The primary means of determining heading in most small aircraft is the directional gyro, as this instrument is unaffected by the magnetic dip errors that cause a magnetic compass to give the wrong heading when at a nonzero angle of bank. However, due to the rotation of the earth under the aircraft, the aircraft's movement over the curved surface of the earth, and friction within the gyro assembly, the directional gyro has to be periodically realigned, using the magnetic compass as a reference.

All well and good, but there are large parts of the earth where a magnetic compass cannot be trusted, because the horizontal component of the earth's magnetic field is too weak or erratically variable. Examples of such areas of magnetic unreliability include the areas around the earth's magnetic poles (for instance, northern Canada), where the magnetic-field lines go almost straight up-and-down rather than side-to-side, and areas with large localized changes in magnetic deviation (for instance, western Lake Superior, where all the iron in the ground causes magnetic compasses to read almost 90° off in places). Additionally, these areas go by true heading rather than magnetic (and the two differ considerably at high latitudes), so you can't even extrapolate a magnetic heading that you obtained before the compass became unreliable - you need to realign the gyro to true north as soon as you enter the area of magnetic unreliability.

Big airliners have GPS and IRS to give them an accurate measure of true heading and keep the directional gyro honest, but, if you're flying a small general-aviation plane not equipped with either, how do you correct the directional gyro's drift if you're somewhere your magnetic compass isn't reliable (especially if you're IFR and can't use ground landmarks to find where north is)? If you're departing from within an area of magnetic unreliability, how do you align the directional gyro in the first place; if you're just entering a magnetically-unreliable area, how do you realign the gyro from magnetic to true?


1 Answer 1


You initially align it while on the ground, using a known reference like the runway. While in flight, you'll need to use an astrocompass or sextant, if you have one and know how to use it and can see the sun or stars.

If you're a bush pilot chugging along VFR in your Beaver, you can make an eyeball alignment by judging heading against the apparent alignment of the airplane itself with landmarks. For example, there may be a feature like a river or shoreline along your route that has a straight section with a true track you can measure on your chart, so when you come to it you can align yourself with the feature below you and set your DG. If you're flying along the Alaska Highway in your 172, it's easy to keep the DG aligned using the track of the highway below you (If you're flying your 172 to Alaska, you'll be sticking close to the highway anyway if you have half a brain).

Aside from tricks like that, there's not much you can do. If you've measured the rate of precession of your DG, you could just apply an equivalent correction at a suitable interval. For example, if you know that your DG precesses 5 degrees in 20 minutes, you just apply a 5 degree counter adjustment every 20 minutes. It'll keep the DG roughly in the ballpark, good enough for VFR flying, but in any case a lot of the time you are going to be navigating by pilotage (landmark to landmark). My bush flying wasn't in the far north, but still I rarely flew precise headings, just found my way from lake to lake, pointing the nose in the general direction I wanted to go and tracing my position/progress on a 50,000:1 scale topographical map, sometimes at 300 ft agl under a 500 ft cloud deck.

I wouldn't want to be doing an IFR flight on DG heading information alone unless I could keep the DG in alignment with an astrocompass/sextant (where you're going to need to get above cloud layers).

Up until about 15 years ago, knowledge of how to take a star or sun shot with a sextant and work out position and heading was a requirement for the Canadian ATPL license, but it was dropped.

  • $\begingroup$ What do you do when aligning the directional gyro before departure from a bush site without an obvious heading reference (given that it's probably easier to identify things on your map from the air than from the ground)? Would that be a situation requiring the sextant in order to figure out which way you're going? $\endgroup$
    – Vikki
    Dec 15, 2021 at 2:09
  • $\begingroup$ Not that I'm complaining; great answer! $\endgroup$
    – Vikki
    Dec 15, 2021 at 2:11
  • 1
    $\begingroup$ Thing about bush operations, not many of them are operated in conditions requiring accurate compass headings, as they are mostly VFR/map navigation (or IFR as in I Follow Roads). The ones requiring proper navigation are flown with equipment not relying on gyros (well other than solid state gyros). $\endgroup$
    – Jpe61
    Dec 15, 2021 at 12:27
  • 1
    $\begingroup$ If you're on a lake you will just identify shore line land marks, work out the true direction between them on your chart, taxi adjacent to one, then point the nose at the other, and set the DG. The times I needed a heading reference was when scud running in low vis down on the deck running from lake to lake, and I would leave one lake on a best guess particular heading and hold it until I hit the next lake, maybe 3-5 minutes later. I was flying where the compass worked fine. $\endgroup$
    – John K
    Dec 15, 2021 at 13:18
  • 1
    $\begingroup$ At the close distance between them with the size of the lakes you would hit it if you were +/- 15-20 deg as long as the lakes were too far apart. Sometimes if it meant crossing an area with no features that went for 20 miles, or detouring to follow a chain of lakes or a river, I would detour to follow the chain of lakes. Flying at 3-500 ft agl under a cloud deck with nothing but forest and rocks under you, it is very easy to get lost if the lake you are expecting to loom out of the murk doesn't appear when you expect it, so you tend to lean on following features rather than dead reckoning. $\endgroup$
    – John K
    Dec 15, 2021 at 21:28

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.