The NOνA (NuMI Off-Axis νe Appearance) experiment is a particle physics experiment designed to detect neutrinos in Fermilab's NuMI (Neutrinos at the Main Injector) beam. Intended to be the successor to MINOS, NOνA consists of two detectors, one at Fermilab (the near detector), and one in northern Minnesota (the far detector). Neutrinos from NuMI pass through 810 km of Earth to reach the far detector. NOνA's main goal is to observe the oscillation of muon neutrinos to electron neutrinos. The primary physics goals of NOvA are:^[1]

Neutrino oscillation is parameterized by the PMNS matrix and the mass squared differences between the neutrino mass eigenstates. Assuming that three flavors of neutrinos participate in neutrino mixing, there are six variables that affect neutrino oscillation: the three angles θ12, θ23, and θ13, a CP-violating phase δ, and any two of the three mass squared differences. There is currently no compelling theoretical reason to expect any particular value of, or relationship between, these parameters.

The neutrino masses and mixing angles are, to the best of our knowledge, fundamental constants of the universe. Measuring them is a basic requirement for our understanding of physics. Knowing the value of the CP violating parameter δ will help us understand why the universe has a matter-antimatter asymmetry. Also, according to the Seesaw mechanism theory, the very small masses of neutrinos may be related to very large masses of particles that we do not yet have the technology to study directly. Neutrino measurements are then an indirect way of studying physics at extremely high energies.^[4]

In our current theory of physics, there is no reason why the neutrino mixing angles should have any particular values. And yet, of the three neutrino mixing angles, only θ12 has been resolved as being neither maximal or minimal. If the measurements of NOνA and other future experiments continue to show θ23 as maximal and θ13 as minimal, it may suggest some as yet unknown symmetry of nature.^[4]

NOνA can potentially resolve the mass hierarchy because it has a very long baseline. It is the only experiment likely to run in the near future that has this ability. Many future experiments that seek to make precision measurements of neutrino properties will rely on NOνA's measurement to know how to interpret their results.

An experiment similar to NOνA is T2K, a neutrino beam experiment in Japan similar to NOνA. Like NOνA, it is intended to measure θ13 and δ. It will have a 295 km baseline and will use lower energy neutrinos than NOνA, about 0.6 GeV. Since matter effects are less pronounced both at lower energies and shorter baselines, it will be unable to resolve the mass ordering.^[5]

Neutrinoless double beta decay experiments will also benefit from knowing the mass ordering, since the mass hierarchy affects the theoretical lifetimes of this process.^[4]

Reactor experiments also have the ability to measure θ13. While they cannot measure δ or the mass ordering, their measurement of the mixing angle is not dependent on knowledge of these parameters. One such experiment is Daya Bay, located at the Daya Bay reactor in southern China, which uses a 2 km baseline optimized for observation of the first θ13-controlled oscillation maximum.^[6]

In addition to its primary physics goals, NOνA will be able to improve upon the measurements of the already measured oscillation parameters. NOνA, like MINOS, is well suited to detecting muon neutrinos and so will be able to refine our knowledge of θ23.

The NOνA near detector will be used to conduct measurements of neutrino interaction cross sections which are currently not known to a high degree of precision. Its measurements in this area will complement other similar upcoming experiments, such as MINERνA, which also uses the NuMI beam.^[7]

Since it is capable of detecting neutrinos from galactic supernovas, NOνA will form part of the Supernova Early Warning System. Supernova data from NOνA can be correlated with that from SuperKamiokande to study the matter effects on the oscillation of these neutrinos.^[4]

To accomplish its physics goals, NOνA needs to be efficient at detecting electron neutrinos, which are expected to appear in the NuMI beam (originally made only of muon neutrinos) as the result of neutrino oscillation.

Previous neutrino experiments, such as MINOS, have reduced backgrounds from cosmic rays by being underground. However, NOνA is on the surface and relies on precise timing information and a well-defined beam energy to reduce backgrounds. It is situated 810 km from the origin of the NuMI beam and 14 milliradians (12 km) west of the beam's central axis. In this position, it samples a beam that has a much narrower energy distribution than if it were centrally located, further reducing the effect of backgrounds.^[4]

The detector is designed as a pair of finely grained liquid scintillator detectors. The near detector is at Fermilab and samples the unoscillated beam. The far detector is in northern Minnesota, and consists of about 500,000 4 cm × 6 cm × 16 m cells filled with liquid scintillator. Each cell contains a loop of bare fiber optic cable to collect the scintillation light, both ends of which lead to an avalanche photodiode for readout.

The near detector has the same general design, but is only about 1/200 as massive. This 222-ton detector is constructed of 186 planes of scintillator-filled cells (6 blocks of 31 planes) followed by a muon catcher. Although all the planes are identical, the first 6 are used as a veto region; particle showers which begin in them are assumed to not be neutrinos and ignored. The next 108 planes serve as the fiducial region; particle showers beginning in them are neutrino interactions of interest. The final 72 planes are a "shower containment region" which observe the trailing portion of particle showers which began in the fiducial region. Finally, a 1.7-meter long "muon catcher" region is constructed of steel plates interleaved with 10 active planes of liquid scintillator.

The NOνA experiment includes scientists from a large number of institutions. Different institutions take on different tasks. The collaboration, and subgroups thereof, meets regularly via phone for weekly meetings, and in person several times a year. Participating institutions as of April 2018 are:^[8]

In late 2007, NOνA passed a Department of Energy "Critical Decision 2" review, meaning roughly that its design, cost, schedule, and scientific goals had been approved. This also allowed the project to be included in the Department of Energy congressional budget request. (NOνA still required a "Critical Decision 3" review to begin construction.)

On 21 December 2007, President Bush signed an omnibus spending bill, H.R. 2764, which cut the funding for high energy physics by 88 million dollars from the expected value of 782 million dollars.^[9] The budget of Fermilab was cut by 52 million dollars.^[10] This bill explicitly stated that "Within funding for Proton Accelerator-Based Physics, no funds are provided for the NOνA activity in Tevatron Complex Improvements."^[11] So although the NOνA project retained its approval from both the Department of Energy and Fermilab, Congress left NOνA with no funds for the 2008 fiscal year to build its detector, pay its staff, or to continue in the pursuit of scientific results. However, in July 2008, Congress passed, and the President signed, a supplemental budget bill,^[12] which included funding for NOνA, allowing the collaboration to resume its work.

The NOνA prototype near detector (Near Detector on Surface, or NDOS) began running at Fermilab in November and registered its first neutrinos from the NuMI beam on 15 December 2010.^[13] As a prototype, NDOS served the collaboration well in establishing a use case and suggesting improvements in the design of detector components that were later installed as a near detector at Fermilab, and a far detector at Ash River, MN (48°22′45″N 92°49′54″W ^[20] ).

Once construction of the NOvA building was complete, construction of the detector modules began. On July 26, 2012 the first module was laid in place. Placement and gluing of the modules continued over a year until the detector hall was filled.

The first detection occurred on February 11, 2014 and construction completed in September that year. Full operation began in October 2014.^[14]