The mystery behind the origin of quasars – the brilliantly bright centres of distant galaxies – may have been solved by a group of New York astronomers who suggest that giant clouds of hydrogen gas falling onto the centres of galaxies are being consumed by the supermassive black holes lurking within.

The majority of quasars (also called active galactic nuclei, or AGN) are found to have existed during the first four or five billion years of the Universe’s history, during an epoch in which galaxies were being assembled by giant clouds of gas. Now Barry McKernan and K E Saavik Ford of the American Museum of Natural History and City University of New York, and Ariyeh Maller also of City University, have put the two together.

“For a while, people have known that gas clouds are falling onto galaxies, and they’ve also known that active galactic nuclei are powered by gas falling onto supermassive black holes,” says McKernan. “But no one put the two ideas together until now and said, ‘Hey, maybe one is causing the other!’”

These gas clouds are vast, millions of times more massive than the Sun. As they fall onto a galaxy they generate large amounts of star formation, but they are also the ideal mechanism for getting gas into galactic centres to provoke quasar activity. Previous theories had suggested that galactic bars could funnel gas into the centre to feed the black hole, or that the gravitational effects of mergers could divert large quantities of gas towards the black hole. However, if mergers were the main trigger we’d see quasars wherever we see lots of galaxies merging, and similarly if bars were responsible we would see quasars only in barred spirals, but we see them in other types of galaxy too. The new model has the advantage, says McKernan, of already being known to happen and being able to provide enough gas. For instance, a full blown quasar gobbles up thousands of solar masses per year. And during the galaxy building epoch there were plenty of these gas clouds around, which were closely linked to star-formation rates.

“We assume that the rate of cloud impact goes as the star-formation rate,” McKernan tells Astronomy Now. “In other words, at a redshift of 1 [about seven billion years ago], we’d expect about ten times the number of impacting clouds than in the present era. At a redshift of 2 [about ten billion years ago, when star formation hit its peak rate] you’d expect maybe 100 times the number of impacting clouds.” Because all galaxies are believed to have been constructed in the same process, including our Milky Way, then at one time all galaxies will have been quasars.

Some of these clouds still float around today, flotsam and jetsam leftover from the quasar epoch. “The leftover clouds are still bombarding galaxies, including ours, occasionally,” says McKernan. “Every so often one of these clouds must hit the centre and light things up.”

However, this would not create a fully-fledged quasar in the modern era. Seyfert galaxies, which have low power AGN, consume an amount of gas equivalent to about half the mass of the Sun per year. Even lower level AGN activity has been seen in galaxies, where the central black hole gobbles up, at most, the equivalent of just one percent of the mass of the Sun per year. McKernan believes this low level activity could be blamed on individual rogue clouds striking at random.

“However, in rare cases, our mechanism could produce a train of impacting clouds (rather like the impact of Shoemaker-Levy 9 with Jupiter), which might be enough to promote quasar-like activity,” suggests McKernan. But if more than ten million years passes between galaxies being hit, their centres will appear normal, and so such activity is relatively rare in the local Universe.