They arrived at that cause and date by modeling the birth of the universe. The researchers plugged stars and data into the computer simulation Synthetic Universe and ran it until they encountered an event that would possibly have emanated the actual gravitational waves. In this case, the culprit was likely a recent merger of two black holes made soon after the dawn of the universe.
"We play God," lead study author Chris Belczynski, an astrophysicist at Warsaw University, told The Verge. "We have a model of the entire Universe in our computer. We populate the computer with stars from the beginning, from the Big Bang, and you let them go ahead, evolve, produce black holes, etc."
Since Synthetic Universe's simulation also includes a mock-LIGO to chronologically sync when we detected the waves, the model is also predictive, the study argues. If correct, we should see LIGO pick up to 60 detections when it starts "listening" for the waves again in fall. At its peak sensitivity, it could hear up to 1,000 detections annually.
Belczynski's speculation specifies the size of black hole mergers that the LIGO should be able to detect from gravitational waves, a combined mass between 20 and 80 times the mass of our sun. That large size indicates that they're likely from just after the Big Bang, when stars had lower metal content and formed proportionately larger black holes.
Belczynski's model strongly suggests that the ones that collided to make these gravitational waves were stars that formed 12 billion years ago, became black holes 5 million years later, and then merged 10.3 billion years after that. 1.2 billion years later LIGO detected those reverberations in space-time. As more data comes in to LIGO and other detectors from these gravitational waves, the more Belczynski can refine his Synthetic Universe model and theorize the life cycles of stars.