It revolutionized our view of the Universe when the LIGO announcements – and we’re up to three, now – came out. They indicated the direct detection of gravitational waves from merging black holes, teaching us about a new population of stellar remnants, confirming the existence of gravitational waves, and showcasing yet another victory for Einstein’s General Relativity. But it all rests on one critical assumption: that what LIGO detected was a gravitational wave signal, not just noise in the detector.
A critical test of this is whether the noise is truly random between detectors, as one would expect, or whether the noise is somehow correlated between the detectors, which would run contrary to expectations. An independent team from Denmark, outside of the LIGO collaboration, put this idea to the test, and what they found has cast significant doubts on the LIGO results.
On September 14, 2015, LIGO directly detected gravitational waves for the first time, from the in spiral-and-merger of two quite massive black holes. Despite the fact that no electromagnetic radiation signal was expected, the Fermi GBM instrument measured a high-energy X-ray event just 0.4 seconds after LIGO’s 200 millisecond detection occurred. According to NASA scientists working on the Fermi mission, there was just a 0.2% chance of a false positive.
In a new paper out today, however, a reanalysis using a superior statistical method shows that the 0.2% chance is, in fact, true, and that LIGO’s black holes didn’t have an electromagnetic counterpart after all. This new statistical technique will be incredibly useful down the road for discriminating between robust astrophysical signals and spikes in the noise that have hitherto been a source of incredible confusion.