⚛️ high-energy experiments

Search for an all-sky and a Galactic Ridge diffuse neutrino emission with the first 2 years of KM3NeT/ARCA data

Using 640 days of data from the initial KM3NeT/ARCA configurations, this study reports no statistically significant detection of diffuse astrophysical neutrinos from the all-sky or Galactic Ridge regions, instead providing flux measurements and upper limits that establish a methodological foundation for future observations.

Original authors: The KM3NeT Collaboration

Published 2026-08-07

📖 4 min read🧠 Deep dive

CC BY 4.0

Original authors: The KM3NeT Collaboration

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine the universe is a giant, chaotic cosmic party where invisible messengers are constantly zooming through space. Most of these messengers are light or charged particles, but they get easily distracted by magnetic fields or blocked by dust, making it hard to trace where they came from. Then there are the "ghosts" of the particle world: neutrinos. These tiny, neutral particles barely interact with anything, allowing them to zip straight through stars, planets, and entire galaxies without getting lost. Because they travel in straight lines, they act like perfect postcards, pointing directly back to the most violent and energetic events in the cosmos, like exploding stars or black holes gobbling up matter. Scientists have been hunting for these ghosts for decades, hoping to figure out what powers the universe's most extreme engines. The big question is: are these ghosts coming from everywhere in the sky equally, or are they spilling out of our own cosmic neighborhood, the Milky Way galaxy?

This paper is the latest report from a team of ghost hunters using a massive underwater telescope called KM3NeT/ARCA, located deep in the Mediterranean Sea. Think of the telescope as a giant, three-dimensional net made of thousands of light sensors, waiting to catch the faint blue flash that happens when a neutrino occasionally bumps into a molecule of seawater. The team looked at data collected over 640 days, using the detector while it was still being built (with anywhere from 6 to 21 "strings" of sensors active). They had two main missions: first, to scan the entire sky to see if they could find a general background glow of these cosmic ghosts; and second, to look specifically at the "Galactic Ridge," a busy, crowded strip of our galaxy where stars are born and die, to see if that specific neighborhood was leaking neutrinos.

The results are a bit like a detective story where the clues are still too faint to solve the case, but the detective is getting better at looking. When the team scanned the whole sky, they didn't find a clear, undeniable signal of new cosmic ghosts. Instead, they found a number that could be a signal, but it's so close to zero that it might just be random noise. They calculated that the strength of this potential signal is about $3.0 \times 10^{-18}$ (with a big range of uncertainty), and the "color" or energy pattern of these ghosts has a spectral index of $3.00$. However, because the numbers wiggle so much, they can't say for sure that they've found a new source yet. They did set strict "upper limits," which are like saying, "If there is a signal, it is definitely weaker than this line." These limits are very competitive and match what other telescopes have seen, but they don't prove the signal exists yet.

When they turned their attention to the Galactic Ridge, the hunt was even quieter. They used a clever trick to separate the signal from the background: they compared the number of ghosts coming from the Galactic Ridge against ghosts coming from empty patches of sky that the telescope saw at different times. It's like checking if a specific street corner has more people than a random corner down the block to see if there's a concert happening. The result? The number of ghosts from the Galactic Ridge was exactly what you'd expect if there were no special source there at all. The data was consistent with the background noise, meaning they couldn't confirm the presence of a neutrino stream from our galaxy's center.

In short, this paper doesn't announce a discovery of a new cosmic power source. Instead, it shows that the KM3NeT telescope is working perfectly and is ready for the big reveal. The team has proven they can filter out the noise and look for these faint signals with high precision. While they didn't find a "smoking gun" in this specific dataset, they have laid the groundwork for future searches. As the telescope grows bigger and collects more data over more years, it will eventually have the sensitivity to catch those elusive neutrinos and finally tell us exactly where the universe's most energetic particles are coming from. For now, the ghosts remain shy, but the hunters are getting closer.

Technical Summary: Search for All-Sky and Galactic Ridge Diffuse Neutrino Emission with KM3NeT/ARCA

Problem and Objectives
The primary objective of this work is to search for a diffuse astrophysical neutrino flux using the first two years of data from the KM3NeT/ARCA (Astroparticle Research with Cosmics in the Abyss) detector. The analysis targets two distinct regions: the full sky (all-sky) and the Galactic Ridge (defined by Galactic coordinates $|l| < 30^\circ$ and $|b| < 2^\circ$ ). While previous experiments like IceCube and ANTARES have established the existence of a diffuse high-energy neutrino flux and provided constraints on Galactic contributions, KM3NeT offers unique advantages due to its location in the Northern Hemisphere, allowing for sub-degree angular precision observation of the Galactic center. The study aims to independently assess potential excesses in these regions, determine best-fit values for signal flux parameters under a single power-law assumption, and evaluate the relative contribution of Galactic emission to the cosmic diffuse flux.

Methodology
The analysis utilizes data collected from four detector configurations (ARCA6, ARCA8, ARCA19, and ARCA21), corresponding to a total livetime of 640 days. The detector consists of vertical strings of Digital Optical Modules (DOMs) anchored to the seabed, optimized for the TeV–PeV energy range.

Event Reconstruction and Selection: The analysis focuses on track-like events, primarily resulting from charged-current $\nu_\mu$ interactions. A multi-stage reconstruction algorithm determines the arrival direction and energy of muons. A preliminary selection suppresses random noise and selects up-going events (zenith angle $> 84^\circ$ ) to reject atmospheric muons.
Background Rejection: To further purify the sample, a Boosted Decision Tree (BDT) classifier is employed. Two separate BDT models were optimized for different detector configurations (ARCA6/8 and ARCA19/21) using the ROOT TMVA package. The training and optimization were performed exclusively on Monte Carlo (MC) simulations to maintain a blind analysis strategy, with 10% of the data reserved as a control sample.
Statistical Framework:
- All-Sky Analysis: The background is estimated using MC simulations of atmospheric muons and neutrinos. A Bayesian approach is used to fit the flux normalization ( $\phi_0$ ) and spectral index ( $\gamma$ ), incorporating Gaussian priors for systematic uncertainties (PMT efficiency, light absorption length, and atmospheric flux normalization). A frequentist analysis using a Poissonian $\chi^2$ is also presented for cross-validation.
- Galactic Ridge Analysis: A model-independent, data-driven approach is used. The signal region (ON region) is compared against control regions (OFF regions) obtained by time-shifting the ON region in right ascension. This method leverages the Earth's rotation to ensure equivalent instrumental conditions for background estimation without relying on MC background normalizations.
Systematic Uncertainties: Key systematic uncertainties considered include PMT efficiency ( $\pm 10\%$ ), light absorption length in water ( $\pm 10\%$ ), and atmospheric flux normalization ( $\pm 40\%$ for the all-sky analysis).

Key Results

All-Sky Diffuse Flux:
- The analysis yields a best-fit single-flavour astrophysical neutrino flux normalization of $\phi_{0}^{1f} = 3.0^{+2.1}_{-2.0} \times 10^{-18} \text{ GeV}^{-1} \text{ cm}^{-2} \text{ s}^{-1} \text{ sr}^{-1}$ and a spectral index of $\gamma = 3.00^{+0.30}_{-0.35}$ at the 68% credible level.
- The atmospheric background normalization was found to be consistent with $1.53 \pm 0.24$ , indicating that the MC simulation underestimates the atmospheric neutrino contribution for the larger detector configurations (ARCA19 and ARCA21).
- No statistically significant excess of cosmic neutrino events was observed. Consequently, 90% confidence level (C.L.) upper limits were derived. These limits are consistent with and competitive to previous measurements from IceCube and ANTARES.
Galactic Ridge:
- The number of events observed in the ON region is consistent with the background estimates derived from the OFF regions.
- The fit does not yield constrained parameters for the flux or spectral index due to the lack of a significant signal.
- Upper limits for the Galactic Ridge diffuse flux were derived and compared with results from ANTARES and IceCube. The results are compatible with previous constraints but do not confirm the signal excess reported by ANTARES.

Significance and Claims
The paper claims that while the current dataset (640 days) does not yield statistically significant detections for either the all-sky or Galactic Ridge analyses, the work establishes a critical foundation for the KM3NeT collaboration. The author states that the developed methods, including the BDT-based selection and the data-driven background estimation for the Galactic Ridge, provide a "solid basis for future measurements."

The significance of the results lies in demonstrating that KM3NeT/ARCA can yield competitive diffuse flux measurements even with partial detector configurations (6 to 21 active Detection Units). The author emphasizes that the expected increase in exposure, driven by longer livetimes and further detector expansion, will significantly enhance sensitivity to both all-sky and spatially localized emissions, particularly for sources in the southern sky and along the Galactic plane. The work serves as a validation of the analysis pipeline and detector performance prior to the completion of the full cubic-kilometer array.

More like this