There are two types of returns that show up on an air traffic controller’s radar screen:

  • Secondary returns are not, strictly-speaking, radar “returns” at all, but, rather, signals automatically broadcast from an aircraft’s transponder, containing encoded location, altitude, airspeed, identification, flightplan, aircraft type, etc., data retrieved from the aircraft’s onboard instruments. These are highly valuable, but, obviously, can only be used with aircraft equipped with a functional transponder which have said transponder turned on and transmitting non-bogus data.

  • Primary returns, in contrast, are true radar returns – direct reflections of the radar’s beam from the surface of an aircraft, bird, cloud, piece of debris, missile, tree, UFO, balloon, hailstone, or whatever else happens to be in the air at that moment. As they merely require that an object be located where the radar beam can reach it, they are of great use in tracking aircraft with nonfunctional transponders (due to, for instance, a general electrical failure or simply a faulty transponder), aircraft in combat zones (for whom broadcasting an identification signal would be a great way to get shot down), pieces of aircraft, flocks of birds, or whatever other non-transponder-equipped objects one has a desire to track.

One common limitation given for primary radar returns is that they provide no altitude information, but only positional information. But this makes no sense, as determining the position of a target requires knowing its elevation angle, azimuth, and distance relative to the radar installation (without knowing the target’s distance, it could be located anywhere along a line extending from the radar’s location out to infinity; without knowing its elevation angle, it could be located [within the altitude limitations of the object generating the target] anywhere along a circular arc extending from the horizon to the zenith at the specified distance from the radar; without knowing its azimuth, it could be located anywhere along a horizontal circle centered in the sky directly above the radar), and, if the elevation angle, azimuth, and distance of the target are all known, that also pins down the target’s altitude – not just its location. Besides, military radars can and do provide altitude information for primary targets (they would be useless otherwise, as intercepting an enemy aircraft requires knowing both its position and its altitude, and enemy aircraft are unlikely to oblige a radar’s request to provide a transponder beacon signal that would aid immensely in shooting them down), which has proven valuable numerous times; for instance, the accident investigation that eventually produced the NTSB’s very first AAR used data from a military air-defence radar to determine that a 727 that crashed into Lake Michigan had descended steadily into the water without levelling off, instead of suffering an uncontrolled excursion from level flight, while, more recently, primary returns received at several military radar sites in Massachusetts showed that EgyptAir Flight 990 pulled out of its initial dive before making a second and final plunge.

So what prevents civilian ATC radars from displaying altitude information for primary targets?

  • 1
    $\begingroup$ Also note that ATC uses more accurate altitude information than position information. If primary radar gives position to within 1/2 mile, that's good, but if it has the same accuracy for altitude, that's not good enough for separation. $\endgroup$
    – Thomas
    Dec 7 '18 at 17:53
  • $\begingroup$ because math... $\endgroup$
    – ryan1618
    Dec 8 '18 at 17:01

Civilian Primary Surveillance Radar (PSR) does not provide elevation angle, and thus no altitude information can be displayed. The location is therefor not accurate either, but that doesn't matter for ATC purposes.

Civil ATC does not need altitude or height information to provide traffic separation when using PSR. They simple make sure the dots on the screen don't collide. For traffic separation you don't need to know the exact position of the aircraft; as long as the plots are separated on the 2D screen, the will be separated in the 3D world. Typical PSR displays are based on a flat world assumption; they only plot azimuth and range. As long as the plots are separated by 5NM, no matter what the altitude difference between the aircraft is, they are safely separated for ATC purposes.

As you correctly observed, if you would want to determine the exact 3D position of the aircraft based on a primary radar, you need azimuth, range and elevation angle. The PSRs used by civilian air traffic control don't measure the elevation angle, because this would require more sophisticated radar hardware and would therefore be more expensive.

The added value of the elevation angle is that the exact position including the height of the aircraft could be measured. However, that is of limited value. In air traffic control, vertical separation is not based on geometric altitude but on barometric altitude or flight level. Using elevation angle, the primary radar would give the geometric height. However, the barometric height showed in the cockpit to the pilot can easily be several hundreds or thousand feet off from the geometric height. So for the communication between ATC and pilots, geometric height measured by primary radar is mostly useless.

Aircraft are vertically separated by putting them on different altitudes/ flight levels, based on the baro altimeter in the aircraft. This altitude is transmitted to the secondary radar which allows display of that altitude and separation based on altitude difference. This is the basis for communication of vertical position information between ATC and pilots.

In short, the costs of adding elevation angle measurement to civilian primary ATC radar outweighs the benefits.

  • $\begingroup$ Aircraft are separated by putting them on different barometric altitudes/flight levels... above the transition altitude (usually 18 kft). Which still leaves quite a lot of traffic below there... $\endgroup$
    – Vikki
    Dec 8 '18 at 3:47
  • $\begingroup$ @Sean I don't understand your comment. Below the transition altitude vertical separation is achieved by using different altitudes. $\endgroup$
    – DeltaLima
    Dec 8 '18 at 9:28

without knowing its elevation angle, it could be located anywhere along a circular arc extending from the horizon to the zenith at the specified distance from the radar

Well, not quite. Aircraft are generally much more restricted in their vertical positioning. If there's a return from 30 miles away, that aircraft can't be overhead. (And radars normally have a maximum elevation to the beam, often around 70 degrees).

This document suggests that (at least as of 1989) determining elevation angle was very difficult for ATC primary radar. It was possible by cross-correlating information from multiple feeds, but that this was not routinely done. The document focuses on how useful that would be for helping to remove ground clutter.

Even if some elevation information were available, that data might not be presented to controllers unless it were of a useful precision. If the precision were worse than say 1500 feet, would that be useful?

Besides, military radars can and do provide altitude information for primary targets

Radartutorial suggests that vertical discrimination (or 3D radar) requires extra equipment and is therefore more expensive. Since ATC can get this information by secondary radar, the expense of gathering it via a 3D primary system is avoided.

Besides cost, older 3D radar would scan a region more slowly than a 2D radar.

  • $\begingroup$ Yes, I realise that aircraft have limits on how high they can fly, but a) different aircraft can fly at different heights, and b) that still leaves a large portion of the arc within the aircraft's flight envelope, so the general idea is still valid. $\endgroup$
    – Vikki
    Dec 5 '18 at 22:23

Primary (2D) ATC radar provides azimuth and slant distance but not elevation. If you use the slant distance as horizontal distance, it will be somewhat inaccurate, but since aircraft don't fly all that high (and shouldn't be above 10kft MSL at all without a working transponder), the error is minimal--and separation standards account for it.

Adding elevation scan (3D) would increase cost and slow scan rates yet add nothing when most aircraft have transponders anyway, so it's more efficient to live with the limitations.

  • 3
    $\begingroup$ Not all readers may realize that radar distance is diagonal without it being pointed out, and that is crucial to understanding it as a potential source of position error in the ATC context. $\endgroup$
    – StephenS
    Dec 5 '18 at 17:52

The limitations of ATC radar are simple: they are not designed to provide more than distance and heading for primary returns.

PAR (precision approach radar) is an example of a radar system which provides altitude information, and is configured to allow the controller to provide vertical approach guidance information.

There are many radar systems which provide elevation information, for example on some SAM (surface to air missile) systems. It's possible, but someone has not spent the additional taxpayer funds for ATC purposes.


The new types on the market can do this. See for instance Thales STAR-NG (altimetry option), Hensoldt (previously Airbus) ASR-NG, ...

You can also upgrade existing radars, such as the FAA ASR-9, as you can for instance see on the website of Intersoft.

There are however a number of limitations in performance, and more importantly, civil ATC procedures are not designed to PSR derived altitudes. But even if not used operationally, it still provides a benefit in tracking and clutter rejection.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.