I was on a flight1 from San Antonio, TX to Toronto and, when arriving in Toronto, the plane went quite a way past the airport, made a big U-turn, and then landed. We were coming from the west.

I thought it might have been a need to adjust for the runway, but as I said, we went quite a way past the airport, meaning after the turn we had to fly "a little bit to get back to the airport". All that even before starting the process of descending for landing (if it was just a matter of adjusting the direction for the runway, what would be the need to keep the high altitude and go so far to make the U-turn?).

Does anyone know the reason? It seems that some flights coming from the west do that move.

Air Canada 7722, October 8, 2017 (history will not persist)

enter image description here
(flightradar24.com) Image shows the flight's route with a final left U-turn prior to landing.

  • 7
    $\begingroup$ Could be a lot of things, most likely they had to enter the traffic flow, so they had to fly out and then get in line behind somebody on a long final. You can't just jump in the middle because there are spacing requirements. $\endgroup$
    – Ron Beyer
    Oct 13, 2017 at 3:36
  • 5
    $\begingroup$ Ask yourself what is the alternative; should some planes land from the west end while other planes are landing from the east end of the runway? What could go wrong? If nothing goes wrong on the runway, what could go wrong on the taxiway? $\endgroup$ Oct 13, 2017 at 14:45
  • 1
    $\begingroup$ See faraim.org/aim/aim-4-03-14-187.html $\endgroup$
    – Steve Kuo
    Oct 13, 2017 at 14:45
  • $\begingroup$ I live near an airport and often take my kids to watch takeoffs/landings. From my experience, the answer to @EricLippert question seems obvious. They take off one after another. In your airplane seat you don't realize there are other planes, but from my picnic bench (just outside the restricted runway area) there is a clear queue of planes.and it is obviously better if everyone was going the same way. $\endgroup$
    – emory
    Oct 13, 2017 at 15:22
  • $\begingroup$ Pilots always maintain as much altitude as possible until they are ready to land. If the plane looses power you want as much time and distance as possible to glide the plane to a clear landing spot. $\endgroup$ Oct 14, 2017 at 15:09

8 Answers 8


Airplanes prefer to land in a headwind. This reduces their groundspeed compared to landing with the same tailwind.

Braking distance grows quadratically compared to groundspeed. So any significant reduction in groundspeed is worth having.

Successive airplanes should all come from the same direction to land. This removes any danger of collision when a plane fails a landing and goes around into what would otherwise be oncoming air traffic. Planes also prefer to take off into headwind, so if takeoffs are intermingled with landings (very common in a busy airport) they won't interfere with each other if they all go the same way.

The tower monitors the wind and then decides which way the airplanes should come in to land and directs landing craft to land from that direction.

  • 1
    $\begingroup$ -1 for being too generic: while the information here is correct, I don't think it answers the specific question stated by the OP. $\endgroup$
    – kevin
    Oct 13, 2017 at 10:31
  • 69
    $\begingroup$ @kevin Au contraire. I especially liked that this answer was tailored towards the obvious non-existance of any prior knowledge in the OP. Compare with Geocrafter's answer, which, while technically correct and ultra specific, is almost worthless to the OP, because it's loaded with aviation lingo and places the wind and serialization aspect as an aside. OP appeared to need a general answer worded for a non-aviation receipient. $\endgroup$ Oct 13, 2017 at 12:41
  • 9
    $\begingroup$ @kevin I think you're interpreting the question as being more specific than it is. $\endgroup$ Oct 13, 2017 at 14:34
  • 5
    $\begingroup$ @kevin What exactly do you think the specific question is that this doesn't answer? The exact wind speed and direction? What the ATC was thinking? $\endgroup$
    – Kevin
    Oct 13, 2017 at 19:02
  • 6
    $\begingroup$ @kevin that is not true and that's why IAF and FAF are not the same thing. This answer is very good and you are just nitpicking. Obviously, when the aircraft arrives from a different direction than the runway in use direction, it will have to turn somewhere. I really find that obvious enough no matter where some STAR terminates. $\endgroup$ Oct 14, 2017 at 15:15

As a Virtual Controller on vatsim.net at Toronto, I can tell you the following:

  • Depends on runways, most of the time they operate runways 24R and 23, which lands planes facing south west. Runways are chosen from the wind direction, as planes preferably land into the wind. Winds at Toronto just most likely will favor the 24 runways.
  • The STAR(s) "Standard Terminal Arrival Routes" coming from the west puts planes in a direction that is completely opposite (downwind) direction for runway 24R or 23, then get directions (vectors) to align with the runway by a controller.
  • The approach airlines mostly fly into Toronto is the ILS which you can think of like a beam coming from the runway that lines the plane up with the runway hence for the plane the intercept that beam, the plane cant be turned in right away. Therefore, ATC will turn the plane perpendicular to the runway, then a 30 degrees difference to the runway for the aircraft to intercept the beam.

Here is an image of how it looks like:

enter image description here
Example approach. Big light blue arrow indicates what the wind direction may have been like. See: What determines in which direction a runway is used?

Here, the plane is coming from the South-West, following GPS locations until the downwind. It then gets turned inbound for the ILS.

  • 1
    $\begingroup$ ILS = instrument landing system. "A radionavigation system which provides aircraft with horizontal and vertical guidance just before and during landing and, at certain fixed points, indicates the distance to the reference point of landing." — Article 1.104, ITU Radio Regulations (ITU RR), p. 16, itu.int/dms_pub/itu-s/oth/02/02/S02020000244501PDFE.PDF $\endgroup$
    – Pocketsand
    Oct 13, 2017 at 18:12
  • $\begingroup$ What is a virtual controller? You get paid or just do that for hobby? $\endgroup$
    – Cloud
    Jul 11, 2018 at 9:40
  • $\begingroup$ @Cloud, it means he plays video games. In my mind that dilutes credibility, although his answer is fair enough to stand on its own merit. $\endgroup$ Sep 12, 2019 at 17:30

I think the other answers have covered the "Why U-turn" part well, but to address the other part about why so far past

I thought it might have been a need to adjust for the runway, but as I said, we went quite a way past the airport, meaning after the turn we had to fly "a little bit to get back to the airport". All that even before starting the process of descending for landing (if it was just a matter of adjusting the direction for the runway, what would be the need to keep the high altitude and go so far to make the U-turn?).

I most often travel through Harsfield-Jackson in Atlanta (busiest in the world) and doing the long U-turn is pretty common. The reason it's so long is that you're getting in line behind all the other traffic. I've watched the plane I was on pass as many as a dozen other aircraft before turning and falling in line (this is less common now since they've added more runways).

By contrast, flying into a smaller airport with less traffic often involves very little distance between passing the airport and turning. On one occasion we were running late into Jacksonville, FL due to an equipment malfunction in ATL. The pilot pulled a fairly sharp turn on that one when we were just past the airport. We had barely straightened up before we were over the runway.

  • 5
    $\begingroup$ Yeah, you typically fly a pretty long downwind at ATL. It's kind of neat to drive through or fly over Atlanta when it's dark, as you can see two giant swirls for the traffic patterns headed into the North and South arrival runways. Flights from Florida to Nashville frequently fly over midfield at ATL, so they provide a neat view of the traffic patterns if it's dark out. $\endgroup$
    – reirab
    Oct 14, 2017 at 1:27
  • $\begingroup$ Weather could be a factor. For full ils you might want to be lined up with localizer for a few miles before the glide slope is intercepted. Also wake turbulence for larger aircraft. Also noise abantmance. Some times they let them do short finals too. $\endgroup$ Oct 14, 2017 at 4:02

I live about 100km west of Pierson Airport at Toronto and can inform you, as others have mentioned, the runways used and landing /takeoff direction is dictated by the current prevailing winds.

My house is situated directly under the flight lane from a local beacon to Toronto, and I can tell which way the wind is blowing simply by listening and watching the aircraft go overhead.

When there is an east wind and at the peak time of day, around 5pm - 7pm, aircraft will fly overhead at a regular separation distance on their decent approach to the airport. When winds are from the west they pass MUCH higher overhead.

Air traffic control vectors the aircraft into this line. When the wind blows from the east, aircraft that arrived in Toronto from the east fly past the airport and, if I look east from my house, I can see them turning into gaps in the line at appropriate separation distances. If no gap is available due to traffic, they circle at appropriate, stacked, heights till a suitable gap opens up. The lowest aircraft joins the line and any remaining aircraft drop down a level in the stack.

The prevailing winds in this part are normally from the west, so most of the time all this happens to the east side of Toronto, as it was in your case.

Intrestingly, since the incident with Air France Flight 358, when a storm is happening, aircraft are now stacked from both directions during severe weather events at the airport.

The busier the airport the more complex all that stacking becomes. The image below is for London, Heathrow. Note they have stacks for planes approaching from the four corners of the compass that are then filtered into two runways as appropriate by ATCs. Again, this whole setup will reverse depending on wind direction.

enter image description here
(Image source)

  • $\begingroup$ The prevailing wind at Heathrow (and, indeed, in most of the UK) is from the west, so the diagram shows the most common scenario. $\endgroup$ Oct 15, 2017 at 13:04
  • $\begingroup$ FYI, the reason that the arrows between the stacks and the runways are so fat is that exactly where each plane turns will vary as needed to establish proper separation for merging traffic streams and variations in the ground speed in each direction due to wind intensity. $\endgroup$
    – StephenS
    Jan 28, 2019 at 18:00

It looks suspiciously like the NUBER2 STAR into CYYZ, possibly with radar vectors early for traffic or weather separation/deviation, followed by radar vectors to the initial fix for either a visual or instrument approach to RWY 24 L/R.

enter image description here


Shared rules must meet everyone's requirements

First, aircraft land in a headwind, hence the U-turn.

They're in a stack of other aircraft approaching, so they have to "get in the queue". Even if this Cessna could easily tailwind-land on a 10,000 foot runway in the tailwind, the 777 behind it really shouldn't.

Making the stream of aircraft manageable requires a standard approach. As you can imagine, this is "lowest common denominator" to accommodate all aircraft.

A lot of work to do: stabilizing the approach

Why is it so long? Because there's a lot of work to do. The pilots' time is a chokepoint, that is to say it's a high workload time. If the work isn't done, they can't land and have to go around.

They have to either find the localizer (horizontal beacon) or get roughly lined up on the runway. Then they have to line up vertically on the glideslope beacon (or the four lights which are a visual version of the same thing).

They have to get their speed where it belongs (several times during the approach), get power where it belongs, drop landing gear, set flaps in coordination with their speed reductions, work their checklists, and prepare the aircraft in all other respects.

If they are dealing with crosswinds, they also must have a strategy to correct for that.

All this is called "stabilizing the approach".

Then they reach a decision point called "minimums", some ways back depending on their flight mode. Here they decide whether they are able to continue the landing: Can they see the field? Is the approach stablized? Continuing to press a bad approach is a bad habit.

So the approach pattern is long to give everybody time to do this.

Also, that approach is slow

The final (upwind) leg of the approach also has the slowest airspeed of the whole flight, and it's into the wind. So it will tend to take a lot longer than the downwind leg where you flew past the airport. So it may seem like it's really far, when it's just really slow.


Although it felt strange to you, it is very common. For efficiency reasons, sometimes it occurs at large airports that aircraft land on a direct approach. But, in general aviation, what you saw is completely standard..

  • 2
    $\begingroup$ Do you perhaps have this backwards? The asker was on a commercial flight to a large airport. At large airports, it's much more efficient for everybody to land (and take off) in the same direction since you can pack vehicles moving in the same direction much more tightly than vehicles moving in opposite directions. $\endgroup$ Oct 13, 2017 at 10:14
  • $\begingroup$ Precisely. In order to have everyone land in the same direction, a pattern is established. $\endgroup$ Oct 13, 2017 at 14:09
  • $\begingroup$ You describe everybody landing in the same direction as completely standard in GA. However, you begin the sentence with "but", which suggests that this situation is different the previous thing were talking about (large airports). It's not different at all. Indeed, at large commercial airports, it is more important for everybody to land in the same direction, because it's the only way you can possibly get the throughput. At an airport with only a few plane movements an hour, it wouldn't affect throughput at all. $\endgroup$ Oct 13, 2017 at 14:28
  • 1
    $\begingroup$ @DavidRicherby The difference he was referring to is more applicable for the case of traffic arriving from the direction where they would already be on-course to the current runway direction. At controlled fields, this traffic will often just be vectored straight onto final. At uncontrolled fields, it's more customary (at least for VFR traffic in the U.S.) to still fly past the airport and enter the pattern 45 degrees to the downwind (opposite direction of runway traffic) and then turn base to final (which is the "U-turn" that the OP asked about.) $\endgroup$
    – reirab
    Oct 13, 2017 at 14:43

Pilots always maintain as much altitude as possible until they are ready to land. If the plane looses power you want as much time and distance as possible to glide the plane to a clear landing spot...

runways are always entered from one direction for all traffic. If you approach a runway from the wrong direction then you have to make a u-turn and come back to turn downwind to face the runway. the course your plane took may have been extended for spacing other planes or the pattern may have appeared long to your sense of scale. Large planes fly fast and need a lot of space to turn around while maintaining safe distances from other aircraft.

Small airports fly a standard clockwise pattern that always uses a u-turn. Larger airports with air traffic control and instrument guided approaches will have approach patterns published in charts. Unless otherwise instructed pilots entering the airport airspace are expected to use the published approach; although really an approach would never be made without contacting air traffic control.

  • $\begingroup$ This answer mostly explains "what" rather than "why". Essentially, you're saying "There's some procedure that requires pilots to do this", which is good to know but doesn't really explain why the asker's plane took the route that it did. Why are these procedures in place? $\endgroup$ Oct 15, 2017 at 13:06

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .