The Carrington model refers to the scientific framework built around the most powerful geomagnetic storm in recorded history, the September 1859 event observed by Richard Carrington and Richard Hodgson, and used today to estimate what a repeat of that storm would do to modern infrastructure. On the morning of September 1, 1859, both astronomers independently watched an intense white-light solar flare erupt from the sun’s surface, and roughly 17 hours later, the strongest geomagnetic disturbance ever documented slammed into Earth’s magnetic field.1Space Weather. Observations of Low‐Latitude Red Aurora in Mexico During the 1859 Carrington Geomagnetic Storm That event has become the benchmark against which space weather researchers model extreme scenarios, and the stakes of getting those models right have grown enormously since 1859.
What Actually Happened in 1859
Before the Carrington Event, nobody had drawn a firm connection between activity on the sun and disturbances on Earth. The 1859 storm changed that permanently: for the first time, observations demonstrated that the sun and the aurora were linked, and that auroral activity generated powerful electrical currents in the upper atmosphere.2PubMed Central. Duration and extent of the great auroral storm of 1859 The auroral display itself was extraordinary. Brilliant red auroras were seen at latitudes far closer to the equator than normal, including in Mexico, confirming that the storm’s reach was truly planetary.3Space Weather. Observations of Low‐Latitude Red Aurora in Mexico During the 1859 Carrington Geomagnetic Storm
The practical damage was limited only because the world’s electrical infrastructure in 1859 consisted of telegraph lines and not much else. Even so, the induced currents were strong enough to shock telegraph operators, set paper on fire, and in one famous case allow the Boston telegraph office to disconnect its batteries entirely and continue sending messages powered solely by the current flowing through the wires from the storm. The event demonstrated that geomagnetic storms could inject usable amounts of electrical energy into long conductors, a fact that would prove far more consequential as civilization became dependent on electrical grids, pipelines, and communication cables.
How Fast and Powerful Was the Ejection
A solar flare on its own is a burst of electromagnetic radiation that reaches Earth in about eight minutes. What makes a Carrington-class event so dangerous is the coronal mass ejection (CME) that accompanies it: a massive cloud of magnetized plasma hurled outward from the sun. The speed of that cloud determines how quickly it arrives and how hard it hits. Most CMEs take two to four days to reach Earth. The 1859 ejection covered the distance in roughly 17 hours, implying extraordinary velocity.
Recent modeling of fast CME transit times has refined the estimates. Using observational data on how interplanetary drag slows ejections during transit, researchers estimate that the Carrington CME traveled at an average speed of about 850 km/s, with a peak one-hour speed reaching roughly 1,200 km/s.4Journal of Geophysical Research: Space Physics. Near‐Earth Solar Wind Speed of Fast Coronal Mass Ejections Those speeds are consistent with the short transit time and with what is known about how the solar wind slows CMEs through aerodynamic-like drag. The relationship between transit time and arrival speed turns out to be sublinear rather than a simple straight-line extrapolation, which matters for predicting future events: a CME that is twice as fast at the sun does not arrive twice as fast at Earth, because faster ejections experience proportionally more deceleration.
What a Carrington-Scale Storm Does to Earth
When a fast CME slams into Earth’s magnetic field, it compresses the magnetosphere violently and sets off a cascade of electromagnetic effects. Modeling of an extreme CME impact suggests the interplanetary shock would arrive with a magnetosonic Mach number around 45, creating a sudden magnetic impulse of roughly 234 nT and a rate of magnetic-field change of about 30 nT per second.5Geophysical Research Letters. An extreme coronal mass ejection and consequences for the magnetosphere and Earth Those numbers may sound abstract, but the rate of change is the critical quantity. A rapidly fluctuating magnetic field induces electric fields on the ground, and those electric fields drive currents through anything conductive: power lines, pipelines, railroad tracks, undersea cables.
Global magnetohydrodynamic (MHD) simulations have been developed specifically to reproduce the magnetospheric response to Carrington-class driving conditions. One such model, built at the University of Michigan, feeds idealized but extreme solar wind parameters into a three-dimensional simulation of the magnetosphere-ionosphere system and tracks the resulting geoelectric fields on the ground.6Journal of Geophysical Research: Space Physics. Modeling extreme “Carrington‐type” space weather events using three‐dimensional global MHD simulations These simulations are the core of what people mean by a “Carrington model” in the operational sense: a scenario-based tool for estimating how bad things could get if a similar storm struck today.
Why Local Geology Changes Everything
One of the less intuitive findings in Carrington-class modeling is that the danger is not distributed evenly by latitude. You might expect higher latitudes to always face the worst effects, and in a general sense they do, since the aurora and its associated currents are concentrated near the poles. But the actual electric fields induced at the surface depend heavily on what is beneath the surface. Rock and soil conduct electricity at vastly different rates depending on their composition, and those differences create hotspots that can appear far from the auroral zone.
Modeling of geoelectric fields in the UK during three major geomagnetic storms (March 1989, October 2003, and September 2017) showed that the amplitude of the ground-level electric field was not determined solely by latitude but also by local geology.7Space Weather. Developing a New Ground Electric Field Model for Geomagnetically Induced Currents in Britain Based on Long‐Period Magnetotelluric Data Similarly, simulations of geomagnetically induced currents (GIC) in Japan’s 500 kV power grid found that the induced electric field showed localized, uneven distribution driven by charge accumulation at boundaries where underground conductivity changes sharply.8Space Weather. Time Domain Simulation of Geomagnetically Induced Current (GIC) Flowing in 500‐kV Power Grid in Japan Including a Three‐Dimensional Ground Inhomogeneity
This has a further complication. The direction of the induced electric field, not just its strength, is shaped by the subsurface conductivity structure. Research has shown that geomagnetic storms produce electric fields with a preferred directionality that depends on local geology, and that the two horizontal components of the electric field can become highly correlated in certain regions. Power grids whose transmission lines happen to align with that preferred direction face disproportionate risk.9Space Weather. The Influence of 3‐D Earth Conductivity, Geoelectric Field Polarization, and Power Grid Topology on GIC Risk A grid in a region of resistive bedrock oriented in an unfortunate direction could experience far higher GIC than a grid at the same latitude sitting on conductive sediment.
Power Grids and Transformers
The most talked-about vulnerability is the electrical grid, and within the grid, the high-voltage transformers. When geomagnetically induced currents flow into a transformer through its grounding connection, they push the transformer’s magnetic core into a state called half-cycle saturation. This distorts the normal alternating-current waveform, generates harmonics, and forces the transformer to consume excess reactive power, all of which can degrade grid stability or, in the worst case, physically damage the transformer.
Observations from New Zealand during elevated GIC events illustrate how sensitive some transformers are. Even at relatively low GIC levels (under 7 amps), single-phase bank transformers showed clear linear increases in reactive power consumption, consistent with core saturation. The researchers described the problem as a lack of “headroom”: these transformer designs have almost no tolerance for even small amounts of quasi-DC current.10Space Weather. Geomagnetically Induced Currents, Transformer Harmonics, and Reactive Power Impacts: Observed Response of a Single‐Phase Transformer in New Zealand Large-scale transformer damage during a Carrington-class storm could leave regions without power for weeks or months, because high-voltage transformers are custom-built, expensive, and have lead times measured in months to over a year.
Satellites and Low-Earth Orbit
Geomagnetic storms heat and expand the upper atmosphere, increasing drag on satellites in low-Earth orbit. A real-world demonstration of this risk occurred in February 2022, when SpaceX launched 49 Starlink satellites into an initial low orbit just as consecutive geomagnetic storms struck. Thermospheric density at 210 km altitude was at least 20 to 30 percent higher than in the days before launch, and 38 of the 49 satellites could not raise their orbits fast enough to overcome the extra drag. They reentered the atmosphere and were lost.11Space Weather. The Thermosphere Is a Drag: The 2022 Starlink Incident and the Threat of Geomagnetic Storms to Low Earth Orbit Space Operations
That storm was moderate by historical standards. A Carrington-scale event would drive far greater atmospheric expansion. Satellites at altitudes below a few hundred kilometers would experience dramatic increases in drag, potentially deorbiting assets that were expected to operate for years. Even satellites at higher altitudes would face degraded orbit predictions, making collision-avoidance maneuvers less reliable. Research into orbital drag during magnetic storms caused by CMEs has confirmed that the effect scales with storm intensity, making extreme events a genuine threat to the growing population of low-orbit satellites and mega-constellations.12Space Weather. Satellite Orbital Drag During Magnetic Storms
Submarine Cables and the Internet
A less widely appreciated vulnerability involves the undersea fiber-optic cables that carry the vast majority of intercontinental internet traffic. Although the data itself travels as light through glass, these cables contain a copper conductor that carries electrical power to the optical repeaters spaced along the cable’s length. During a geomagnetic storm, induced voltages along the cable can add to or subtract from the power feed, potentially disrupting the repeaters.
Historical experience confirms this is not hypothetical. Submarine cables have experienced problems during past extreme geomagnetic disturbances precisely because of induced voltages interacting with the power supply.13Space Weather. An Examination of Geomagnetic Induction in Submarine Cables Modern fiber-optic cables remain vulnerable because of their internal copper conductors used for powering repeaters.14Space Weather. Validating SCUBAS Predictions of Geomagnetically Induced Voltage in Submarine Cables Using Legacy Superstorm Observations There is growing concern in the research community that a large CME event could induce enough perturbation in the power supply of transoceanic cables to cause a widespread internet blackout.15arXiv. Effects of geomagnetic field perturbations on the power supply of transoceanic fiber optic cables Unlike power grid damage, which is regional, cable disruption could isolate entire continents from one another digitally.
How Likely Is Another Carrington Event
This is where the research community has gone back and forth considerably. Early estimates placed the probability of a Carrington-class storm in any given decade at around 12 percent, a number that got widespread media attention. More recent statistical work using extreme value methods has produced substantially lower figures. One analysis using the Weibull distribution estimated the probability of a Carrington-strength or stronger event occurring in the decade 2018–2027, given that none has occurred since 1859, at about 0.92 percent, with a 95 percent confidence interval stretching from roughly 0.5 to 1.9 percent.16Scientific Reports. Probability estimation of a Carrington-like geomagnetic storm That is roughly one-thirteenth of the earlier estimate, and the authors explicitly noted the sharp contrast.
Approaching the question from the flare side rather than the geomagnetic side, extreme value theory applied to satellite X-ray flux data suggests that a Carrington-class flare is roughly a one-in-100-year event, with an upper bound estimate placing it as frequent as one in 30 years. The expected 150-year return level is approximately an X60 flare, with a worst-case estimate of X90.17Space Weather. Using Extreme Value Theory for Determining the Probability of Carrington‐Like Solar Flares These two approaches measure slightly different things: one looks at how intense the geomagnetic disturbance was, the other at how powerful the flare itself was. A strong flare does not always produce a strong geomagnetic storm, because the CME might miss Earth or arrive with its magnetic field oriented unfavorably for coupling with Earth’s field.
There is also an important distinction between the 1859 storm and truly ancient extreme events recorded in tree rings and ice cores. The 774 CE event, identified through spikes in carbon-14 and other cosmogenic isotopes, was far more powerful than the Carrington Event in terms of solar energetic particle (SEP) fluence. Interestingly, high-precision carbon-14 measurements from Alaskan trees covering 1844–1876 found no significant increase characteristic of the extreme SEP signatures seen in events like 774 CE, suggesting the Carrington storm’s particle output was a fundamentally different class of event.18Journal of Space Weather and Space Climate. No signature of extreme solar energetic particle events in high-precision 14C data from the Alaskan tree for 1844–1876 CE A separate study did detect a transient offset in carbon-14 at high-latitude sites after the Carrington Event, but the finding required explanations beyond current models, possibly involving unusually rapid transport of carbon-14 from the upper atmosphere to the surface at polar latitudes.19Geophysical Research Letters. Transient Offset in 14 C After the Carrington Event Recorded by Polar Tree Rings The upshot is that the Carrington Event, while the worst documented geomagnetic storm, was not the worst solar event in the last few thousand years, and the modeling framework needs to account for events potentially much larger.
Early Warning and Forecasting
The good news is that a Carrington-class CME would not arrive without warning. Even the 1859 ejection took about 17 hours to reach Earth after the visible flare. Current space weather monitoring relies on spacecraft positioned at the Sun-Earth L1 Lagrange point, roughly 1.5 million kilometers sunward of Earth. A CME passes these sentinels about 15 to 60 minutes before hitting Earth’s magnetic field, giving grid operators a narrow but valuable window to take protective action.
There is a push to expand this capability. A workshop on future monitoring missions concluded that pairing an L1 spacecraft with one at the L5 Lagrange point, which trails Earth by 60 degrees in its orbit, would provide complementary views of solar and interplanetary conditions that a single vantage point cannot.20Space Weather. L1L5Together: Report of Workshop on Future Missions to Monitor Space Weather on the Sun and in the Solar Wind Using Both the L1 and L5 Lagrange Points as Valuable Viewpoints An L5 observatory would see CMEs from the side as they propagate toward Earth, allowing far better estimates of their speed, size, and magnetic structure before they arrive. The European Space Agency’s Vigil mission, planned for the L5 point, is one effort to realize this concept.
Forecasting models have also advanced. Ionospheric prediction during storms can maintain reasonable accuracy under moderately disturbed conditions, though accuracy degrades noticeably during extreme storms.21Radio Science. Assessment of ionosphere tomographic modeling performance using GPS data during the October 2003 geomagnetic storm event GPS-based navigation, aviation communication, and precision agriculture all depend on ionospheric models that assume relatively stable conditions, and a Carrington-class event would push those models well outside their reliable range. This is one of the less dramatic but practically important consequences: even if the power grid survives, GPS accuracy could degrade for hours or days.
Hardening the Grid
The most direct defense against GIC damage to transformers involves blocking the quasi-DC current from entering through the transformer’s neutral grounding connection. GIC blocking devices typically insert a capacitor in the neutral grounding path, which blocks the slow-varying induced current while allowing normal alternating current to pass. Testing of these devices on a real-time simulation platform across different transmission voltages and fault conditions has shown that, for typical capacitance values, the devices do not interfere with distance protection relays, meaning they can be installed without requiring changes to the grid’s existing protective relay settings.22Electric Power Systems Research. Modeling the impact of GIC neutral blocking devices on distance protection relay operations for transmission lines
Despite the demonstrated effectiveness of blocking devices, deployment has been slow and uneven. The devices add cost, and the perceived probability of a Carrington-class event in any given year is low enough that many utilities treat it as a low-priority risk. In the United States, FERC and NERC have established reliability standards requiring utilities to assess their GIC vulnerability and develop mitigation plans, but the specifics of what counts as adequate mitigation vary. Some utilities have installed blocking devices on their most vulnerable transformers; others rely on operational procedures like reducing power transfers and taking reactive power reserves online when warnings arrive.
Radiation Risk at Flight Altitudes
Beyond electromagnetic effects on the ground, extreme solar particle events pose a direct radiation hazard to people at high altitudes. Aircraft crew and passengers are partially shielded by Earth’s atmosphere, but that shielding thins considerably at cruising altitude and near the poles. An assessment of what a 774 CE-class event would mean for flight-altitude radiation found that the integrated ambient dose at polar latitudes would reach roughly 9.5 sieverts at 50,000 feet, 6 sieverts at 40,000 feet, and 2 sieverts at the more typical cruising altitude of 30,000 feet, all representing a severe threat including the potential for acute radiation syndrome.23J. Space Weather Space Clim.. Assessment of the radiation risk at flight altitudes for an extreme solar particle storm of 774 AD
The 774 CE event was far more intense in particle output than the 1859 Carrington storm, as the tree-ring evidence discussed earlier shows. For a repeat of the 1859 event specifically, the radiation doses at flight altitude would likely be substantially lower but still a concern for polar routes. Airlines already reroute flights away from polar paths during strong solar particle events, but the speed at which particles arrive (traveling close to the speed of light, they reach Earth in minutes rather than hours) means that flights already in the air over the poles when an event begins would have little time to descend to lower altitudes. Developing faster particle-event alerts and automated rerouting protocols remains an active area of work in aviation safety and space weather operations.