Coronal mass ejections are massive explosions of solar material that hurl billions of tons of plasma into space. When these magnetic clouds strike Earth, they trigger geomagnetic storms that can induce damaging electrical currents in power grids, degrade satellite communications, and expose astronauts to dangerous radiation levels. Forecasting exactly when these solar storms will hit is critical for mitigating their destructive impacts, but the technology to track them accurately has lagged for decades.
Until recently, scientists could only observe coronal mass ejections for about one-fifth of their journey from the Sun to Earth. Once the eruptions left the immediate vicinity of the Sun, they became invisible to existing instruments. Scientists were left to guess the trajectory, speed, and evolution of the solar storms for the remaining distance. This blind spot resulted in massive uncertainty windows in space weather prediction, leaving satellite operators and power grid managers unsure of exactly when to take protective action.
How the PUNCH Spacecraft Track Coronal Mass Ejections
Launched in 2025, the Polarimeter to Unify the Corona and Heliosphere (PUNCH) mission closed this critical observation gap. The NASA PUNCH mission relies on four small spacecraft flying in low Earth orbit. By working in tandem, these satellites provide continuous, three-dimensional observations of the inner solar system. The spacecraft capture a new image every four minutes, allowing scientists to routinely track solar explosions nearly all the way to Earth.
Researchers recently tested this tracking capability using data from a coronal mass ejection that erupted on May 31, 2025. They fed the continuous stream of PUNCH imagery into a computer model designed to analyze the leading edge of the solar storm. As the eruption moved and evolved across the inner solar system, the model used the storm’s speed and geometry to calculate exactly when it would reach our planet.
Twelve hours after the coronal mass ejection left the Sun, the computer model settled on a final prediction. It projected that the solar storm would arrive at Earth eight hours later. The predicted arrival time ultimately proved accurate to within 30 minutes. This precision represents a 10-fold improvement over currently used forecasting methods, which typically only provide a five-hour window of uncertainty.
The system also provided an added layer of reliability. The computer model itself revealed when its estimate had stabilized. This self-awareness means that a space weather forecaster would know exactly when to trust the prediction and issue alerts with confidence, rather than waiting for corroborating data.
A Leap in Solar Storm Forecasting
Researchers presented these initial proof-of-concept results Tuesday at the Committee on Space Research Scientific Meeting. The findings, which could revolutionize how Earth-impacting storms are forecasted, are currently under review at the journal Space Weather.
“We thought PUNCH would be good at this, but it’s a stunning result,” said Craig DeForest, principal investigator for PUNCH at Southwest Research Institute’s Solar System Science and Exploration Division in Boulder, Colorado. “To put it in perspective, this could be the space weather equivalent of going from a steam engine to a modern internal combustion engine.”
The breakthrough stems directly from the continuous stream of tracking data provided by the four satellites. Previous models relied on sparse snapshots, forcing scientists to make broad assumptions about how a storm behaved in the dark zones of the solar system.
“We accomplished an order of magnitude better result than the state-of-the-art method with a really basic process, just informed by the fact that the coronal mass ejection could be tracked continuously across the solar system,” DeForest explained. By simply watching the storm move, rather than guessing its path, the model achieved unprecedented accuracy.
New Insights Into Solar Plasma Behavior
Beyond revolutionizing solar storm forecasting, the high-resolution imagery from the NASA PUNCH mission is reshaping fundamental astrophysics. The wide-field imagery revealed new structures within coronal mass ejections. The data showed that clouds of material ejected from the Sun are far clumpier than scientists previously believed. The images also demonstrated that these coronal mass ejections continue to evolve dynamically as they cross the solar system, rather than remaining static blobs of plasma.
The mission is also clarifying how plasma—the superheated, magnetized solar material launched during these eruptions—moves through the vacuum of space. This localized data helps astrophysicists better understand plasma behavior on a galactic scale. The same physics governing solar storms apply to distant star-forming regions and nebulae, which are nearly impossible to study on small scales from Earth.
What Happens Next
As the NASA PUNCH mission continues to gather data over its operational lifespan, scientists expect to refine their computer models further. The research team believes that combining higher-fidelity PUNCH data with advanced machine learning algorithms could eventually push coronal mass ejection arrival forecasts even further in advance. This would give infrastructure operators, space agencies, and aviation regulators critical additional time to prepare for incoming space weather events.
Southwest Research Institute, based in San Antonio, leads the mission and operates the four spacecraft from its Boulder facilities. The mission is managed by Space Science Mission Operations at NASA’s Goddard Space Flight Center in Greenbelt, Maryland, for the agency’s Science Mission Directorate in Washington. With this operational framework firmly in place, the future of space weather prediction looks sharply focused.
— Priya Nair, science desk, AXO News