have it the barometer stops being a superstition and becomes an instrument.

Low pressure means less air overhead than usual. Air from around it flows in to fill the gap — and since it can't go down through the ground, it has to go up. Watch the side view: the arrows converge at the surface and then turn skyward over the centre.

Rising air expands, because there's less pressure squeezing it the higher it goes. Expanding air cools — about 10°C for every kilometre it climbs, which is why the top of a mountain is colder than its base even though it's nearer the Sun. And cool air cannot hold as much water vapour as warm air. So the vapour that was invisible down here condenses into droplets on the way up, and droplets in bulk are what we call cloud. Pile enough of them together and they start falling out as rain.

And high pressure is the same story in reverse

More air overhead means air is sinking. Sinking air is compressed, which warms it, and warm air holds more vapour — so any cloud in it evaporates away. That is why a high gives you clear skies, and it's not a correlation, it's the same physics running backwards.

The chain in one line: low pressure → air rises → it cools → vapour condenses → cloud → rain. High pressure → air sinks → it warms → cloud evaporates → clear.

Why the wind goes around a low instead of into it

Air heading for the centre of a low gets deflected by Earth's rotation — to the right, here in the northern hemisphere — so instead of falling straight in it ends up circling counter-clockwise around the centre. That's what the curved arrows on the map are showing, and it's why every satellite picture of a Pacific storm is a spiral.

There's a piece of practical seamanship hiding in that, called Buys Ballot's law, and it works standing in your driveway: turn your back to the wind and the low is on your left. A little forward of your left shoulder, in fact. You have just located the storm without any instrument at all.

The change matters far more than the number why the dashboard shows an arrow

Here is the thing that takes a barometer from decorative to useful: a single reading tells you almost nothing. 1008 hPa could be a pleasant grey afternoon that stays that way for a week, or it could be the leading edge of something serious. What separates them is which way it's going, and how fast.

That's why the dashboard shows a trend arrow and a 24-hour trace rather than just a number. Drag the storm slider slowly and watch: the needle passes through exactly the same value on the way down and on the way up, and those two moments mean completely opposite things. Falling through 1008 is rain arriving. Rising through 1008 is rain leaving.

Change over 3 hoursWhat it means
−3 hPa or moreA strong front, moving fast. Wind first, then rain. The steeper the fall, the harder it blows — this is the one worth respecting.
−1.6 to −3A front on the way. Cloud thickening now, precipitation within six to twelve hours.
−0.4 to −1.6Slow decline. More cloud, maybe something later. From a high starting point this just means "the fine spell is ending".
−0.4 to +0.4Steady. Whatever it's doing now, it keeps doing for several more hours. The most common and least exciting reading.
+0.4 to +1.6Recovering. Showers tapering, sky starting to break up.
+3 or moreA sharp clearance behind a front. Bright, cold and often windy — and a rise this fast rarely holds.

Those are the exact bands the dashboard sorts your real pressure into, which is why the sentence it prints changes as the trace bends.

The old sailor's version: "Long foretold, long last; short notice, soon past." A slow, gradual fall brings weather that settles in for days. A sudden plunge brings something violent that's over quickly. Both halves are still true, and both fall out of the same physics — the width of the dip on the map is the length of the weather.

Why it reads 1017 when the air up here is really 915 the Hyak problem

The dashboard shows two pressures, and they're about a hundred apart. That is not an error, and the difference is entirely about where you live.

Hyak sits at roughly 882 metres — near 2,900 feet. A tenth of the atmosphere is already below you. The air genuinely pressing on the Pass is around 915 hPa, and it is never anything else, because you can't move the mountain.

If weather reports used that raw number, every mountain station on Earth would read permanently "STORMY" and every coastal one "FAIR", and no two places could ever be compared. So every station in the world quotes sea-level pressure instead: the raw reading, plus a correction for the weight of the imaginary column of air that would be there if you dug down to sea level. That's the ~1017 figure. It is a calculated number, not a measured one, and it exists purely so that a map of it shows weather rather than topography.

Measured at Hyak
~915 hPa
What a barometer sitting on the ground actually reads.
Reduced to sea level
~1017 hPa
The comparable number. What the dashboard's dial points at.
Rule of thumb
1 hPa / 30 ft
Near sea level, pressure drops about a hectopascal for every 28 feet you climb.

A thing you can feel on the drive home

Coming down I-90 from the Pass to Seattle, the real pressure around you rises by about 100 hPa in under an hour. That is nearly twice the entire range that weather manages at a fixed spot in a whole year, compressed into a drive — and it's why your ears pop on the descent and never pop because of a storm.

Same reason aircraft altimeters need setting. An altimeter is just a barometer with a height scale printed on it, so it can't tell a low-pressure day from being lower down. Pilots dial in the local sea-level pressure to correct for it, and above a certain altitude everyone switches to the standard 1013.25 instead — not because it's right, but so that every aircraft is wrong by the same amount and they still miss each other.

Three things people get wrong including the dial

1. The words on the dial are folklore, not physics

STORMY, RAIN, CHANGE, FAIR, VERY DRY — those come off 18th-century English parlour barometers, and they encode the typical weather of lowland southern England at sea level. They are a rough guide at best anywhere else, and at 2,900 feet in the Cascades they're barely even that. The dashboard prints them because they're charming and traditional. Trust the trend arrow over the word every single time.

2. A deep low isn't automatically a storm

What hurts is the gradient — how tightly the pressure changes over distance — because that's what drives the wind. On the map view, a storm with its lines crammed close together is a windy one; the same depth spread over twice the distance is a damp, boring day. And a low parked motionless overhead can sit there for days at 995 with a perfectly flat trace and nothing much happening, which is a real Pacific Northwest special. The dashboard has a phrase for exactly that one: "Low and not budging."

3. Pressure says nothing about temperature

A barometer cannot tell you whether what's coming falls as rain or as snow — it doesn't know how cold it is. That's why the dashboard reads the temperature separately before choosing its words, and why at 34°F and below the same falling trace gets described as snow instead of rain. Try the Make it freezing button above and watch the same storm get re-described without the needle moving at all.

And one that's actually right: "the calm before the storm" has a real basis. Ahead of an approaching warm front the wind often drops off and the sky goes flat and heavy while the pressure is already falling hard. The barometer knows before the wind does. That's the whole reason anyone bothered to invent it.

What this simulation simplifies the honest footnotes

The physics above is real. The toy at the top of the page is a cartoon of it, built to make the needle move in a recognisable way:

  • The low is a tidy mathematical dimple in a flat 1023 hPa background. Real systems are lumpy, have troughs and ridges hanging off them, and interact with each other.
  • It moves at a constant 42 km/h, due east, and never deepens or fills. Real Pacific lows speed up, slow down, and often do most of their deepening on the way in.
  • The rise behind is steeper than the fall ahead, which is true of real fronts and is the one asymmetry that got modelled.
  • The clouds are decorative. Cloud is keyed to the pressure field rather than computed from anything; real frontal cloud arrives hours ahead of the pressure minimum as high cirrus, then thickens and lowers.
  • The mountains are missing entirely. The Cascades wring out air forced over them, which is why the Pass gets weather the same system never delivers to Ellensburg.
  • The fronts are drawn in the textbook positions rather than derived from the pressure field.

The dashboard's own readings are real measurements from Open-Meteo, not this model. This page only exists to explain what they mean.