Strong geomagnetic storms can create harmful currents in power grids and pipelines, increase drag on low orbit satellites and disrupt satellite navigation. That is what NOAA states. For IT teams the useful question is not whether a storm will end civilization but which of your dependencies fail together.
What NOAA says
NOAA's Space Weather Prediction Center describes a geomagnetic storm as a major disturbance of Earth's magnetosphere. It happens when solar wind transfers energy into near-Earth space, especially when its magnetic field points south for many hours. The largest storms are linked to coronal mass ejections, which usually take days to arrive but can take as little as 18 hours.
Effects NOAA lists: harmful geomagnetically induced currents in power grids and pipelines, heating that expands the upper atmosphere and adds drag to low earth orbit satellites, and disruption of GNSS navigation systems.
The historical reference
The 1859 storm often called the Carrington event is the usual example. Per Wikipedia, telegraph systems across Europe and North America failed, with sparks and some fires at stations. The same page says a comparable storm today could cause widespread electrical disruptions and blackouts as well as damage to the power grid and satellites. Treat that as a plausible scenario described by that source, not a prediction of timing.
What it means for IT
Three dependencies are worth tracing.
- Power. Does your site, data center or office have backup power and a plan if the grid is down for hours or days?
- Time and location. Many systems use GPS time. See GPS timing: the hidden dependency in your stack.
- Connectivity. If your links run over satellites, what is the fallback?
A proportionate plan
- Write down which services need power, time and satellite links.
- Test failover to backup power on a schedule.
- Check that time sync has a second source that does not rely on GPS alone.
- Keep an offline copy of runbooks and contact lists.
- Follow NOAA alerts if you operate infrastructure that is sensitive to them.
None of this is exotic. It is the same continuity work you would do for any large outage, with one more cause on the list.
What to test, not just plan
Plans on paper hide gaps. Pick one dependency and test it. Switch to backup power during a quiet period and watch what fails to restart. Disconnect GPS from a time server in a lab and see how long clients stay accurate and what they do when it returns. Block a satellite link and check that failover works and alerts you.
Each test takes a few hours and produces a concrete fix list. It also gives you evidence for managers who ask whether the organization is prepared for a major outage of any cause.
Quick check
How much warning is there before a major solar storm reaches Earth?
NOAA says coronal mass ejections usually take days to arrive but can take as little as 18 hours.
Sources
Last reviewed: 19 October 2026.
