Sea, lake and ice: telescopes with complementary skies

A source’s background depends on the site’s latitude

Sky coordinates and local coordinates are different. As Earth rotates, a source can rise above and sink below the horizon of a mid-latitude site. At the South Pole its elevation is essentially fixed by its declination. The same source therefore meets different detector backgrounds at different sites.

Looking through Earth filters out many atmospheric muons, but atmospheric neutrinos remain in the sample. At high enough energy, Earth also absorbs part of the neutrino flux. Whether a view is favorable depends on energy and selection, not just on whether the source is geometrically visible.

Exposure is the accumulated probability of accepting a signal from a given direction. It includes usable time, detector configuration and response. A sky-coverage map should show those weights; shading a whole hemisphere as equally observed hides much of what decides a source search.

Sources: [1] · [2] · [3]

ANTARES proved the method; KM3NeT carries it on

ANTARES showed that neutrino observation in Mediterranean seawater could be sustained for years. It stopped taking data in February 2022. Its dataset is still available for analysis, so its scientific contribution continues after shutdown.

KM3NeT extends deep-sea observation with two arrays. ARCA, off Sicily, targets high-energy neutrino astronomy. The denser ORCA, off southern France, is designed to study how atmospheric neutrinos change flavor at lower energies. The two share module technology but differ in spacing and science goals.

The collaboration reported 62 detection units in ARCA on 23 September 2026 and 44 installed units in ORCA on 9 October. These dated installation counts are not the live exposure of any earlier physics paper, and they do not mean the full planned arrays are finished.

Sources: [4] · [3] · [5] · [6]

Lake Baikal adds an independent northern instrument

Baikal-GVD deploys modular clusters in the deep water of the lake. In winter the ice can serve as a platform for construction and maintenance, while the sensors watch the water below. IceCube is different: its sensors are frozen into the optical medium itself.

A JINR report dated 19 February 2026 describes an expedition to add clusters and cables. It reports ongoing expansion and planned work; it does not show that every planned component was already in a physics dataset. This article therefore does not attach an undated full-array size to published results.

Baikal-GVD’s diffuse-flux measurement uses a selected cascade sample and stated spectral assumptions. It tests the same population question with an independent response and background. It cannot be compared with a track sample simply by dividing event counts by nominal detector volumes.

Sources: [7] · [8] · [9]

Water and ice need different calibrations

All these optical telescopes infer interactions from light, but light travels differently in each medium. Absorption removes photons; scattering changes their paths and arrival times. What an analysis can use is the calibrated distribution of possible sensor signals.

Deep-water detector lines move, so positioning and timing must follow the changing geometry. Natural light in the water also reaches the triggers. IceCube’s sensors are fixed, but they need a detailed model of how the ice changes with depth and of local optical effects around each module.

These differences can make direction or energy information complementary, but no medium wins at everything. Compare a stated event class, energy range and reconstruction, with uncertainties. A good resolution quoted for selected tracks does not automatically apply to every cascade.

Sources: [2] · [10] · [3]

What a joint search needs

A joint search can multiply the likelihoods of independent samples under a shared source hypothesis. Each instrument keeps its own acceptance, background and response parameters. The gain comes from complementary information and added exposure; the events are not interchangeable.

First check whether the selected samples overlap or share modeled uncertainties. Then fix the source spectrum, time window and trial correction. Seeing nothing in one telescope means something only after calculating how many accepted events the source model predicts there.

The status statements above were checked on 11 October 2026. A construction count that keeps changing belongs next to its dated source. The lasting lesson is about method: another sea, lake or latitude gives an independent test of the same astronomical hypothesis.

Sources: [11] · [3] · [8]

Try it in the Lab

Primary sources & revision

1. Francis Halzen · Astroparticle Physics with High Energy Neutrinos: from AMANDA to IceCube (2006)
2. IceCube Collaboration · Instrumentation and Online Systems (2017; arXiv v3)
3. KM3NeT Collaboration · Letter of Intent for KM3NeT 2.0 (2016)
4. ANTARES Collaboration · Detector dismantling (2022)
5. KM3NeT · Welcome, ARCA62! (23 September 2026)
6. KM3NeT · ARCA infrastructure and ORCA44 (9 October 2026)
7. JINR · Winter Baikal expedition started (19 February 2026)
8. Baikal-GVD Collaboration · Diffuse neutrino flux measurements (2022/2023)
9. Baikal-GVD Collaboration · Observatory and location
10. KM3NeT Collaboration · The Calibration Units of KM3NeT (2021)
11. Particle Data Group · Statistics (2025)

First published 2026-10-11; last revised 2026-10-13. Original explanatory text and diagrams by Sandbox Physics. Illustrations are schematic; no experimental event records are reproduced here. This is an independent educational publication, not an official Nobel or experiment collaboration publication.