Magnetic north has moved more than 2,200 km since 1831.
That part is measured fact.4 Whether the magnetic field is about to flip, or the planet is about to roll over, are separate questions with very different answers. Poleshift.earth sorts the evidence, lets you run every scenario on a live globe, and gives you a specific plan for each one.
Space weather right now
Current conditions from NOAA’s Space Weather Prediction Center, translated into what they mean where you are.31
Latest NOAA alerts, watches and warnings
Get a heads-up before the next big storm
We’ll email you when NOAA issues a geomagnetic storm watch, warning or alert at the level you choose, with what it means and what to do. Watches often come a day or more ahead.
Prefer NOAA’s raw bulletins? Its Space Weather Prediction Center offers free email alerts too.12
“Pole shift” means three different things
Most of the confusion starts here. Only one of the three can move water, and it’s the one with the least evidence for a sudden version.
| What people mean | What actually moves | How fast | Evidence | Moves oceans? |
|---|---|---|---|---|
| Magnetic pole drift | The spot where a compass needle points straight down, driven by molten iron flowing in the outer core. | About 35 km a year now, after peaking at 50 to 60 km a year in the 2000s.4 | Measured | No |
| Magnetic excursion or reversal | The whole field weakens, tangles and may flip polarity, so compasses end up pointing south. | Centuries to millennia. The Laschamps excursion took about 250 years to flip and stayed flipped for about 440.8 | Established | No |
| Geographic or crustal shift | The solid Earth itself, relative to its spin axis. | Slow and small today: groundwater pumping alone shifted the spin pole about 80 cm east between 1993 and 2010.10 A sudden shift over days has never been documented. | Established for slow drift, Speculative for sudden shifts | Yes, and it’s the only kind that does |
Polaris stays the North Star in every scenario on this site. Even the sudden-shift theories keep the spin axis pointed at the same spot in the sky; it’s the ground underneath that would move.23
How claims are labeled
Pole-shift talk mixes satellite measurements with forecasts nobody can test yet. Every claim here carries one of four labels, so you always know which is which.
- Measured
Observed directly by satellites, observatories and field surveys.
- Established
Mainstream science from peer-reviewed research and the geological record.
- Debated
Real research that specialists still argue about.
- Speculative
Predictions and theories without supporting evidence, or that conflict with it.
The drift: what’s been measured
James Clark Ross found the north magnetic pole in northern Canada in 1831, and it has been crossing the Canadian Arctic toward Russia ever since.1 For most of the 20th century it crept along at 5 to 15 km a year. In the 1990s it took off toward Siberia, peaking at 50 to 60 km a year, and over the last five years it slowed to about 35 km a year, the biggest deceleration on record.4
The cause is a tug-of-war between two patches of strong magnetic field deep under Canada and Siberia. Lately the Siberian patch has been winning.4 The geographic North Pole, the one Earth spins around, hasn’t moved in any meaningful way.
Your phone’s compass corrects for all of this using the World Magnetic Model, which NOAA and the British Geological Survey update every five years.2 The pole’s sprint forced an early update in 2019,5 and the current WMM2025, released in December 2024, runs through 2029.4
The field is weaker, but probably not flipping
Earth’s main magnetic dipole is about 7.6% weaker than in 1900, by the same IGRF-14 model,3 part of a decline of roughly 5% per century since at least 1840.7 The South Atlantic Anomaly, a weak zone where satellites take extra radiation, has grown since 2014 by an area nearly half the size of continental Europe.6
Is a flip starting? The best evidence says probably not. A 2018 study compared today’s field with the two most recent excursions and found neither resembled current changes. Similar weak patches 46,000 to 49,000 years ago recovered on their own.7
Run the drift in the simulator to see the pole’s position year by year and how far your own compass reading has moved.
- 1831
Ross locates the north magnetic pole in northern Canada.1
- 1948–1994
Canadian surveys track it across the Arctic.1
- 2001 and 2007
International expeditions measure it moving north-northwest at about 55 km a year.1
- 2019
The World Magnetic Model gets an early, out-of-schedule update because the pole moved faster than forecast.5
- December 2024
WMM2025 and a new high-resolution version are released.4
- 2025
The pole sits at 85.76°N, 139.30°E, closer to Siberia than to Canada.1
- October 2025
ESA’s Swarm satellites show the South Atlantic Anomaly still expanding.6
Space weather, explained
Nearly everything the Sun does to Earth arrives in one of three forms. Each travels at a different speed and hits different systems.
- 8 minutes
Solar flares
A burst of X-rays at light speed. It blacks out HF radio on the sunlit side of Earth. There’s no warning: the light is the warning.
- Minutes to hours
Solar energetic particles
Radiation storms that threaten astronauts, satellite electronics and crews on polar flights.
- 15 hours to 3 days
Coronal mass ejections
Billion-ton clouds of magnetized plasma. When one hits, it sets off a geomagnetic storm: grid currents, GPS errors, satellite drag and auroras.
Forecasters can see a CME leave the Sun, which usually gives a day or more of notice. The detail that decides how bad it gets, the direction of the cloud’s magnetic field, can only be measured when it passes a monitoring spacecraft about 1.5 million km upstream, roughly 15 to 60 minutes before impact. That’s why NOAA issues watches days ahead and warnings minutes ahead.12
NOAA’s geomagnetic storm scale
| Level | Kp index | Typical frequency per 11-year cycle | What it can do |
|---|---|---|---|
| G1 Minor | 5 | About 1,700 times | Weak grid fluctuations; auroras at high latitudes. |
| G2 Moderate | 6 | About 600 times | Voltage alarms on high-latitude grids; auroras to about New York and Idaho. |
| G3 Strong | 7 | About 200 times | Voltage corrections needed; GPS and HF radio problems; auroras to about Illinois and Oregon. |
| G4 Severe | 8 | About 100 times | Widespread voltage-control problems; GPS degraded for hours; auroras to about Alabama and northern California. |
| G5 Extreme | 9 | About 4 times | Some grids may collapse and transformers may be damaged; HF radio out for a day or two; auroras to about Florida and southern Texas.15 |
Scale descriptions summarized from NOAA’s Space Weather Prediction Center.11
What it hits
Power grids Measured
Storms drive currents through long transmission lines into transformers. In March 1989, Québec’s grid collapsed in about 90 seconds and 6 million people lost power for 9 hours. A G5 storm in 2003 caused outages in Sweden and damaged transformers in South Africa.15
Satellites Measured
Storms heat and swell the upper atmosphere, adding drag. In February 2022, a minor storm helped bring down 38 of 49 newly launched Starlink satellites.
GPS and navigation Measured
A disturbed ionosphere bends and scrambles satellite signals. During the May 2024 G5 storm, GPS-guided farm equipment lost accuracy in the middle of planting season.
Radio and aviation Measured
Flares black out HF radio on the dayside, and radiation storms force airlines to reroute polar flights.
Radiation Established
Astronauts and high-altitude crews take real doses. At ground level the atmosphere still shields you; no solar storm on record has caused radiation sickness on the ground.
Pipelines and rail Measured
Induced currents speed up pipeline corrosion and can trip railway signals.
Undersea internet cables Debated
The glass fiber is immune, but the repeaters along each cable are powered over long conductors. How exposed they’d be in a Carrington-class storm is still argued.
People and animals Debated
There are no established direct health effects on the ground. Some studies link storms to navigation errors in migrating birds and to whale strandings.
How likely is the big one?
Estimates for a Carrington-class storm in any given decade range from 0.46–1.88% in a 2019 analysis to about 12% in a 2012 one, depending on the statistical method.1617 A 2008 National Research Council report put the first-year cost of a severe event in the US at $1–2 trillion, with 4 to 10 years to fully recover.18
The deep record is sobering. Stars like the Sun produce superflares, far bigger than Carrington’s, roughly once per century,19 and tree rings record five extreme solar particle events plus three candidates in the past 12,000 years, about one every 1,500 years.20 Whether our Sun can produce a superflare is still unknown.
- 774–775
Tree rings worldwide record a sudden carbon-14 spike from an extreme solar particle event, the first known “Miyake event” (Miyake et al., Nature, 2012).
- 1859
The Carrington Event. Telegraph systems failed and sparked, and auroras reached the tropics. It’s still the benchmark for a worst case.
- 1921
The “New York Railroad Storm” starts fires in telegraph and signal stations.
- 1972
An ultra-fast storm sets off a large number of US Navy sea mines off North Vietnam, between the Apollo 16 and 17 missions (Knipp et al., Space Weather, 2018).
- 1989
Québec’s grid collapses in about 90 seconds; 6 million people lose power for 9 hours.
- 2003
The Halloween storms reach G5, cutting power in Sweden and damaging transformers in South Africa.15
- 2012
A Carrington-class eruption crosses Earth’s orbit about a week after Earth passed that point (Baker et al., Space Weather, 2013).
- 2024
May 10–11: the first G5 storm since 2003, with auroras deep into the southern US.15
- 2025
G4 storms on January 1, June 1 and November 11. USGS expects more G4 storms, possibly a G5, within a year or two.1314
Run the scenarios
Choose an outcome. The globe shows what changes, the readout tells you what it means where you are, and the plan underneath updates to match.
What it means
Who’s predicting what
From space agencies to independent forecasters, in the same format: the claim, the timeframe, the evidence label and where it stands.
| Source | Claim | Timeframe | Evidence | Where it stands |
|---|---|---|---|---|
| NOAA and British Geological Survey (WMM2025) | Magnetic north keeps moving toward Siberia, at a slower pace than in the 2000s.4 | 2025–2029 | Measured | Tracking. The pole has slowed to about 35 km a year. |
| IGRF-14 international model | The pole reaches about 84.7°N, 125.9°E by 2030.3 | 2030 | Established | Model forecast. |
| Brown, Korte and others (PNAS, 2018) | Today’s weakening field probably isn’t heading into a reversal or excursion and is likely to recover.7 | Coming centuries | Established | Consistent with observations so far. |
| USGS Geomagnetism Program (2025) | More G4 storms, and possibly a G5, are likely within a year or two.13 | 2025–2027 | Established | Open. |
| Vasilyev and others (Science, 2024) | Sun-like stars produce superflares roughly once per century.19 | Ongoing | Established for stars, Debated for our Sun | No superflare in the instrumental record. |
| Ben Davidson (Suspicious0bservers, SpaceWeatherNews) | A recurring disaster cycle of about 12,000 years: a solar “micronova,” a magnetic excursion, and Earth turning over about 90°, putting new polar regions near India and Brazil.2122 | Presented as approaching; no specific date in the sources reviewed here | Speculative | No geological record of 90° turnovers. Astronomers don’t recognize a solar “micronova,”24 and even other catastrophists dispute the 90° model.22 |
| ECDO thesis (2024), popularized in the book Inversion | Core–mantle decoupling flips the crust and mantle 104° within days, moving Polaris over Africa’s Rift Valley.23 | Proponents say the warning signs are visible now | Speculative | Earth spins around its axis of greatest inertia, which is stable; the tumbling “Dzhanibekov effect” applies to spin around an object’s middle axis. |
| Charles Hapgood (1958, 1970) | The crust periodically slides over the mantle; he placed the ice-age North Pole in Hudson Bay. | Recurring | Speculative | Overtaken by plate tectonics and the paleomagnetic record. |
| 2012 doomsday claims | A pole shift or planetary flip around December 21, 2012. | 2012 | Speculative | Didn’t happen. |
Myths and straight answers
Will a magnetic pole shift flood the coasts?
No. Magnetic changes happen in the liquid outer core and move compass needles, not water. Only a shift of the solid Earth relative to its spin axis could move oceans, and today that motion is measured in centimeters to meters per year.10
Is Earth’s magnetic field about to flip?
Probably not. It has weakened about 5% per century since at least 1840, but a 2018 comparison with the last two excursions found today’s pattern doesn’t match how they began.7 Reversals unfold over centuries to millennia, not days.
Does a reversal strip away the atmosphere?
No. Earth’s field has reversed many times while the atmosphere and life carried on, and people lived through the Laschamps excursion 41,000 years ago. Cosmic-ray production roughly doubled during that event, which matters for satellites and aircraft far more than for people on the ground.8
Do I need a Faraday cage for a solar storm?
Not for your phone or radio. Storm-driven currents build up in conductors many kilometers long, like power lines and pipelines. The practical risk to you is losing grid power, not fried electronics. A nuclear EMP is a different threat.
Is the pole’s speed-up a warning sign?
Its speed rose sharply from the 1990s, then dropped from about 50 to 35 km a year over the past five years.4 The speed reflects core flows under Canada and Siberia, and it hasn’t been shown to predict excursions or crustal shifts.
Is the Sun about to “micronova”?
There’s no observation of one on the Sun, and astronomers don’t treat it as a solar phenomenon.24 What is real: rare extreme solar particle events, about one every 1,500 years in tree rings, and superflares on other Sun-like stars.20
Would the North Star change?
Not in any of these scenarios. Even in the sudden-shift theories, the spin axis keeps pointing at the same spot in the sky; the ground beneath it would move.23
Sources and method
- NOAA NCEI. Wandering of the Geomagnetic Poles. ncei.noaa.gov
- NOAA NCEI. World Magnetic Model, including declination calculators. ncei.noaa.gov
- IAGA. International Geomagnetic Reference Field, 14th generation (IGRF-14). Used here for the pole path, speeds, dipole strength and city declinations. doi.org/10.5281/zenodo.14012302
- ZME Science. Coverage of the WMM2025 release, with BGS geomagnetic modeler William Brown, December 2024. zmescience.com
- ScienceAlert. Magnetic north’s drift and the early World Magnetic Model update. sciencealert.com
- ESA. Swarm reveals growing weak spot in Earth’s magnetic field, October 2025. esa.int
- Brown, M., Korte, M., Holme, R., Wardinski, I. and Gunnarson, S. (2018). Earth’s magnetic field is probably not reversing. PNAS 115(20). doi.org/10.1073/pnas.1722110115
- Universe Today. Reconstruction of the Laschamps excursion (Panovska, EGU 2024). universetoday.com
- Nowaczyk, N. and others (2013). The unstable geomagnetic field during the last glacial. EGU General Assembly abstract. copernicus.org
- ScienceAlert. Groundwater pumping shifted Earth’s rotational pole (Seo and others, Geophysical Research Letters, 2023). sciencealert.com
- NOAA Space Weather Prediction Center. NOAA Space Weather Scales. swpc.noaa.gov
- NOAA Space Weather Prediction Center. Alerts, watches and warnings. swpc.noaa.gov
- USGS Geomagnetism Program. Severe magnetic storm, November 11, 2025. usgs.gov
- USGS Geomagnetism Program. Severe magnetic storm, June 2, 2025. usgs.gov
- CBS News. NOAA confirms an extreme G5 storm, May 2024. cbsnews.com
- Riley, P. (2012). On the probability of occurrence of extreme space weather events. Space Weather. doi.org/10.1029/2011SW000734
- Moriña, D. and others (2019). Probability estimation of a Carrington-like geomagnetic storm. Scientific Reports 9:2393. Summary: phys.org
- Gizmodo. Riley’s estimate and the 2008 National Research Council cost figures. gizmodo.com
- Vasilyev, V. and others (2024). Sun-like stars produce superflares roughly once per century. Science 386. doi.org/10.1126/science.adl5441
- Max Planck Institute for Solar System Research. Solar superflares once per century, including the tree-ring record. mpg.de
- Davidson, B. (2021). The Next End of the World: The Rebirth of Catastrophism. Author’s site: suspicious0bservers.org
- Catastrophism (Substack). The Mammoth Conundrum: a critique of the 90° flip model, quoting Davidson. catastrophism.substack.com
- Interview on the ECDO thesis and the book Inversion. seagertp.substack.com
- Plait, P. No, there’s no such thing as a solar micronova. SYFY Wire. syfy.com
- Ready.gov. Build a kit. ready.gov/kit
- Ready.gov. Water. ready.gov/water
- Ready.gov. Power outages. ready.gov/power-outages
- FoodSafety.gov. Food safety during power outages. foodsafety.gov
- US EPA. Emergency disinfection of drinking water. epa.gov
- NOAA Space Weather Prediction Center. Real-time data products: NOAA scales, planetary Kp forecast, alerts, solar wind and GOES X-ray flux. services.swpc.noaa.gov
- Map data: elevation from the ETOPO 20-minute grid distributed with Basemap; coastlines from Natural Earth via world-atlas.
How the models work
Pole path and speed. For each year from 1900 to 2030, the north magnetic pole is located by finding where the IGRF-14 field points straight down. Speeds are averages over the model’s five-year epochs, and 2025–2030 uses the model’s forecast.
Sea level in the crustal-shift scenario. When the crust moves relative to the spin axis, the spin’s outward pull on the oceans changes, raising water where land moves toward the new equator and draining it where land moves toward the new poles. The model applies that change for a sudden shift, scales it by an elastic-Earth factor of about 0.7, runs it on a 0.5° elevation grid, and keeps total ocean volume constant. It leaves out the surges during the shift, ice and the slow readjustment of the solid Earth, so treat it as a first-order illustration of a hypothetical.
Storm reach, excursions and reversals. Aurora boundaries follow NOAA’s scale descriptions and are illustrative. Excursion and reversal animations are keyed to the Laschamps reconstruction for strength and timing; the pole paths are illustrative, not forecasts.