How the atmosphere flies your airplane — density altitude, airspeeds & parcels of air · interactive simulations, tuned for KANP · weather hub →
One idea runs through this whole page: everything your airplane does starts with how many air molecules it can grab per second. The dots floating behind this page are those molecules — they thin out, speed up, and moisten as you scroll and as you move the sliders, always matching the panel you're looking at. Start by grabbing a parcel yourself:
Watch the page background as you drag: heat the parcel and the dots speed up and thin out (warm air expands) · squeeze it with pressure and they crowd together · raise the dewpoint and blue vapor replaces heavier dry air — push it to the temperature and the vapor condenses into cloud.
Air gets thinner, colder and lighter as you climb. Drag up and down the column (or use the altitude slider) and watch pressure, temperature and density fall together — then warm or cool the day and see the whole column change. Watch the background dots thin out as you climb: every other panel on this page is a consequence of these three curves.
Lift and engine power both come from how many air molecules you can grab per second — that's density (ρ). Pressure pushes molecules together; temperature pushes them apart. The ideal gas law ties the three together:
Climb and pressure drops → fewer molecules. Heat the air and it expands → fewer molecules again. A hot day is a form of altitude: the wing can't tell whether the air is thin because you climbed or because it's 95 °F on the ramp — that equivalence is exactly what density altitude measures in panel 2.
The ISA "standard day" (29.92 inHg and 15 °C at sea level, cooling 1.98 °C per 1,000 ft) is just an agreed reference so that altimeters, performance charts and controllers all speak the same language.
Panel 1 showed that climbing and heat both steal molecules — this panel turns that into the two numbers pilots plan with. Set the field conditions and watch the airplane's effective altitude move: pressure altitude is what the altimeter reads on 29.92; density altitude is where the airplane performs like it is — the number that stretches your takeoff roll on a July afternoon.
Pressure altitude corrects the field for the day's pressure: set 29.92 and the altimeter shows it. Roughly ±1,000 ft per inch of mercury away from standard:
Density altitude then corrects PA for non-standard temperature — about 120 ft per °C away from ISA:
This lab computes both exactly (NWS formulas, including the humidity term via vapor pressure — muggy air is less dense than dry air, because H₂O molecules are lighter than N₂/O₂, so a humid day quietly adds a few hundred feet more. Watch the blue dots multiply behind the page as you raise the dewpoint: each one displaced a heavier dry-air molecule).
Why the performance numbers move so fast: takeoff distance suffers twice — the engine makes less power in thin air, and the wing needs a higher true speed for the same indicated liftoff speed. The classic planning rule: expect roughly +15 % ground roll and −7½ % climb rate per 1,000 ft of DA.
Here's the missing link between thin air and your instruments. The pitot tube doesn't measure speed — it catches molecules and measures how hard they pile up: ram (dynamic) pressure, ½ρV². Fly faster and each molecule hits harder; fly higher and there are fewer of them to hit. The gauge can't tell the difference — that's not a flaw, it's the design, because the wing can't tell the difference either.
Every aerodynamic force on the airplane — lift, drag, control feel, the stall — scales with the same quantity the pitot tube measures:
So when the gauge reads 63 kt on final, the wing is making exactly the lift it makes at 63 kt at sea level — even though at a high-DA airport your true speed (and ground speed, and tire speed, and runway consumed) is much higher. Rotate, approach, and stall speeds are flown indicated at any altitude for exactly this reason.
Engine performance lives here too. A normally aspirated engine is its own molecule-catcher: the intake stroke pulls in a fixed volume, and thin air puts fewer oxygen molecules in that volume — roughly 3 % power lost per 1,000 ft of density altitude. Fewer molecules over the wing and fewer in the cylinders is why the panel-2 takeoff numbers deteriorate twice as fast as intuition expects. (A turbocharger is simply a machine for putting the missing molecules back.)
Try the "climb and hold the needle" button: the needle stays put while true airspeed grows — you get free speed with altitude. Panel 4 shows the forces doing it; panel 5 counts the knots.
The same ram pressure that works the gauge is what actually flies the airplane. The airfoil bends the stream — upwash ahead, downwash behind — and the reaction is lift; the propeller — just a wing spun sideways — throws molecules back for thrust, and its blast weakens as the air thins. The nose rides at the angle of attack the lift equation demands: slow down or load up and watch it rise toward the 16° critical angle where the flow lets go. Against all of it: weight, the one force density can't touch, and drag, the toll the molecules charge for passage.
Lift = weight, always (in level flight). The airplane doesn't get to choose how much lift to make — weight sets the bill and the wing must pay it. What density changes is the cost of paying: thin air means a higher true speed or a higher angle of attack for the same lift. Load the airplane heavier and the bill itself grows — stall speed rises with √weight, and the wing pays extra induced drag for the privilege (drag born from making lift — strongest when slow and heavy, exactly where takeoff lives).
Thrust is molecule-throwing. Newton, not magic: the prop accelerates a column of air backward and the reaction pushes forward. Thin air cuts it twice — the engine inhales fewer oxygen molecules (less power to spin the prop) and each blade bites fewer air molecules to throw (less grip per horsepower). That double cut is why high-DA climb performance collapses faster than the power chart alone suggests.
Drag is honest. At the same indicated speed, drag is the same at any altitude — same q = ½ρV², same toll. That's the quiet gift of altitude: at 8,000 ft the same drag buys you ~12 % more true speed. The whole cruise-performance game in panel 5 is that trade: climb until the engine's fading power available meets the drag you must pay.
Weight is immune. Gravity doesn't care about density, temperature, or humidity — every other player on this page weakens with thin air except the one pulling down. That asymmetry is the density-altitude problem.
Now chain it together. Panel 3 showed the gauge under-reads in thin air — so true airspeed grows with altitude for the same indicated, and panel 4 showed the engine fading while drag stays honest: put together, density picks your cruise speed and your best altitude. Then the wind, which the airplane can't feel at all, decides what the ground sees. Left: how TAS grows with altitude. Right: the wind triangle — drag on the compass to steer the wind.
Temperature, pressure and humidity all funnel into one number — density — and density writes this curve:
Climbing helps at first: drag at the same indicated speed is the same, so every foot of thin air is free true airspeed (+2 %/1,000 ft).
Then the engine gives out: a normally aspirated engine loses ~3 % power per 1,000 ft of DA. Where fading power meets the drag bill is the hump — the best-cruise altitude.
Indicated (IAS): the ram pressure from panel 3, dressed up in knots. It's the speed the wing cares about: stall, rotation and approach speeds stay honest at any altitude, which is exactly why we fly them indicated.
True (TAS): your actual speed through the air mass. Thin air → fewer molecules → the same true speed makes less pressure, so the indicator under-reads. Recover it with the density ratio σ:
(Strictly the correction applies to calibrated airspeed; below ~200 kt CAS ≈ IAS and compressibility is negligible, so this lab treats them as equal.)
Ground speed (GS): TAS plus the wind vector. The air mass is a conveyor belt — you fly at TAS within it while it carries you wherever it's going. Crab into the wind (the wind-correction angle) to make the track match the course; whatever component of wind remains along the course adds to or steals from GS. Headwind on final is why landing distance charts and fuel-time planning both start from the wind.
So far the air sat still while you flew through it. But the same density physics makes the air move on its own: warm a bubble of surface air and it's lighter than its surroundings — the atmosphere flies it exactly like a wing flies you. Rising air expands and cools at a fixed rate (3 °C / 1,000 ft while dry); whether it keeps climbing — building cumulus and turbulence — or sinks back depends on how fast the air around it cools with height. Drag the parcel, or heat it like a sun-baked field, and watch stability play out.
A rising parcel doesn't mix with its surroundings — it just expands as pressure drops, and expanding gas cools: the dry adiabatic lapse rate, 3 °C per 1,000 ft. Its dewpoint falls much more slowly, so temperature and dewpoint converge at about 2.5 °C per 1,000 ft. Where they meet, water condenses — that altitude is the lifting condensation level, the flat cumulus cloud base you can estimate from the ground:
Once condensing, latent heat is released and the parcel cools more slowly (~1.8 °C / 1,000 ft moist) — which is how storms feed themselves.
Stability is a race of lapse rates. If the environment cools faster with height than the parcel does, a lifted parcel finds itself warmer than its surroundings — lighter — and it accelerates upward: unstable (thermals, cumulus, an afternoon of bumps at pattern altitude). If the environment cools slowly — or warms with height, an inversion — the lifted parcel is colder and heavier and sinks home: stable (smooth air, but haze and poor visibility trapped below, a very Chesapeake-summer combination). In between, stability is conditional on reaching saturation.