Some planes in the sky leave trails that persist and spread, and other planes, in the same sky, leave short-lived trails, or no trails at all.
These trails are actually called contrails, short for “condensation trails”. They are not smoke from the engines, they are formed when the water in jet exhaust (and there’s quite a lot of it, like car exhaust on a cold day) mixes with wet cold air, and condenses and freezes into ice crystals. Contrails are actually a type of cirrus cloud. When the air is wet and cold enough the trails can stay around for a long time, and sometimes spread out.
This difference between trails that fade away, and trails that spread, is often used as evidence of the “chemtrail” theory, which states that the longer lasting trails (or some of them) are being deliberately manipulated for some reason. So you see helpful images like this.
But this is wrong. Contrails can fade away, and contrails can persist and spread. It depends on the air they are formed in.
Now there are two main reasons why some planes leave trails and some nearby planes do not. The less common reason, is that different planes have different engines. Some engines will leave a contrail in air where another engine will not. Here, for example are an Airbus A340 (maiden flight: 1991) on the left, leaving contrails, and a Boeing 707 (maiden flight: 1957) not leaving contrails. Both are flying at 33,000 feet (part of a German test to study contrail formation), but the exhaust of the newer engines of the A340 is at a lower temperature, and so makes contrails in a wider range of conditions*.
You can also get a similar effect with engines at different power settings, especially if it effects the exhaust temperature.
But here’s the main reason why you see trails on some planes but not on others, and I’ll emphasize it, because although it’s simple, it’s also easy to miss.
The planes are at different altitudes.
Yes, it’s really that simple. The reason that one plane makes contrails, or makes contrails that persist, and the other plane does not, is that they are in different regions of air. For simplicity, let’s refer to these regions of air as wet air and dry air, although the differences are a bit more complex.
When the plane is in wet air, it makes a contrail. In dry air it does not.
Surely, you might object, they would have to be miles apart? Well, no, and that brings me to another point I fear I must emphasise:
Wet and dry air can exist within a few feet of each other.
Consider, for example, clouds:
Inside the cloud it’s wet. Outside it’s dry. What’s the difference between inside and outside? It’s a few feet.
Look at the bottom of those clouds, see them extend off into the distance. They form a layer at a specific altitude. Above that altitude there are clouds. Below it there are no clouds. The difference between clouds and no clouds is just a few feet.
Now those are low altitude cumulus couds. Let’s look at high altitude clouds.
Again they are in a flat layer. The different between being in the layer and not in the layer is just a few feet.
This layering of the air into wet and dry layers is not limited to clouds. Seemingly clear air also contains exactly the same kind of variation in layers. This was very neatly illustrated by the recent launch of the Solar Dynamics Observatory. As it ascended it did not leave a contrail, until it hit a layer of wet air, when it left a contrail that lasted quite a while, and then it went into dry air again, and no more contrail
So, if a plane were flying in that middle region then it would probably leave a persisting contrail. If it were above or below it then it would not.
But, you may cry, the planes are at the same altitude. Now you might even disagree with a “few feet”, and say the planes were too close for them to be in different layers. I’d respond with:
You can’t tell how high a plane is
And you certainly can’t tell if one plane is at the same altitude as another. These planes fly at 30,000 to 40,000 feet. Let’s see what the visual difference is at around that altitude:
I took one image of a jet nominally at 35,000 feet. Then scaled it for 34,000 (102.9%) and 32,000 feet (109.4%). I think you’ll agree they all look pretty much the same. Especially as this is more zoomed in than you’d see with the naked eye, which would be more like:
Or, considering you generally can hardly see the wings of a plane with the naked eye, a fully realistic representation would be:
If the planes are flying lower, then it’s still similar. If the top plane was flying at 20,000 feet, then the bottom would be at 18,285 feet, still nearly 2,000 feet apart, and looking pretty much the same to the naked eye.
And that is with the same model of plane, directly overhead, and right next to each other. A situation that almost never occurs. If the planes are different, or separated, or at an angle to you, then it is IMPOSSIBLE for you tell the relative altitudes when they are high in the sky. Just look at this:
Or from the ground, with the planes at 30,000 feet.
They look about the same height, right? In fact if they were not overlapping, you’d think the JAL plane was lower, as it seems bigger, hence closer. But actually the JAL plane (a B777) is at least 1000 feet above the DHL plane (an A300).
And look at some planes on the ground, where we know they are all the same distance from the camera. The differences in size are very significant:
So, a simple question gets a simple answer:
The planes leave different trails because the planes are at different altitudes.
http://contrailscience.com/broken-contrails/ – Why contrails are often broken, and start and stop abruptly.
Debunked: High Bypass Turbofans do not make Contrails [actually they make more] - A more detailed look at why modern engines make contrails in a wider range of conditions.
*(Edited 3/15/2015): when I wrote this article in 2010 I’d said that the newer more efficient engines produced “more water”, and while it’s true that the cleaner the combustion the more water is produced, the difference is negligible for the two engine types discussed. The key difference is the exhaust gas temperature, as explained in the more recent Metabunk article