A Million-Mile Bubble: Inside the New Push to Forecast the Solar Storms That Can Black Out a Continent
Researchers at Johns Hopkins APL are building a model that simulates space weather across a million miles of territory around Earth, betting that better physics means fewer blackouts when the next Carrington-class storm hits.
On 10 July 2026 a team at the Johns Hopkins Applied Physics Laboratory (APL) published a new model that aims to forecast solar storms across roughly 1.6 million kilometres of the space surrounding Earth — the region where geomagnetic disturbances that can cripple power grids actually build up. The work, described in reporting from Phys.org on 10 July 2026, treats the cislunar environment as a single coupled system rather than a string of disconnected zones, and is part of a broader federal push to harden the North American grid against extreme space weather.
The bet is straightforward. When the Sun ejects a billion-tonne plume of magnetised plasma — a coronal mass ejection — the resulting geomagnetic storm is not born at the moment it touches Earth. It assembles itself, slowly, across the magnetosphere and the surrounding plasma sheet. Predicting what arrives at the surface means modelling that whole volume, not just the moment of impact. APL's new model is an attempt to do exactly that.
The geometry of a blackout
Geomagnetic storms do their damage by inducing direct currents in long conductors: high-voltage transmission lines, oil and gas pipelines, undersea cables. A sufficiently strong event can saturate transformer cores, trigger protective relays and, in the worst documented cases, take grid nodes offline for hours or days. The benchmark remains the Carrington Event of 1859, which drove auroral displays as far south as the Caribbean and set telegraph paper on fire. A modern repetition would not need to be quite that strong to inflict serious harm; an event on the scale of the 1989 Quebec storm, which knocked out Hydro-Québec's grid for nine hours, is now treated as a routine planning scenario rather than an outlier.
The U.S. federal government has been preparing for the larger version. The National Space Weather Strategy and Action Plan, refreshed across both the Trump and Biden administrations, treats extreme geomagnetic storms as a national-security risk on par with a major hurricane or a cyberattack on the bulk-power system. FEMA and the North American Electric Reliability Corporation (NERC) have run joint exercises; the Department of Energy has funded grid-hardening research; insurance markets have begun pricing the exposure.
What has been thinner is the forecasting layer that sits in front of all of it. The Sun is watched constantly — the NOAA/NASA Solar Dynamics Observatory, the Parker Solar Probe and the ESA/NASA Solar Orbiter together give an unusually rich picture of the corona and the inner heliosphere. The problem is what happens between roughly 0.5 and 30 solar radii, the region threaded by the solar wind and the interplanetary magnetic field, where a coronal mass ejection transforms from a coherent plume into the structured disturbance that will eventually hit the magnetosphere. That intermediate zone is where APL's million-mile bubble lives.
A different kind of model
Most operational space-weather forecasting today relies on a chain of empirical and semi-empirical models: a coronal model that estimates when an ejection leaves the Sun, a heliospheric model that propagates it outward, and a magnetospheric model that estimates the geoeffective impact. Each link in that chain passes a simplified summary to the next. Errors compound at every handoff.
APL's model, as described in the 10 July 2026 reporting, attempts to simulate the physics of the cislunar region as one coupled domain. That includes the solar wind's interaction with Earth's magnetosphere, the behaviour of the plasma sheet that wraps the night side of the planet, and the ring current that builds up during a storm. The practical effect is to keep more of the underlying physics in play from the moment an eruption is observed on the Sun to the moment a disturbance reaches geostationary orbit, where most of the satellites that modern life depends on — GPS, weather, communications — sit.
For grid operators the difference is felt in lead time and false-alarm rate. Today's best forecasts give utility control rooms somewhere between 12 and 24 hours of useful warning before a major geomagnetic storm, with confidence intervals that can swing an order of magnitude. A model that retains the underlying physics through the propagation step should, in principle, narrow both the uncertainty and the false alarms — letting operators stage protective actions without paying the cost of precautionary blackouts that turn out to have been unnecessary.
What the wire is not saying
Space-weather coverage tends to cluster around two frames: the spectacle of an auroral display visible from unusual latitudes, and the catastrophic scenario of a multi-week continental blackout. Both are real, but neither is the everyday case. The everyday case is a moderate storm that knocks a transformer offline here, degrades GPS accuracy for precision agriculture there, and forces a few airlines to divert polar routes because high-frequency radio has gone patchy. Those events rarely make headlines, which is part of why preparedness lags.
The other under-reported angle is the international dimension. The same coronal mass ejection that pushes geomagnetically induced currents through the U.S. Eastern Interconnection pushes them through the European grid, the Nordic grid, and the Russian and Chinese networks. Forecasting capacity is unevenly distributed. The U.S. and the EU operate relatively mature space-weather services (NOAA's Space Weather Prediction Center, the ESA Space Weather Office). China has built up a parallel capability through the Chinese Meridian Project and the China National Space Administration's Fengyun satellite series. Coordination between these services is improving but is not yet routine; during a real Carrington-class event the question of whose forecast a given grid operator acts on is not a settled matter.
What to watch next
Three things are worth tracking over the next 12 to 24 months. First, whether APL's model is operationally adopted by NOAA's Space Weather Prediction Center or remains a research tool — the gap between a published physics model and a forecast that runs reliably on the operational cadence is large and historically slow to close. Second, whether the Federal Energy Regulatory Commission moves on the long-discussed rulemaking that would require NERC to set explicit geomagnetic-disturbance operating procedures for the bulk-power system; the technical case has been made for years, the regulatory plumbing less so. Third, whether the next solar maximum — now forecast to peak in 2025–2026 — produces an event large enough to test all of this in earnest.
The honest read is that the science is converging faster than the engineering, and the engineering is converging faster than the institutional plumbing. APL's million-mile bubble is a serious piece of work. It is not, on its own, what keeps the lights on when the next Carrington-class storm arrives. That will still come down to transformers, relays, operating procedures, and a forecast that a control-room operator in, say, PJM or MISO actually trusts enough to act on.
This article draws on a single research thread. Monexus reported the APL development in plain editorial language rather than the wire's framing in order to surface the grid-operations and international-coordination angles that the source material only gestures at.
Wire provenance
This editorial synthesis draws on the following public wire/social posts:
- https://www.swpc.noaa.gov/
- https://science.nasa.gov/heliophysics/