The MiniBooNe wave swept through the blogosphere a week ago, right after first results were published by the collaboration. Here at CERN, however, we take our time ;-) Only today, the local neutrino gurus organized a discussion, which i am gladly reporting. The delay is convienent for me, as i don't need to explain the basics. The physics of neutrino oscillations and the kitchen of the MiniBooNE experiment are explained at length here on Cosmic Variance and here on Tommaso's blog. I'm going to concentrate on the most interesting thing, which is theory. On the impact of the MiniBooNE results on the theory of neutrino masses, to be precise.
The MiniBooNE experiment was looking for an appearence of electron neutrinos in a beam of muon neutrinos. The main objective was to confirm or exclude the signal of the earlier experiment called LSND, which had claimed to observe oscillations of muon anti-neutrinos into electron anti-neutrinos. If the LSND signal is real, we need at least one new neutrino species in addition to the Standard Model particles. MiniBooNE did not find any oscillation signal for neutrino energies larger than 475 MeV. The negative result means that the simple two-neutrino-oscillation interpretation of the LSND data is no longer available. As seen on the plot above, MiniBooNE finds some excess of electron neutrinos at smaller energies. However, in this energy range the expected background is large. For the moment, it is not clear if the excess is due to new physics or simply due to an underestimation of the background.
Thomas Schwetz was discussing which theoretical scenarios are able to accommodate the LSND and MiniBooNE data, together with the abundant data from other thousand neutrino experiments. His message is that it is tough. In fact, it was tough even before MiniBoone. Explaining the LSND signal requires introducing at least one sterile neutrino (sterile, means not coupled to the Z boson, unlike the three Standard Model active neutrinos). The problem is that the mixing angle between sterile and active neutrinos required to fit the LSND signal must be rather large. This is at odds with the negative results of certain experiments like CHOOZ and BUGEY, which searched for neutrino disappearance.
Thomas presented his preliminary analysis of the situation after MiniBooNE. Fits to all neutrino data in 3 active + 1 sterile scenarios become even worse than before. The fits in 3 + 2 scenarios do not change much. If one allows for CP violation, the \chi^2 is around 100 (with about 100 degrees of freedom). According to Thomas, this is discouraging. According to astrophysicists, this is an indication for 3+2 scenarios ;-) What is strange, nobody's willing to plunge into 3 + 3 scenarios. Chicken?
There are more exotic models proposed to accommodate the LSND results. Thomas mentioned three:
- Sterile neutrino with CPT violation, proposed by Vernon Barger & Co. The LSND signal could be more easily accommodated if neutrinos and anti-neutrinos had different masses. This violates CPT, the hallmark of local relativistic quantum theories. It seems that this scenario is not much affected by MiniBooNE, who studied neutrinos rather than anti-neutrinos.
- Decaying sterile neutrinos, proposed by Thomas Schwetz in person. Here, the LSND signal is a muon anti-neutrino oscillating into a sterile one, the latter decaying to an electron anti-neutrino and an exotic scalar. This model predicts a signal in MiniBooNE which is not there. The model can be saved, however, by introducing another sterile neutrino and CP violation in the decays.
- Shortcuts in extra dimensions, proposed by Heinrich Paes & Co. Here, the sterile neutrino propagates in extra dimensions and travels a different path than the active ones, which changes the oscillation pattern. By shaping the extra dimensions properly, the model can accommodate the MiniBooNE excess at lower energies as well as the suppression of oscillations at higher energies. My impression is that, with a little bit more work, the model could accommodate Harry Potter, too.
If you're interested in my opinion, I think the truth is boring. The LSND signal is fake, while the miniBooNE excess will go away after re-examining the background. The theory of neutrino oscillations will remain the theory of a 3x3 matrix.
Everything about neutrinos can be found on this page, updated in real-time by Allessandro Strumia.