# How close can planes fly to each other over the ocean?

While on a recent Air Transat flight from Toronto Pearson to London Gatwick, several hours into the flight and over the Atlantic Ocean, my daughter and I noticed a plane behind ours to the left, flying quite close. Assuming there must be protocol for how close planes can fly to each other, I was more interested than concerned.

It then became apparent that the other plane was flying faster than our plane and coming closer. The plane came close enough for my daughter and me to see all the windows as its path led it to cross in front of our plane. It was only very slightly higher than our plane by mere meters. As it very quickly approached us we lost sight of it as it crossed in front of our path.

My daughter and I figured a crash was imminent. Our plane then altered its path, moving slightly down and to the right.

This was an overnight flight and at the time the cabin lights were turned down and the passengers sleeping. There was no flight crew to speak with at the time and we didn't want to disturb anyone considering we were... well.. still alive.

Upon return from our trip I called Air Transat and spoke with Customer Relations who had the flight records pulled. There was no report of the incident. I cannot believe any regulation could possibly allow for two planes to fly so close to one another under any circumstances. I need to understand what happened here, the distance planes must remain from one another and why such an occurrence would not be logged.

The plane was not marked. It was not a Lear jet but a small, dark, unmarked passenger plane. Thank you for your help. Obviously I am still very disturbed by this event.

• How recently was this? It would help to have an exact date and flight number. Jul 26 '17 at 20:38
• Did you feel any turbulence after the other plane moved away?
– mins
Jul 26 '17 at 22:04
• If you provide a date/flight number often we can locate the specific aircraft you saw and see exactly how close it was. Jul 27 '17 at 1:13
• we didn't want to disturb anyone considering we were... well.. still alive. so, would you have considered disturbing anyone if you weren't still alive? ;-) Jul 27 '17 at 20:26
• A few examples of what 1000 ft separation looks like: 1 2 3 4 5 Jul 29 '17 at 4:13

It was only very slightly higher than our plane by mere meters.

You think you saw a plane at the exact same vertical level as yours, when in fact it was at least 300 metres (1000 feet) above or below your level. It can be very hard to judge differences in level for the untrained eye, especially over the ocean and at night when there is nothing to reference for scale. Your confusion is understandable, but there was absolutely no safety risk here.

The minimum vertical separation between two aircraft is 1000 feet, and the plane you saw was - with 100% certainty - at least 1000 feet above or below you. This is ensured by air traffic control and a ton of complex regulations and rules governing air traffic.

When over the ocean, where no radar coverage exists, if two planes are flying at the same level, they will either be at least 10 minutes flying time apart (if following the same route), or 25 nautical miles (about 50 kilometres) if flying on parallel routes. Over land where there is radar, two aircraft at the same level will generally be at least 5 nautical miles apart (10 kilometres). But when 1000 feet or more vertical separation exist, aircraft can cross each others paths without any risk. It is perfectly normal for aircraft to fly this close, but passengers seem to take notice more often over the ocean, probably because there is nothing else to look at.

There was no risk of collision during your flight. If, for some inexplicable reason, the standard separation described above was violated, pilots and air traffic controllers would have responded immediately to establish separation again, long before any sort of collision risk would occur. There is very reliable technology in place to warn pilots and air traffic controllers if aircraft are getting too close, and they are trained to react to such warnings without delay.

• If you don't think it's difficult to estimate distances, go outside during daylight and try to estimate how tall the nearest radio or (cell) phone tower or windmill is. Then using the references in the answers here go look up how tall it actually is. For the untrained eye, estimating height/distance is quite difficult. Jul 26 '17 at 19:42
• Since the OP is 'still very disturbed by this event' is it worth mentioning TCAS in the penultimate paragraph? en.wikipedia.org/wiki/Traffic_collision_avoidance_system Jul 26 '17 at 21:56
• @DaveGremlin Since OP clearly knows quite little about aviation, I wanted to keep the answer simplistic. I added a general note about technology instead. Jul 27 '17 at 5:47
• @mickburkejnr No it wouldn't. TCAS provides a warning to the pilots, it does not control the plane. It's up to the pilots to follow the recommended action of TCAS Jul 27 '17 at 9:17
• @canadianer I guess I'm 99,99% sure and decided to round up to 100 ;) Jul 27 '17 at 19:58

This is an excellent question, and helps illuminate several important aspects of the way perception works in aviation. But first...

# Any near-miss or other air-proximity incident would have been reported

If there had been any risk of a collision, it would have been reported, unless somehow (and this is astoundingly unlikely) the entire flight crew of both aircraft were entirely piratical.

But flight crews are not swashbuckling daredevils; they do everything by the book, because that's what keeps them alive to fly another day, and because everything in their training and professional discipline is about following procedures correctly.

Even if there were not a risk of collision, but the merest compromising of safety, the incident would have been reported.

# What you saw

But let's consider what you saw. Our binocular vision is excellent; it can track movements, and judge speeds, sizes, angles and relative position with amazing accuracy and precise judgements, which is why we human beings can do things like return Roger Federer's serve and shoot apples off people's heads with arrows.

However, our binocular vision can do all these wonderful things only when it's operating in a rich and reliable frame of reference. In an aircraft, at night, over the ocean, we lose almost all of the reference that allows us to make accurate and precise judgements. You could be looking at a small aircraft a short distance away or a huge one far away, and you'll have little chance of knowing which. It gets even harder to judge speeds. Relative position can also be very hard to judge.

However compelling the feeling that you know what you saw, it's simply not reliable.

In this case, the compelling evidence of your perception led you to conclude that there was a risk. In aviation, we usually hear about the opposite case: where trust in compelling-but-unreliable perception causes a pilot to underestimate or even just ignore a risk. Pilots learn that perception is not a reliable friend, and that senses that work very well when playing tennis or with crossbows are hopelessly inadequate in the air.

Instead, they place their trust in a battery of proven instruments, from simple things like gyroscopes and compasses that have been key to flight safety for a century or more, to advanced radar systems, and industry wide systems and processes. They are what keeps flight safe.

It's hard to ignore the clamouring of one's alarmed perceptions, but the fact is that when you climb on board on airliner you should accept that nothing you will see or feel for the next few hours is a good indication of what it really is.

# What if the pilot took evasive action?

You mention that your flight changed course slightly, with the implication that this may have been evasive action.

If your pilot had been required to take some sort of emergency last-second evasive action to avoid a collision, not one of your sleeping fellow passengers would have remained asleep: those of you not strapped down would have been hurled to the ceiling or slammed to the floor and your stomach would be heaving in a different direction from the rest of you.

You'd really know about it, in other words. And then there would be plenty of aviation industry paperwork to follow.

However, that would be an extraordinarily rare thing. Evasive action, in the still very rare cases when it is required, is taken well before it's necessary to hurl the plane around in the sky to avoid something at close quarters.

Aircraft of the kind you're talking about are fitted with equipment that monitors their sector of the sky (for miles and miles around) for other aircraft, and will advise each flight crew to climb or descend appropriately so that there is not a proximity risk.

Either way, it certainly doesn't sound like you experienced evasive action at that moment.

# On being disturbed

It's completely natural to be disturbed by what you saw - but if you accept that aviation is completely unnatural as far as our senses and perception go, and they just don't work very well at making good judgements in that context, that might help.

Even if you can't let go of the reports of your perceptions (I can't), knowing that the reality is otherwise can still help.

In short: anyone else looking out of the window may well have been alarmed by what they saw. Once that moment of alarm has passed though, what you know can help prevent it from continuing to disturb.

• Very good! "It's hard to ignore the clamouring of one's alarmed perceptions". Perception (auditive, visual...) is pre-processed by filters in the brain. Those filters deliver a modified information to other parts of the brain which need it. Those filters are built during our childhood from experience. This is why we'll keep affirming squares A and B are not of the same color in this checkerboard, while actually they are. Here, our brain trusts the filters (A is black, B is white), and cannot retrieve the original colors.
– mins
Jul 26 '17 at 23:39
• Good answer, except the bit about evasive action. In situation like this, evasive action would be taken based on TCAS advisory and those leave enough time so there is no need for drastic manoeuvres. Few notice response to TCAS RA when it happens. However, there is a way to tell this was not an evasive manoeuvre: evasive manoeuvres are always vertical. Both because TCAS has accurate method for comparing altitudes, but not for solving how the horizontal paths intersect, and because climb or descent are initiated faster than a turn. Jul 27 '17 at 17:54
• @JanHudec Edited to clarify that I was describing a hypothetical WTF is that? emergency manoeuvre, not TCAS-initiated evasive action. Jul 27 '17 at 20:24
• @corsiKa, well, I would believe the OP that some manoeuvre did occur—it just was a completely coincidental course correction (the North Atlantic tracks are not straight) and not evasive manoeuvre. Jul 28 '17 at 13:01
• I know this sort of comment is frowned upon, but I think this is among the best answers ever written on this entire site! I liked specifically that you discuss aviation from the POV of both the experienced pilot and the inexperienced passenger - while the latter believes in their "compelling-but-unreliable perception", the former knows to trust the instruments and ATC, even at the expense of disregarding what they see out the window - especially above FL100 Feb 3 '19 at 8:38

A "small" transatlantic plane probably isn't very small -- it was probably further away than you think. Compare these silhouettes (can't embed due to image rights, also can't vouch for accuracy):

(Just look at the first and last for a quick indication of what I mean)

In brackets are the lengths (ranges where I don't know which model the silhouette represents). A trained eye could identify the model, then make a decent estimate of the distance. A untrained eye, with no distance cues, would really struggle to rank them in order of size, despite the factor of around 2 difference in length betwen a 737-500 and the widebodies. So if you're used to seeing 737s fairly close at your regional airport and then see a 777 fly past, it will seem half as far away as you think based on size alone -- and in an empty sky things seem closer than they are.

• Right, shapes are very similar, even a Messerschmitt 262 doesn't look that different than an A350. So a distant 747 will look like a near Learjet. On Bush's famous Thankgsiving 2003 trip to Baghdad, another pilot radioed that he thought he saw Air Force One, they radioed back saying "we're just a Learjet", and the pilots bought it. So it really is hard to tell. Jul 28 '17 at 20:30
• @Harper an extreme example but a good one. About 20 years ago I had a little experience with silhouettes of aircraft on approach (almost head-on) but never got to be a real expert. Even then I had the advantages of familiar conditions and ground cues Jul 31 '17 at 10:50

How close can planes fly to each other over the ocean?

The visual perspective from the cockpit or passenger window often makes another airplane look much closer than it actually is. As previously stated, the minimum vertical separation during enroute flight is 1000 feet. But, as I mention, when aircraft are somewhat close laterally, but separated vertically by a 1000 ft, the visual perception can result in the aircraft looking much closer.

Over the ocean, beyond radar coverage, the vertical separation minimum can be a little as 1000 ft. With respect to the lateral and longitudinal separation, the separation criteria are a bit more complex.

Depending on a variety of circumstances, which include, the equipment capabilities of the aircraft involved, the speed of the aircraft involved, and where the aircraft are operating, the separation minimums can be:

1. As low as 5 minutes longitudinal separation (leading aircraft is much faster than trailing aircraft)[ref: [JO 7110.65W], Chapter 8. Offshore/Oceanic Procedures]]. OR

2. As low as 30 NM longitudinal separation (both aircraft are appropriately equipped with certain Performance Based Nav equipment, RNP-4).ref: [North Atlantic Separation -NBAA document]

3. As low as 30 NM lateral separation (both aircraft are appropriately equipped with certain Performance Based Nav equipment, RNP-4).ref: [North Atlantic Separation -NBAA document].

Oceanic (non-radar) separation can be quite complex. But as aircraft equipment and satellite technology increase in sophistication, separation minimums will likely continue to safely decrease, thereby increasing ATC/Airspace system capacity.

Something to note here would be that modern aircraft contain a Traffic collision avoidance system (TCAS). So pilots in both aircraft would have been notified that they were on a collision and one would be told to change altitude. The system is so reliable that in the event TCAS and ATC orders conflict, TCAS is to be obeyed (due to a midair collision over Germany in 2002)

When an [Resolution advisory (RA)] is issued, pilots are expected to respond immediately to the RA unless doing so would jeopardize the safe operation of the flight. This means that aircraft will at times have to manoeuver contrary to ATC instructions or disregard ATC instructions. In these cases, the controller is no longer responsible for separation of the aircraft involved in the RA until the conflict is terminated.

So at the bare minimum, your aircraft were separated by enough distance not to set off the alarm (600 ft)

Aircraft operating within the NAT High Level Airspace(Atlantic Ocean), from Flight level 285 and 420 inclusive are required to conform to the Reduced Vertical Separation Minimum (RVSM) equipment requirements.

This means that they are required to have an autopilot capable of maintaining 1000ft vertical separation and have equipment to alert them if the aircraft strays from the assigned level. In addition in order to operate in RVSM airspace you must have a Traffic Collision and Avoidance System (TCAS) fitted.

So in the situation you describe the autopilot would have been on maintaining the cleared level and the TCAS would of been showing the aircraft that passed you on the display in the flight deck. If the other aircraft strayed from its cleared level (1000ft clear) then the TCAS would of alerted the crew and if required provide a resolution advisory to climb or descend. This would be the case in both aircraft involved.