The Forgotten History of the Compass Needle's Magnetic Declination

The schoolroom version of the compass is that the needle points north. The full version is that the needle points to magnetic north, which is not the same place as geographic north, which itself is not in the same place as the rotational axis of the Earth, which is not in the same place as the celestial pole, which itself moves over millennia due to precession. The error between magnetic north and true north is called magnetic declination, it varies by location on Earth, it varies over time at any given location, and figuring out what it actually was at any given place and time was one of the central scientific problems of the seventeenth, eighteenth, and nineteenth centuries.

The story is interesting for three reasons. First, the discovery sequence is unusually clean: a single observable phenomenon (the compass needle does not point to the pole) revealed an entire branch of physics (geomagnetism) and an entire branch of geophysics (the Earth's iron core). Second, the practical consequences were enormous: every long-distance navigation problem from 1500 through 1900 depended on knowing the declination, and the project to map it generated some of the longest-running scientific expeditions in history. Third, the modern relevance is large: the magnetic field is still moving, modern aviation still depends on knowing where it is, and the declination has changed faster in the last 40 years than at any time in the last 400.

The discovery

The fact that the compass needle does not point exactly to true north was probably known to skilled navigators in the Mediterranean by the 13th or 14th century, though it is not well documented in surviving texts. The first written acknowledgment in Europe is in Christopher Columbus's logbooks from his 1492 voyage, where he records that the needle deviated from the polar star and that the deviation changed as the ship moved westward. He did not have a theory; he just noted the observation and adjusted his navigation accordingly.

The systematic study began with William Gilbert's De Magnete in 1600, which proposed that the Earth itself was a giant magnet with poles offset from the rotational axis. Gilbert's argument was based on the behavior of small spherical magnets ("terrellas") that he had constructed in his London laboratory. The terrella showed the same kind of declination pattern as the Earth, supporting the idea that the Earth was, in effect, one of these spherical magnets at a planetary scale. Gilbert is also responsible for the modern terminology: he called the deviation "variation of the compass," which evolved into the term magnetic declination as the science matured.

The next step was to measure the variation at as many points as possible to produce a map. This was an enormous undertaking. The variation was different at every location, so measurements had to be taken in many places. The variation also changed over time, so old measurements became increasingly inaccurate. And the measurement itself required reasonably good astronomical observation (to determine true north) plus a reasonably good compass (to determine magnetic north), neither of which was routinely available in the 17th century.

The Halley expedition

The first significant attempt to produce a global declination map was Edmond Halley's voyages from 1698 to 1700 on HMS Paramore, commissioned by the Royal Navy. Halley, better known today for the comet that bears his name, was also one of the most important geophysicists of his era. The expedition crossed the Atlantic twice, taking declination measurements at hundreds of locations from Newfoundland to South Africa.

Halley published the resulting map in 1701 as the first known chart of magnetic declination. The map used a then-novel technique called isogonic lines: contour lines connecting points of equal declination. The same technique would later be used for elevation contours, atmospheric pressure isobars, and dozens of other geographic data types, and Halley is sometimes credited with inventing the contour-mapping convention itself.

The Halley map was a navigational tool of considerable practical value. Ships' navigators could look up the declination for their estimated location and correct their compass readings accordingly, improving the accuracy of dead-reckoning navigation. The map was reprinted and revised throughout the 18th century as new measurements accumulated.

The secular variation

The bigger puzzle that emerged from comparing declination measurements taken at different times in the same place was that the value changed. London's declination was 11 degrees east of north in 1580, zero in about 1660, 24 degrees west in 1820, and is currently about 1 degree east. The total range of variation at one location over 400 years is about 35 degrees, which is enormous. The pattern of change came to be called the secular variation, and explaining it was the next major problem.

Halley himself proposed an explanation in 1692: he suggested that the Earth had a solid outer shell magnetized one way, a solid inner sphere magnetized differently, and a fluid layer between them. The relative rotation of the inner and outer parts would produce a slowly-changing magnetic field at the surface. The model was hand-wavy but the basic idea (the magnetic field is generated by the deep interior of the Earth, and changes in the interior produce changes at the surface) was substantially correct.

The modern explanation, developed mostly in the 20th century, is that the Earth's outer core is a layer of liquid iron about 2200 km thick, that convection in this iron generates electric currents, that the currents produce the magnetic field, and that the convection patterns shift over time, causing the field to shift. The full theory (called the geodynamo) requires magnetohydrodynamics, fluid mechanics, and substantial numerical computation, and the first complete computational models did not exist until the 1990s.

The 19th-century mapping projects

The 19th century saw an explosion of declination measurement and mapping driven by the needs of increasingly long-distance navigation. The British Admiralty, the French Bureau des Longitudes, the German Bonn observatory, and similar institutions in Russia, the United States, and elsewhere ran long-running programs to measure declination at known stations and to map the field at sea via ship-based surveys.

Carl Friedrich Gauss made the next major theoretical contribution in 1832 with his measurement of the absolute magnetic field strength (previously, only the direction had been measured, not the magnitude). Gauss also developed the mathematical machinery (spherical harmonics) that allowed the global magnetic field to be represented as a sum of terms with known geographic distributions. The first global representation using this approach was Gauss's 1838 model based on 84 stations.

The international Magnetic Crusade of 1839-1845 set up a network of permanent magnetic observatories around the British Empire and at allied institutions worldwide, taking continuous synchronized measurements. The Crusade was one of the first global coordinated scientific projects and the precursor to several 20th-century projects like the International Geophysical Year of 1957-58.

The current situation

The Earth's magnetic field continues to change, and the current rate of change is unusually fast. The magnetic north pole, which was located in the Canadian Arctic for most of the 20th century, has been moving toward Siberia since the 1990s at roughly 50 km per year. The current declination at most populated locations is changing by tens of arc-minutes per year, which is large enough that aeronautical charts have to be reissued every five years to keep up.

The World Magnetic Model, maintained by the British Geological Survey and the United States National Geospatial-Intelligence Agency, is the current authoritative global declination model. The model is updated every five years and is used by aviation, marine navigation, smartphone compasses, and military systems worldwide. The 2020 model had to be re-released early in 2019 because the field had drifted faster than predicted; the 2025 model was released in late 2024 with revised drift rates.

The underlying physics is that the geodynamo is in an active phase, and some researchers (notably the Leeds group around Phil Livermore) have argued that the rapid pole motion since 1990 reflects a competition between two regions of strong magnetic flux in the core, one under Canada and one under Siberia, with the Siberian patch currently winning. The interpretation is contested, but the observation that the field is changing faster than at any time in the modern instrumental record is not.

What this means for the rest of us

The practical relevance of declination has shrunk dramatically with the advent of GPS, which determines position from satellite signals rather than from compass readings. Most modern navigation does not require knowing the declination at all. The exceptions are situations where GPS is unavailable (deep mines, polar regions during certain solar events, underwater, hostile electronic environments) and situations where the compass is being used as a backup for GPS failure. The redundancy applications keep declination knowledge alive at the institutional level even as it disappears from daily life.

The smartphone compass that points to "north" in your maps app is consulting the World Magnetic Model in real time to correct for the local declination. The aviation chart that gives a runway heading lists the magnetic heading rather than the true heading and notes when the heading changed because the declination changed. The deep-water-drilling industry uses declination corrections to align well casings, and the recalculated 2025 corrections were a small operational nuisance to that industry when they were released.

Three observations

First: the magnetic declination project is one of the cleanest examples in the history of science of a single observable phenomenon (the compass needle does not point to the pole) revealing successive layers of underlying structure (Earth has a magnetic field, the field is offset from the rotational axis, the field changes over time, the changes are produced by liquid iron in the core, the core dynamics can be modeled mathematically). Each layer required new instruments, new theories, and new institutions to study. The phrase "more is different" applies cleanly: the phenomenon is one thing, but the explanations span 400 years and span chemistry, physics, geology, and fluid mechanics.

Second: the scientific institutions built to study magnetism (observatories, expeditions, international coordination programs) became templates for the institutional structures of other geosciences and eventually of climate science. The Magnetic Crusade of 1839 is recognizably the same institutional form as the International Geophysical Year of 1957 and the Intergovernmental Panel on Climate Change of 1988. The long-running, multi-decade, internationally coordinated scientific monitoring project is one of the most consequential 19th-century institutional inventions, and magnetism was where it was first developed at scale.

Third: the technological obsolescence of compass navigation is more recent than people typically appreciate. GPS became commercially available in 1995; before that, every blue-water shipping company in the world had crews trained in compass navigation with declination corrections, and the Notice to Mariners updates that distributed new declination values were among the most-read scientific publications in the maritime industry. A skill that defined an entire profession for 500 years was retired by an unrelated technical advance in the span of about 15 years.

The deeper observation is that the contingencies of which scientific problems become culturally salient are stranger than retrospect tends to suggest. Magnetic declination was a top-tier scientific problem for four centuries because it was operationally critical to navigation. The moment navigation no longer needed it, the field receded into specialist obscurity within a generation. The cultural memory of geomagnetism is now smaller than the institutional infrastructure required to maintain the modern World Magnetic Model, which is itself smaller than it was forty years ago. The historical curiosities of any field correlate poorly with the field's eventual practical importance, and they correlate even worse with the field's cultural memory.