Most of IceCube is designed to look at high energy neutrinos, but there is a special case that I think is pretty interesting. If a certain type of supernova happens relatively near to Earth, we should be able to detect a burst of lower energy neutrinos hours before the supernova is visible to other kinds of telescopes.
In 1987, light from a supernova – an exploding star – in the Tarantula Nebula reached the Earth. From our perspective on Earth, it was the brightest supernova in a long time, and we haven’t seen a brighter one since. Lots of interesting astronomy and astrophysics research has been done on SN1987A, as it’s now known. In a sense, it was also the beginning of neutrino astronomy; a couple hours before the light from SN1987A reached Earth, detectors saw a burst of neutrinos!
This doesn’t mean that the neutrinos went faster than the light. SN1987A is what we call a core collapse supernova. We think that in this type of supernova, a star runs out of fuel, and collapses due to it’s own gravity. The collapse then causes a massive explosion in the centre of the star, which produces lots of neutrinos. Those neutrinos can travel out through the rest of the star mostly unimpeded, since they don’t generally interact with normal matter. On the other hand, light we see from a star comes from it’s surface, which means normal telescopes can only tell that a supernova has occurred when the explosion gets all the way out to the surface of the star.
Stars are so big (interesting graphic here) that the difference in time between the neutrino burst and the supernova light becoming visible can give us hours of warning before a supernova becomes visible. IceCube can detect this neutrino burst, and will automatically alert other observatories around the world through SNEWS if one occurs. Estimates I’ve heard are that there’s a 1/4 to 1/2 chance that we’ll see one of these events within the lifetime of IceCube – it would be big news.
Unfortunately, we can’t currently tell what direction one of these supernova neutrino bursts comes from. The problem is due to IceCube being designed to look at tera-electron volt (TeV) neutrinos, but the supernova neutrinos are only 10s of mega-electron volt (MeV). Basically, the lower energy – by a factor of about a million – means that supernova neutrinos will make much smaller tracks in the detector than it is designed to observe.
We think that there will be lots of these short tracks in the detector, if a supernova happens nearby. Rather than the normal operation where one TeV neutrino is “seen” by many DOMs (this is how we know what direction the neutrino was going), we would have a situation where many MeV neutrinos are simultaneously seen by many DOMs. People are currently looking into ways to enhance IceCube software to tease out the direction of the neutrino burst, but in the meantime at least we can alert people to “look up” if there’s about to be a visible supernova.