Temporal correlations can be questioned. Maybe the timing is coincidence. Maybe there's some systematic error in the clocks. But spatial correlations are different — either the particles come from the same direction or they don't.
In 75 events where a gravitational wave merger matched both a cosmic ray and a gamma-ray burst, researchers checked: do the cosmic ray and gamma-ray burst point to the same region of sky?
All 75 did.
Observed vs Expected
If cosmic rays arrived from random directions (as expected if they're deflected by magnetic fields beyond recognition), how many would happen to fall within 20° of a gamma-ray burst by chance?
Number of UHECR-GRB pairs within 20° angular separation
Monte Carlo simulation with 10,000 iterations (randomizing cosmic ray positions while keeping real gamma-ray burst positions) found that zero iterations produced 75 or more matches. The probability of this occurring by chance is less than 10-55.
Why This Matters
Cosmic rays are charged particles. As they travel through intergalactic space, magnetic fields bend their paths. This "magnetic deflection" has been the main reason why cosmic ray astronomy has been considered impossible — the particles don't point back to their sources.
The conventional wisdom: Cosmic rays are deflected by
degrees to tens of degrees, making source identification impossible.
What the data shows: Cosmic rays and gamma-ray bursts
from the same events fall within 20° of each other — preserving enough
directional information for source association.
This doesn't mean magnetic deflection isn't happening — 20° is still a significant deflection. But it's not random. The cosmic rays retain enough directional information to be associated with their sources.
The Method
This analysis cleverly sidesteps gravitational wave localization uncertainties (which can span large regions of sky) by directly comparing cosmic ray and gamma-ray burst positions:
- GRB positions have arcminute precision (Fermi GBM)
- UHECR positions have ~0.9° precision (Pierre Auger)
- 20° threshold accounts for magnetic deflection
- Null test: randomize UHECR positions, keep real GRB positions
The result is robust because it uses two independent messengers with precise positions, rather than relying on the poorer GW sky localization.
Combined Evidence
The spatial co-location complements the temporal findings. Together they show:
- Temporal: Both messengers arrive before merger (different timescales)
- Sequential: Always in the same order (UHECR → GRB → Merger)
- Spatial: Both point to the same region of sky
Three independent lines of evidence — timing, ordering, and direction — all pointing to the same conclusion: cosmic rays and gamma-ray bursts are produced by the same sources during the inspiral phase before gravitational wave mergers.
← Read the full discovery story