# Why some planes leave trails and others don't

> Trails form in very cold air and last only where it's humid enough for ice to grow. Three balloon soundings from one evening show the difference.

By Derek Johnson. Updated Nov 7, 2026. Source: https://skycapapp.com/articles/overhead/why-some-planes-leave-contrails/

The air usually decides. A trail forms when hot, wet exhaust hits air cold enough to freeze it, which at cruising height is about −40 °C or colder. Whether it lasts depends on how much water vapour the surrounding air already holds. In dry air it's gone in seconds or minutes. In air humid enough for ice to grow, it can spread for hours.

Burning jet fuel makes water vapour. A contrail is a cloud made by mixing, the same way your breath shows on a cold morning. Your breath doesn't stay long on most mornings because the air around it is dry enough to take the water back. Same idea, 7 miles up.

## Three balloons on one evening

The National Weather Service launches weather balloons twice a day, and each one reports temperature and moisture on the way up. I took three from the launch of 00:00 UTC on Nov 7 2026, which is the evening of Nov 6 in the US, and worked out humidity with respect to ice near cruising height.

| Launch site | Height | Temperature | Humidity over ice |
|---|---|---|---|
| Upton, Long Island | 35,445 ft | −52.4 °C | 134% |
| Oakland, California | 35,402 ft | −47.3 °C | 79% |
| Lincoln, Illinois | 34,352 ft | −53.1 °C | 62% |

All three are cold enough for a trail to form. So a jet in any of them could leave one. What differs is the last column. Over 100% means the air holds more vapour than ice can tolerate, so ice crystals in the trail grow instead of shrinking.

Over Long Island, a trail that evening should have lasted and spread. Over Illinois, it should have faded fast. Oakland sits between, and it shows something more useful.

## Why two planes in one sky differ

The same Oakland balloon passed through a thin wet layer. At 33,392 feet the humidity over ice was 111%. At 33,554 feet, 162 feet higher, it was 89%. Ice-saturated layers can be that thin, and they have edges in every direction.

That's your two contrails. One plane cruises inside the layer, and its trail hangs on and widens. Another crosses the same sky 2,000 feet higher, in drier air, and leaves a short line that stops almost behind the tail. Engines differ a little, and a modern engine can make trails in slightly warmer air than an old one. But for two jets above you at the same moment, altitude is the first thing to check.

It also explains a trail that starts and stops. The plane hasn't changed anything. It went in and out of a patch of humid air.

## Other things you might be seeing

Short white streaks from the wingtips, usually on a damp day near an airport, are different. Low pressure over the wing cools the air enough to condense it for a moment. NASA scientist Patrick Minnis calls these aerodynamic contrails, and at low height they're short-lived liquid water.

A trail that drifts sideways and widens into a sheet is a persistent one sheared by wind at height. It can turn into something that looks like natural cirrus. Researchers can't always tell them apart from the ground either.

And a plane with no trail at all may be flying well below 30,000 feet, where the air is usually too warm. If you're checking, a tracker's altitude number tells you.

## How I got the numbers, and how far to trust them

Humidity over ice comes from each balloon's temperature and dew point, using a standard vapour-pressure formula over water and another over ice. The soundings came from a public archive.

Balloon humidity sensors are at their worst in this cold. Treat 134% as "well over 100", not as a measurement to the percent. Some of that Long Island layer may have been real cloud, which means you might not have seen a trail in it at all. A balloon also drifts, and it samples one thin column of air once. And I picked these three launches to make a point, so they say nothing about climate.

Whether a trail forms in marginal air depends on details of the engine and the exact temperature, and forecasts of these layers are still poor. The FAA and NASA both study them, because lingering trails trap some heat.

[What's that plane overhead](/articles/overhead/whats-that-plane-overhead) shows how to find a plane's altitude, and [how to read a flight's path](/articles/overhead/how-to-read-a-flight-path) reads the line once you've found it. If you're at the airport waiting, [this is the next hour](/articles/landing/when-to-leave-to-pick-someone-up).

> With two trails above you, Overhead in Skycap names each plane and gives its altitude, so you can see which is higher. The trail on its camera view is a drawing, shown above 26,000 feet by day, and it doesn't check whether a real one is forming. [What Overhead shows](/app/features#overhead)

## Sources

- [Patrick Minnis, Contrails, NASA Langley](https://www-pm.larc.nasa.gov/site/doc-library/208-Minnis.CONTRAILS.Ed2.pdf)
- [FAA, contrails fact sheet](https://www.faa.gov/contrails/fact-sheet)
- [EPA, contrails](https://www.epa.gov/contrails)
- [NASA, Review of Past Decade of Aviation Contrail Research, 2025](https://ntrs.nasa.gov/api/citations/20250007272/downloads/NASA-TM-20250007272.pdf)
- [National Weather Service, upper air observations](https://www.weather.gov/upperair/)