← Daily Dose · Maps The Pole That Walked to Siberia — Daily Dose · Maps No. 019
Daily Dose · Maps  /  Sep 12, 2026  /  Magnetic North

The Pole That
Walked to Siberia

Every compass on Earth points at a place that will not hold still. Since a British officer first stood on the north magnetic pole in 1831, it has walked out of Canada, crossed the Arctic Ocean, and kept going — fast enough, once, to force an emergency rewrite of the model your phone uses to find north.

2,200km travelled

Distance the north magnetic pole has moved since James Clark Ross fixed its position on the Boothia Peninsula in 1831. Most of that ground was covered after 1990. NOAA NCEI / BGS, World Magnetic Model 2025.

85.76°N 139.30°E

Where it is now — north of the New Siberian Islands, closer to Russia than to Canada.

GEOGRAPHIC NORTH POLE 1831 1948 2001 2015 2025 BOOTHIA PEN. ELLESMERE GREENLAND SVALBARD NEW SIBERIAN IS. SIBERIA ALASKA CANADA SCANDINAVIA ICELAND 80°N 70°N 60°N 90°E 180° 90°W
1831 · Ross's fix, Boothia Peninsula
Track of the north magnetic dip pole
2025 · WMM2025 position
Polar azimuthal equidistant, schematic coastlines. Endpoint coordinates are measured; intermediate track is drawn to published pole positions and is indicative, not survey-accurate.

01 · First Fix, 1831

Nobody knew it moved

Ross planted a flag and assumed he had found a fixed point. Seventy-three years later Roald Amundsen went back to check — and the pole had shifted. That single discrepancy is the origin of everything on this map.

70°05′N
96°46′W · James Clark Ross, 31 May 1831
Boothia Peninsula, Nunavut
70°31′N
96°34′W · Roald Amundsen, 1904
≈50 km from Ross's fix
0.7 km/yr
Average drift implied between the two
observed fixes — 73 years apart

Sources: WHOI Magnetics Group, The Magnetic North Pole. The 0.7 km/yr figure is arithmetic on the two observed positions, not a published rate.

02 · Speed

Then it started to run

For most of the instrumental record the pole crept. Between 1990 and 2005 it accelerated roughly fourfold — and in the last five years it has posted the largest slowdown ever recorded.

1831 → 1904 · implied0.7 km/yr
20th c. historic range0–15 km/yr
1990–2005 · peak50–60 km/yr
2018 · WMM emergency update55 km/yr
2025 · latest annual report36 km/yr
South magnetic pole, 20259 km/yr
0204060 KM / YEAR

The emergency

A model went out of tolerance a year early

The World Magnetic Model is normally reissued every five years. By summer 2018 the gap between the 2015 model and the pole's real position had nearly crossed the acceptable navigation-error threshold. NOAA and the British Geological Survey pushed an out-of-cycle release in February 2019 — an unscheduled patch to the planet's compass.

2019
Out-of-cycle WMM release
1.0°
Stated declination error budget of the model

Sources: Livermore et al., Nature Geoscience (2020); NOAA NCEI 2025 WMM Annual Report (published 9 Jan 2026, data through 31 Oct 2025); NOAA NCEI out-of-cycle release notice (2019). Bar lengths scale to a 60 km/yr axis; range bars are hatched.

03 · Declination

The angle your compass gets wrong

Declination is the gap between the north your needle finds and the north on the map. It is not a constant, and near the poles the field goes so horizontal that a compass stops working at all.

BLACKOUT CAUTION COMPASS USABLE
0 nT 2,000
BLACKOUT LIMIT
6,000
CAUTION LIMIT
16,000 28,000 40,000 nT

Horizontal field intensity, H. Below 2,000 nT the WMM declares a blackout zone — compasses are unreliable. 2,000–6,000 nT is a caution zone. Both rings surround each magnetic pole. Source: NOAA NCEI, WMM accuracy & limitations.

Greenwich, September 2019

For the first time since about 1660, a compass at the Royal Observatory pointed at true north

The agonic — the line where declination is exactly zero — swept across zero longitude. UK compasses had pointed west of true north for roughly 360 years. The agonic is moving west at about 20 km a year; by around 2040 British compasses should point east of true north.

≈20 km/yr
Westward drift of the agonic line (BGS, 2019)
≈1660
Last time declination at Greenwich was zero — near the Observatory's founding

Source: British Geological Survey, True north and magnetic north cross at Greenwich for the first time in 360 years (2019).

04 · On the ground

They have to repaint the runways

Runway numbers are magnetic bearings rounded to ten degrees. When declination drifts far enough, the paint is wrong — so crews go out and change the number on the tarmac.

1L → 2L
Fairbanks International renamed runway 1L-19R to 2L-20R in 2009
≈24 yrs
Interval between renumberings at Fairbanks; next expected around 2033
±1.0°
FAA tolerance: published magnetic variation must track the WMM within one degree
≈56 yrs
Estimated wait before London Stansted, renumbered in 2009, needs it again

The closer to the pole, the faster the change. Sources: NOAA NCEI, Airport Runway Names Shift with Earth's Magnetic Field; FAA Order 8260 guidance via NBAA.

Resolution

Two models now, one ten times sharper

WMM2025 shipped on 17 December 2024 alongside a first-ever high-resolution twin, WMMHR2025. Both run to the end of 2029.

WMM2025 · resolution at equator3,300 km
WMMHR2025 · high resolution≈300 km

Source: NGA / NOAA / BGS joint release, December 2024. The model "can be found in virtually every smartphone" (NOAA NCEI, 2026).

05 · Cause

Two blobs, one tug of war

The pole is not a thing that travels. It is where two deep patches of magnetic flux, 2,900 km down at the core–mantle boundary, happen to balance. One of them is losing.

Over the last two decades the pole's position has been set by two large lobes of negative flux on the core–mantle boundary — one under Canada, one under Siberia. Between 1999 and 2019 the Canadian lobe stretched out lengthwise and split into two joined patches. Splitting into smaller scales made it weaker at the surface, because small features fade faster on the way up through the mantle. The Siberian lobe won by default, and the pole slid toward it.

Source: Livermore, Finlay & Bayliff, "Recent north magnetic pole acceleration towards Siberia caused by flux lobe elongation," Nature Geoscience (2020), using ESA Swarm data.

−9%
Global average loss in field strength over the past 200 years (ESA)
24,000 → 22,000 nT
Minimum intensity in the South Atlantic Anomaly over the past 50 years (ESA Swarm)
80.85°N 72.76°W
The geomagnetic north pole, 2025 — a modelled dipole point, and a much steadier one than the dip pole on the map

Why it matters

What to notice on the map

  • The track is almost a straight line. From Boothia to 85.8°N it barely wanders — the pole is not drifting randomly, it is being pulled.
  • The dots bunch on the left and spread on the right. Nearly two centuries of slow creep in Canada, then most of the distance covered since 1990.
  • It has already crossed the geographic pole. The 2025 position sits on the Siberian side of the map — magnetic north is now a Russian-hemisphere feature.
  • Nothing on this map is a border. It is drawn by two patches of iron-rich flow 2,900 km beneath the seabed, and it is redrawn every five years — or sooner, when it has to be.
Share X WhatsApp LinkedIn Reddit

Get it in your inbox

Like this? Get a new map in your inbox every morning.