Dark matter is that weird stuff that is supposed to make up about 80% of all matter in the universe. It is invisible, and pretty much non-interacting (you can’t see it, and you can’t detect it (easily) by other means).
One way you might detect it is to take a very pure crystal and look for tiny flashes of light inside the crystal caused by a dark matter particle smacking into one of the atoms in the crystal. Nice idea, but insanely hard to do. For a start, natural radioactivity inside your crystal or your light detector will cause many more spurious flashes than there will be ‘real’ (caused by a head on collision of dark matter and an atom) flashes. You can then throw into that all the cosmic rays and particle showers caused by cosmic rays – these further add to the noise swamping the precious flashes of light signaling a dark matter hit (and your Nobel prize…)
You can get away from the spurious flashes somewhat, by building your crystal out of the most pure material you can get – filtering out a lot of the natural radioactivity, and then running the experiment deep underground – or in the centre of a local mountain – to filter out much of the cosmic ray noise.
What you should see now are random flashes of light, some of which might be caused by dark matter impacts, and the rest caused by all the crud you haven’t managed to filter. What you need now, is a way to differentiate the signal (dark matter) from the noise (radioactive crud).
The Sun and the solar system are moving (at around 250 km/sec) though a background of dark matter that fills the galaxy. The earth circles the sun at around 30km/sec, so for half of the year the velocity of the earth though the dark matter is 250+30 km/sec and for the other half of the year, the figure is 250-30 km/sec. This difference in velocity of the earth through the DM background should show up as a difference in the number of flashes of light in your detector – you hit more DM when you plow though it faster than when you move though it more slowly.
This rather stunning graph seems to show just that.
Ignoring the first part of the graph, where the detector wasn’t working at maximum sensitivity (so from 0 to about 3000 days on the bottom axis) and concentrating on the rest of the data, it really looks like the error bars show a sinusoidal variation with time – this is what the black line is fitting to. The data goes up and down once per year.
This type of variation is exactly what you’d expect from differences in the number of particles hit as the earth change velocity (speed and direction) though the DM background.
So, have they found Dark Matter? Or do they have another bug in the detector? Too soon to call on it yet, but this is very interesting.