Why we need a direct mass measurement! 
The g-2 Storage Ring is a unique facility for precision measurements that test the standard model. This proposal to improve the muon neutrino mass limit by a factor of 20 (from 170 keV down to 8 keV) will be the largest factor improvement made in any neutrino species mass within the last 25 years. It utilizes the storage ring as an extremely uniform spectrometer to compare the parent pion and daughter muon momenta from pion decay in flight. The experiment avoids multiple scattering limitations by transforming the momentum measurement into a position measurement; the edge of the decay muon radial distribution referenced on an event-by-event basis to the parent pion depends on the mass of the recoiling neutrino.
In order to put pions on orbit in the ring, 5.2 cm of beryllium degrades the pion energy. Silicon microstrip detectors on either side of the degrader record the position of the particle exiting through S1 and then entering the other side through S2 one cycle later. Because of the excellent magnetic field uniformity (better than 1 ppm), pions which do not decay end up back at the same position on the detector from which they started, no matter what their initial momentum and angle, whereas the daughter muons will cover a range of positions. The resulting radial distribution will consist of a large narrow peak at zero for undecayed pions and a broad distribution representing the decay products. The fact that the pions return to the origin makes the experiment relatively insensitive to the initial pion distribution. The width of the undecayed pion peak provides an in situ calibration of the orbital parameters and magnetic field uniformity.
Since the muons which decay in the forward direction actually have a higher momentum than the parent pion, they will traverse a larger diameter circle. Muons that are produced in the forward direction at a point halfway around the ring will therefore have the maximum radial displacement from the parent pion initial impact point when they return to the detector. In fact, they represent the maximum displacement of any daughter muon with any decay angle. Thus, the radial distribution of the decay muons referenced on a particle-by-particle basis to the radial position of their parent pions has a well-defined edge. If the neutrino has mass, this will reduce the energy of the forward-going muon and shrink the edge of the decay muon distribution by an amount which is sensitive to the neutrino mass, but relatively insensitive to the uncertainty in the pion mass.
Minnesota NuMass Collaborators