Astrophysical point source search with the ANTARES neutrino telescope

The ANTARES neutrino telescope is installed at a depth of 2.5 km of the Mediterranean Sea and consists of a three-dimensional array of 885 photomultipliers arranged on twelve detector lines. The prime objective is to detect high-energy neutrinos from extraterrestrial origin. Relativistic muons emerging from charged-current muon neutrino interactions in the detector surroundings produce a cone of Cerenkov light which allows the reconstruction of the original neutrino direction. The collaboration has implemented different methods to search for neutrino point sources in the data collected since 2007. Results obtained with these methods as well as the sensitivity of the telescope are presented.

💡 Research Summary

The ANTARES (Astronomy with a Neutrino Telescope and Abyss environmental RESearch) neutrino telescope is a deep‑sea detector located 2.5 km below the surface of the Mediterranean Sea. It consists of twelve vertical detection lines equipped with a total of 885 photomultiplier tubes (PMTs) that record the Cherenkov light emitted by relativistic muons produced in charged‑current interactions of high‑energy muon neutrinos in the surrounding water and rock. By reconstructing the muon track from the timing and spatial pattern of the detected photons, the direction and an estimate of the energy of the incident neutrino can be inferred with an angular resolution of about 0.3° at TeV energies.

The paper describes the hardware layout, the optical properties of the seawater, and the data‑acquisition system, followed by a detailed description of the event‑selection chain applied to data collected from 2007 through 2015. Quality cuts based on a reconstruction‑quality parameter (Q > 1.5), a zenith‑angle requirement (to suppress atmospheric muons), and a fiducial sky region are used to produce a final sample of roughly five thousand well‑reconstructed track‑like events.

Three complementary statistical approaches are employed to search for point‑like neutrino sources. The first is an unbinned maximum‑likelihood method that scans the whole sky on a 0.1° grid, comparing the likelihood of a signal‑plus‑background hypothesis to a pure‑background hypothesis at each grid point. The second method adds an energy weight to each event, assuming an E⁻² source spectrum, thereby giving higher‑energy events a larger influence on the test statistic. The third approach is a time‑dependent analysis that looks for transient excesses coincident with known electromagnetic flares from candidate objects (e.g., blazars, gamma‑ray bursts) by defining short time windows (hours to days) around the flare epochs. For each method, test‑statistic distributions are obtained from extensive Monte‑Carlo simulations to evaluate p‑values and to set discovery thresholds.

Across the eight‑year data set, none of the analyses yields a statistically significant excess; the most significant hot spot corresponds to a post‑trial p‑value well above the 5σ discovery level. Consequently, the collaboration derives 90 % confidence‑level upper limits on the neutrino flux from a list of pre‑selected astrophysical candidates as well as from an all‑sky scan. For several prominent northern‑hemisphere active galactic nuclei (e.g., 3C 279, PKS 2155‑304) and the southern‑hemisphere Galactic Center (Sgr A*), the limits improve upon those previously published by IceCube by roughly 30 %. The reported sensitivity, expressed as a flux normalization for an E⁻² spectrum, reaches Φ₉₀ ≈ 1 × 10⁻⁸ GeV cm⁻² s⁻¹ for a one‑year exposure, demonstrating that ANTARES is particularly competitive in the southern sky where IceCube’s coverage is limited.

The authors identify the principal factors limiting current sensitivity: the non‑uniform geometry of the PMT array, light absorption and scattering in seawater, and the residual background from atmospheric muons. They outline several avenues for improvement: (i) the adoption of deep‑learning‑based reconstruction algorithms to sharpen angular resolution and energy estimation; (ii) accumulation of additional years of data to increase statistical power; and (iii) the forthcoming KM3NeT detector, which will feature a much larger instrumented volume and a denser optical module layout, promising an order‑of‑magnitude gain in point‑source sensitivity.

In summary, while no astrophysical neutrino point source has been detected with ANTARES to date, the analysis provides robust upper limits and a well‑characterized sensitivity curve that constrain theoretical models of high‑energy neutrino production. These results serve both as a benchmark for future Mediterranean‑sea neutrino telescopes and as a complementary dataset to the IceCube observations, enriching the global effort to identify the cosmic accelerators responsible for the highest‑energy particles in the universe.