YOUR COMPASS DOESN'T KNOW TRUE NORTH
But the Sun Can Tell You How Wrong It Is
Eric Kelley · EMNAV Field Notes · October 6, 2026 · Historical Navigation
Most people who carry a compass trust the needle. They shouldn't, at least not blindly.
A magnetic compass does not point toward the geographic North Pole. It aligns itself with Earth's local magnetic field. The angular difference between true north and magnetic north is magnetic declination, and depending on where you are, that difference can be large enough to matter considerably over a long movement.
Normally, this isn't a problem. A current map or other reliable source gives you the magnetic declination, and you apply the appropriate correction.
But suppose you don't have that information. Maybe you don't have a map. Maybe the map is decades old. Maybe you're operating in an unfamiliar area and have no idea what the local declination should be.
Can you determine the error yourself?
You can.
One method is surprisingly elegant: observe a celestial body with your compass when it rises, observe it again when it sets, and use the two azimuths to determine how your magnetic reference differs from true north.
There is an important limitation for the land navigator.
This method assumes that you can observe the celestial body at both the eastern and western horizons under reasonably comparable conditions. At sea, or on open terrain with clear horizons, that may be practical. In mountainous, forested, or broken terrain, one or both horizons may be obscured or elevated by terrain. A ridge may block the Sun in one direction while the opposite horizon is much lower. Vegetation can do the same. If the morning and evening observations are made when the celestial body is at substantially different altitudes, the simple rise-and-set method may introduce additional error.
This does not make the historical technique invalid. It means the navigator must understand the conditions under which it works. A clear view of both sides of the horizon makes the method much more useful. In rough terrain, another method of establishing comparable observations may be preferable.
This isn't a new survival trick. The U.S. Army was teaching the method more than 80 years ago.
A Forgotten Army Navigation Technique
In 1941, the War Department published FM 21-26, Advanced Map and Aerial Photograph Reading. This was not an introductory map-reading pamphlet. The manual states that it was intended for military personnel already familiar with basic map reading or those requiring more detailed instruction.
Buried in its discussion of field-expedient direction finding is a technique rarely encountered in modern basic land-navigation instruction.
The navigator observes the magnetic azimuth of the Sun, a planet, or a bright star as it rises and again as it sets. The two magnetic azimuths are added together and compared with 360°. One-half the difference provides the compass declination under the convention described in the manual.
No GPS is required. No electronic magnetometer is required. For this particular method, you don't even need to calculate the celestial body's actual azimuth.
The technique exploits something far more fundamental: the geometry of the sky.
The Geometry Behind the Trick
Imagine, for a simplified example, that a celestial body rises at a true azimuth of 80° and later sets at a true azimuth of 280°.
Add the two:
80° + 280° = 360°
The rising and setting directions are symmetrical about the observer's meridian.
Now suppose your magnetic compass is displaced 10° from true direction. Instead of measuring 80° and 280°, imagine that it reads 70° at rising and 270° at setting.
Add those observations:
70° + 270° = 340°
The expected symmetrical total is 360°, so:
360° − 340° = 20°
Divide the difference by two:
20° ÷ 2 = 10°
The paired observations have revealed a 10° difference between the magnetic reference and the true reference.
FM 21-26 states the convention simply: if the sum of the observed magnetic azimuths is less than 360°, the compass declination is east; if the sum is greater than 360°, it is west.
The important point isn't memorizing the arithmetic. It's understanding what just happened.
The sky became your independent directional reference.
How to Do It in the Field
Choose a celestial body that can be observed close to both rising and setting. The Sun is the obvious choice, although the 1941 manual specifically states that the method can also be performed with a planet or bright star.
At rising, move away from obvious sources of magnetic interference and select a location with a clear horizon. Observe the body as close to the horizon as conditions reasonably permit and measure its magnetic azimuth carefully. Record the azimuth and, as good field practice, the observation location, date and approximate time.
At setting, make a second observation of the same celestial body under comparable conditions. Comparable here includes a similar altitude above the horizon, not merely a second look at the end of the day. A clear eastern horizon and a ridge to the west are not equivalent observations. Again record its magnetic azimuth.
Now add the two magnetic azimuths.
If the sum is less than 360°, subtract the sum from 360°. If the sum is greater than 360°, subtract 360° from the sum. Divide the resulting difference by two.
Using the convention in FM 21-26:
Sum less than 360° = East
Sum greater than 360° = West
Worked Example
Suppose the Sun's magnetic azimuth at rising is:
110°
At setting, you measure:
270°
Add them:
110° + 270° = 380°
The difference from 360° is:
380° − 360° = 20°
Divide by two:
20° ÷ 2 = 10°
Because the sum is greater than 360°, the procedure produces:
10° West
This isn't an example invented for this article. FM 21-26 illustrates the method with those same 110° and 270° magnetic observations and arrives at 10° west.
Why Not Just Call Sunrise East?
Because the Sun does not normally rise due east or set due west.
The familiar rule that the Sun “rises in the east and sets in the west” is a useful generalization, not a precision navigation method. The Sun's actual rising and setting points move along the horizon through the year, and their positions also depend upon the observer's latitude.
The paired-observation method is much more sophisticated because it doesn't require you to pretend sunrise is 090° and sunset is 270°. Instead, it exploits the relationship between the rising and setting directions.
That's what makes the technique interesting.
It isn't a crude directional guess. It's a geometric measurement.
Your Compass May Not Be the Only Problem
There is an important qualification.
If your astronomical observation and compass disagree by 10°, you should not immediately conclude that you've measured regional magnetic declination to exactly 10°.
You have established that your compass measurement differs from the astronomical true-direction reference under the conditions of the observation.
Why the distinction?
Because a magnetic compass can also be influenced by its immediate surroundings.
FM 21-26 warned that iron and electrical fields could deflect the compass needle and specifically cautioned navigators about weapons, helmets, power lines, artillery, telegraph wires and barbed wire.
The equipment around us has changed considerably since 1941, but magnetism hasn't.
Vehicles, steel structures, tools, weapons, electrical equipment, speakers, magnets and other equipment can affect a magnetic compass. Before concluding that you've discovered the magnetic characteristics of an entire region, make sure you haven't actually discovered the magnetic characteristics of your truck.
Move away from suspected interference and repeat the observation whenever practical.
This leads to an important EMNAV principle:
What Are You Actually Measuring?
The rising and setting method can be used to make a field estimate of local magnetic declination when current declination information is unavailable.
But it is important to understand what the observation actually compares.
The celestial observation provides an astronomical reference to true direction. The compass provides a magnetic direction measured by a real instrument, at a particular location, under the conditions that exist at the time of the observation.
In an ideal observation, made with a properly functioning compass in a location free of local magnetic interference, the difference between those references provides a practical field estimate of local magnetic declination.
In actual field conditions, however, the observed difference can contain several components. These can include local magnetic declination, compass or instrument error, nearby magnetic interference, and observational error.
Observed difference
≈ local magnetic declination
+ compass and local magnetic effects
+ observational error
This is a conceptual field relationship, not a rigorous error equation. A single observation does not automatically separate those components.
This distinction gives the technique two useful applications.
If local magnetic declination is unknown, the method can provide a practical field estimate of it.
If local magnetic declination is already known from a reliable current source, the astronomical observation can be compared with the expected value. A significant difference may indicate that the compass, the immediate environment, or the observation itself is introducing additional error.
This is an important distinction for an advanced navigator.
The celestial body provides an independent true direction reference. The compass provides a magnetic measurement. Understanding the difference between the two allows the navigator to evaluate the entire directional system rather than simply trusting the compass needle.
A Second Check: Polaris
There is another way to make the sky check your compass.
At night in the Northern Hemisphere, Polaris provides a natural reference near the north celestial pole. FM 21-26 describes finding Polaris with the pointer stars of the Big Dipper and also identifies Cassiopeia on the opposite side.
The manual goes further. It explains that under favorable geometry, an observation of Polaris can establish true north and allow the navigator to determine compass declination approximately to the resolution of the compass.
There is a refinement that matters for advanced navigation: Polaris is not exactly at the north celestial pole.
For ordinary field orientation it provides an excellent north reference. For higher precision, its displacement from the celestial pole and its position around that pole should be accounted for.
That subject deserves its own Field Notes article.
For now, the important concept is the same: establish an astronomical true reference, compare the compass against it, and investigate the difference.
Magnetic Declination Doesn't Stand Still
There's another reason an independent check can be valuable.
Magnetic declination changes with time.
FM 21-26 was already warning navigators about this in 1941. The manual notes that magnetic declination at a location is subject to change and says standard maps should show the relationship between magnetic and true north for a stated date, along with the annual magnetic change.
That's an important detail.
A 30-year-old topographic map may still describe the terrain perfectly well while its printed magnetic information has become outdated.
The terrain hasn't moved.
Magnetic north has.
A navigator capable of independently establishing true direction has a way to detect that problem instead of unknowingly building it into every azimuth.
A Few Degrees Become a Big Error
Why care about a small compass error?
Because angular error accumulates into positional error with distance.
Suppose you intend to travel 5,000 meters on a straight course but your actual heading is 10° away from the intended line.
Your approximate lateral displacement is:
5,000 × sin(10°) ≈ 868 meters
A 5° error over the same distance produces approximately:
5,000 × sin(5°) ≈ 436 meters
Even a 2° error produces approximately:
5,000 × sin(2°) ≈ 175 meters
You can have an excellent pace count, accurately estimate your distance and execute the movement exactly as planned, and still finish hundreds of meters from your intended line because your directional reference was wrong.
Small angular errors become large positional errors when you give them enough distance.
This is one of the differences between simply operating a compass and understanding navigation.
The Real Lesson Isn't the Trick
The most valuable part of this forgotten technique isn't the formula.
It's the way of thinking behind it.
A compass does not define north.
North exists independently of your compass.
Earth's rotational axis provides the fundamental geographic reference. Celestial bodies provide observable relationships to that reference. Earth's magnetic field provides another reference. Maps and grids provide still others.
Once the navigator understands how those systems relate to one another, one system can be used to check another.
That is considerably more powerful than simply following a needle.
A modern navigator can become dependent upon instruments precisely because they work so well. GPS displays a position. A digital compass displays a heading. Software performs the transformations. The underlying geometry disappears behind the screen.
But the geometry is still there.
More than eighty years ago, Army navigators were being taught to use the sky to determine whether their magnetic reference agreed with true north.
The technique didn't stop working.
We mostly stopped teaching it.
Sources & Further Reading
This manual covers magnetic and grid declination, local magnetic attraction, field-expedient determination of direction, celestial methods, orientation, advanced map reading and aerial-photograph navigation. It was intended for personnel already familiar with basic map-reading concepts.
EMNAV FIELD NOTES
Read the sky. Rule the ground.