A new computer simulation of the Milky Way, which takes into account the gravity of stars and gas as well as the gravity of dark matter, predicts that a disc of dark matter exists within the plane of the Milky Way. We asked lead scientist Justin Read about his research.

How do your computer simulations allow you infer the presence of a dark matter disc, and how does this compare to the idea of dark matter haloes?
We now have a very good idea of what the Universe looked like just a few hundred thousand years after the Big Bang. At this time the Universe cooled enough to become transparent to photons, and we see these relic photons today as the cosmic microwave background radiation (CMB). The faint temperature fluctuations in the CMB allow us to measure the matter fluctuations in the Universe at these early times. This tells us already that most of the mass in the Universe is not the atoms that we are made up of, but something else: dark matter.

Super-computer simulations allow us to take these initial matter fluctuations and model their subsequent evolution as they collapse under gravity and form the galaxies and clusters of galaxies that we see today. So far, almost all simulations modelled just the dark matter alone. This makes sense since the dark matter comprises most of the mass of the Universe. But also dark matter is much easier to model than normal atoms because it only interacts via gravity. In our paper, we made the first attempt to model also the stars and gas in the Universe -- the atoms from which we are made up -- and their effect on the local distribution of dark matter in our Galaxy.

In dark matter only simulations of the Universe the "galaxies" that form are roughly spherical (really cigar-shaped). These are the dark matter "haloes" and visible galaxies should reside within them. Although there is much less normal matter in the Universe than dark matter (about 1/6th of the matter in the Universe is in atoms), the normal matter -- unlike dark matter -- can cool down by radiating photons (this is one of the reasons it is more difficult to simulate!). By cooling down, the gaseous normal matter settles into the very centres of the dark matter haloes. In our Galaxy, for example, although most of its total mass is dark matter, interior to the orbit of our Sun (about 8kpc from the Galactic centre) actually most of the mass is in stars and gas, not dark matter. The dark matter dominates the mass only further out -- some 20-50kpc from the Galactic centre. This means that at our local position within the galaxy, the stars and the gas will have a strong influence on the local distribution of dark matter and this is what we set out to measure.

In the super-computer simulations, we observe that dark matter haloes form hierarchically through the successive mergers of smaller halos. Without stars and gas, these mergers occur with random orientations and this is why the haloes are roughly spherical. The gas, however, cools down and settles into a disc (this is why many galaxies are disc galaxies, like our own Galaxy). Once the gas disc has formed, it gradually forms stars producing the galaxies that we see in the Universe today. The key point about our new work is that we realised that this disc of gas and stars will affect the subsequent accretion of the dark matter haloes (each of which contains its own little galaxy). The haloes are now preferentially dragged towards the disc plane. As they are torn apart by tidal forces, their material then settles into a disc-like structure -- this is the dark matter disc. We see precisely this effect in our simulations, which we use to predict how massive/important this dark matter disc should be. It is important to stress that only those haloes that actually 1) get close enough to hit the stellar/gas disc and 2) are oriented close to the disck plane are strongly affected. For this reason, the global structure of the dark matter halo is still roughly spherical, which agrees well with the simulations that model the dark matter alone. But locally -- i.e. around the Earth and the Sun -- there is this focusing effect and so extra to the standard dark matter halo there is a dark matter disc. If we live in a Universe dominated by dark matter, then the dark matter disc is a natural consequence of such a theory and there must be one in our own Galaxy.

Can dark matter discs and haloes co-exist?
Yes. The haloes are still correct globally. But on smaller scales -- where the Earth and the Sun live -- there is an extra dark matter component in addition to the roughly spherical halo -- the dark matter disc.

What is XENON100 how will it detect dark matter?
Its goal is to try and directly detect dark matter, under the assumption that dark matter is some new fundamental weakly interacting massive particle, or WIMP. It is important for all experiments like XENON to have some idea of how many of these dark matter particles are flowing through the Earth. Firstly, this allows them to design their detectors to be able to better find dark matter. But secondly, if they do actually detect something in the future, it will allow them to understand the signal. Perhaps this analogy will help: Imagine I stand people in a line each 1 metre apart. I then have another group of people run towards the first. They must run in straight lines and start out at random separations. Every now and then the people will bump into one another -- we can think of this as a rare "dark matter detection". Now I can double the number of "detections" (people bumping into each other) either by having twice as many runners, or by having twice as many people standing in the line. Twice as many runners is like changing the amount of dark matter flowing through the Earth. Twice as many people standing still is like changing the dark matter itself. Thus, we need to know how many dark particles are flowing through the Earth in order to then work out what dark matter is. This is where the dark disc becomes important. It rotates almost as fast as the Earth and so, from our point of view, the WIMPs from the dark disc are moving very slowly compared to those from the dark halo [the Earth and Sun are rotating about the centre of our Galaxy at some 220km/s, while the dark halo is roughly stationary]. This source of low velocity WIMPs are interesting because low velocity WIMPs excite a much larger response in a detector than fast moving ones. To take the above analogy too far, it's like the slow moving ones have much more time to give everyone a big push as they walk, rather than run, past. If we expect many of these slow WIMPs, then it would be good to design detectors that are particularly sensitive to them -- increasing our chance of really detecting dark matter for the first time. In this sense XENON100 is exciting because it will be better able than ever before to detect these slow WIMPs. If it sees something, but only at low energy (i.e. low WIMP velocity), this would then be the smoking gun both for dark matter and for the dark matter disc.