I mentioned, a while ago, that I’d been at a meeting and had an idea for a post. Well, this is my attempt to articulate what I thought at that meeting. A good deal of my own research involves trying to understand discs around very young stars, in particular discs that are quite massive, relative to the mass of the central protostar.For example, consider a young Sun with a disc that extends to just beyond where the edge of Solar System is today and that has a mass greater than 10% that of the central young Sun. Such a disc is likely to be susceptible to the growth of a gravitational instability, which can manifest itself as spiral density waves, as illustrated by the figure on right.
One reason why this is interesting is that it might provide a mechanism for transporting mass through the disc onto the central protostar. The waves grow and are then damped, which essentially acts to heat the disc by converting kinetic/rotational energy into thermal energy. In doing so, this must transport angular momentum, which has to go outwards, allowing mass to flow inwards. In fact, this might be the primary way in which mass is transported onto stars during the earliest stages of star formation.Now, there is another possible outcome; if the disc becomes very unstable, it might undergo what we call fragmentation. The spiral density waves start to form clumps that ultimately contract to form bound objects (see the Figure on the left). For a while this was regarded as a possible mechanism for the formation of gas giant planets, like Jupiter and Saturn.
Although I wasn’t the first, I was amongst one of the first to try to quantify the conditions under which a disc would fragment. Essentially, it would need to be susceptible to the growth of the gravitational instability and it would need to be able to cool rapidly. If it didn’t cool rapidly, the dense regions would heat up, and they would be unable to contract to form bound clumps.
Much of this work was done using numerical simulations. About 6 years ago, another group pointed out that if you ran some of these simulations at much higher resolution than we could have done, fragmentation occurred even when the discs cooled slowly. You might think that the immediate response would be to change our general view of this process. However, there had been a lot of other work to try to understand this process. People had shown that the wave amplitudes would depend on how fast the disc cools; if it cools slowly, the waves have a small amplitude. If the amplitude of the waves are small, how can they then form dense regions that clump to form bound objects? People had looked at the power spectra and shown that most of the energy was at scales that were well-resolved by the original simulations. If we were originally resolving these scales, why didn’t we see the fragmentation that was later seen in higher resolution simulations?
Also, if this process was easy, we’d expect to find many giant planets on wide orbits around stars, and we don’t. They do exist, but they are relatively rare. I also published a paper suggesting that if fragmentation were common, we’d expect to see some of the planets that form this way contaminating the known exoplanet population, and – again – we don’t see much evidence for this. So, there was immediately some suspicion, and most of the subsequent work has suggested that there is a numerical issue with some of the simulations, rather than it being likely that fragmentation can happen even if the discs cool slowly.
This isn’t definitive, but the general view – despite some uncertainty – is that if you want to form clumps in discs around young stars, you need a disc that is gravitationally unstable and one that is able to cool rapidly. Consequently, this process is unlikely to play a dominant role in the formation of giant planets.
The point I’m getting at is that science (research, really) can be complicated. It may be that some things challenge our understanding and so we have to consider the weight of the evidence and also whether or not there are reasons why there might be issues with this new evidence. Some amount of expert judgement is often required. There may be cases where the challenge is so convincing that you modify your understanding, and others where it appears that there are problems with the challenge, even if it’s not completely clear what it is.
Of course, it is always possible that our current understanding will change and we should be open to that as a possibility. However, we also shouldn’t change our views the instant someone comes along with some kind of challenge. Often, our current understanding builds up over time and is based on many lines of evidence. Challenging this is, of course, a key part of science, but it also requires that these challenges themselves stand the test of time. In fact, even if these challenges don’t overthrow the current paradigm, they still act to strengthen our understanding, because they force us to look more closely at all the lines of evidence.
The bottom line, though, is that if there appears to be some challenge to a scientific consensus that seems to be being ignored by the scientific community, it’s probably because they’ve looked at all of the evidence and concluded that, despite this challenge, the mainstream view is still probably broadly correct. It can be complicated, in other words.