If you want to know how likely we are to find a ninth planet lurking at the edge of our solar system, it is worth considering hunches elsewhere in science. When discovery relies on hope, inference and statistics, we are easily fooled.
We have been here before. In 1983, when Pluto was still the ninth planet and any new discovery could legitimately be given the moniker “Planet X”, researchers operating the Infrared Astronomical Satellite (IRAS) thought they had seen something big. In 1990, another search began: the suspicion was that a tenth planet might be found in the constellation Centaurus. This was prompted by computer simulations that suggested the existence of a planet that was twice the size of Earth, ten billion kilometres from the sun.
In 2002, researchers used computer modelling to investigate what had made a suspicious area of empty space in the otherwise teeming Kuiper Belt, a region crammed with small rocks, orbiting at the edge of the solar system. According to the computer, the empty region could have been created when some of those rocks agglomerated to form a planet the size of Earth or Mars.
Two years ago, anomalies in the orbits of some objects in the Kuiper Belt led astronomers to claim that two planet-sized objects could be out there: one with a mass ten times that of Earth and another that was even bigger. So far, observations by everything from the IRAS satellite to Nasa’s Wide-Field Infrared Survey Explorer telescope have ruled out a Planet X. Yet the computers say that we should look again.
The latest suggestion comes from simulations of the solar system’s birth. This was a time of planetary violence, with collisions between proto-planets creating debris, swept up to form new bodies. The collisions that formed Uranus and Neptune would also have created the cores of planets up to ten times the size of Earth, slung by gravitational effects out of the inner solar system. It is possible that one of these could have moved into an orbit in the outer solar system. Given the anomalies in some of the orbits of the Kuiper Belt objects, it seems worth a look. Hence the new excitement about Planet X – though after Pluto’s demotion, it would be Planet Nine.
None of this speculation is wasted: it has worked before. We found Neptune because of irregularities in Uranus’s orbit around the sun. Pluto was discovered in the search for Planet X. However, although anomalies and computer models occasionally lead us to a discovery, they can also be notoriously misleading.
It is not just a problem in astronomy. Anomalies in the data at Cern give “hints” of new, undiscovered particles far more often than those particles are discovered. In medicine, understanding the “false discovery rate” saves innumerable blushes.
So how long will we keep searching for Planet X? A paper published last year by Thomas Fanshawe of Oxford University gives us a reliable hint. It looked at the way in which our tenacity in looking for something – whether for weapons in airport baggage, cancers in X-rays, or cracks in road bridges – depends on the expectation of finding it.
Of particular relevance is the “law of increasing returns”: those who believe that the object is there assume that the longer the search continues, the more ground has been covered, increasing the chance of imminent discovery. Judging by the story so far, the search for Planet X (or Nine) will go on for a good while yet, with ever higher hopes.
Content from our partners
Related
This article appears in the 27 Jan 2016 issue of the New Statesman, Should Labour split?