High energy particles, such as protons and helium nuclei, are continuously raining down on the earth from space. When these so-called “cosmic rays” collide with atoms in the atmosphere they produce a “shower” of many kinds of particles. These showers include pions, kaons, and muons, which produce neutrinos, known as “atmospheric neutrinos,” when they decay (Figure 1). Atmospheric neutrinos come in two types: electron neutrinos and muon neutrinos.
While most of a shower's particles are either absorbed by the ground or atmosphere, neutrinos can travel all the way though the earth since they very rarely interact. Super-Kamiokande was made as large as possible (50 kilotons) to observe the few that do interact. In fact, though there are roughly 100 atmospheric neutrinos passing through each square meter of the planet's surface every second, Super-Kamiokande observes only eight per day on average.
Since atmospheric neutrinos are produced all over the planet and neutrinos can easily pass though it, Super-Kamiokande observes neutrinos from all directions. Neutrinos that are produced above the detector (downward-going) travel a short distance, about 10 km, while those that are produced below it (upward-going) on the other side of the earth travel more than 13,000 km (Figure 2) to the detector.
If neutrinos have no mass, as was once thought, one would expect to observe about the same number of upward-going and downward-going atmospheric neutrinos. While this is precisely the case for Super-Kamiokande's electron neutrino data (Figure 3, left) there is a large deficit of upward-going events observed in the muon neutrino data (Figure 3, right). At the same time the downward-going events agree with the expectation. Muon neutrinos that travel long distances before reaching the detector seem to be disappearing!
We now know that this disappearance is due to “neutrino oscillations,” a phenomenon in which massive neutrinos change their type mid-flight. In this case the upward-going muon neutrinos have oscillated into a third neutrino type, the tau neutrino. This result indicates that neutrinos have mass. Downward-going muon neutrinos, on the other hand, interacted inside of Super-K before they had traveled far enough to change types.
So how do we know that muon neutrinos have oscillated into tau neutrinos? Actually its possible to search for signs of tau neutrino interactions inside of Super-K. Since there are no tau neutrinos to start with, finding evidence of tau neutrino interactions would indicate they came from oscillations.
However, tau neutrino interactions are even more rare than those of the other neutrinos and they often produce many particles, which makes them hard to distinguish from those other neutrinos. For this reason Super-K searches for tau neutrino-like interactions in the data (Figure 4). The colored part of the histogram shows the number of events that are attributed to tau neutrino interactions. Notice that the upward-going data, shown as circles, would not agree with the prediction without these events. Also, notice that there is no sign of tau interactions in the downward-going data. Tau neutrinos are appearing just where the muon neutrinos disappeared!
There are still many things about neutrinos which are currently unknown but that can learned by studying using neutrino oscillations. For instance, though it is known that there are three massive neutrinos, it is not known which of the three is the heaviest. Neutrino oscillations depend on the difference of the neutrino masses (more precisely the difference of the squared masses), and the sign of that difference changes the number of events expected at Super-Kamiokande. So it is possible to determine which neutrino is heaviest by studying oscillations with more atmospheric neutrino data.
Atmospheric neutrinos can also be used to study the effects of matter on neutrino oscillations. It is thought that electron neutrinos should undergo additional interactions with matter that are not felt by the other neutrinos as they travel through it. Since the earth's core is made of very dense material ( 13 g/cm3 ) with a lot of electrons the oscillations of neutrinos that pass through it are expected to be enhanced when those neutrinos have certain energies. This enhancement has not yet been seen, but is of great interest since it is closely related to the neutrino mass ordering.
In addition, atmospheric neutrinos can provide insight into the behavior of neutrinos and their antiparticles, antineutrinos, both of which are observed at Super-Kamiokande. If the behavior of neutrinos and antineutrinos is found to be different, it would have profound impact on our understanding of the universe, its origins, and evolution.
Precise measurements of the atmospheric neutrino spectrum itself is another research topic with particular importance, not only in the understanding cosmic rays, but also for the characterization of backgrounds to searches for rare processes at Super-Kamiokande, like the decay of protons.
Figure 5 shows the atmospheric neutrino spectrum measured by Super-Kamiokande and other experiments. Solid lines show the prediction from detailed computations of the expected spectrum, which is in good agreement with the data.
To go even further, Figure 6 shows the horizontal directional distribution of atmospheric neutrino events together with the expectation. Due to the presence of the earth's magnetic field the number of neutrinos from the west is expected to differ from the number from the east. Though neutrinos themselves are neutral particles and have no electric charge, their parent cosmic rays (mostly protons) are charged and feel the effects of the magnetic field. That the data and prediction agree so well suggests that our modeling of the processes that create atmospheric neutrinos is correct. Viewed another way, we can be sure that the neutrinos we are observing are in fact atmospheric neutrinos.