Seafloor Topography Enhances KM3NeT Sensitivity to ANITA-like Events

TL;DR

This work addresses the tension between a KM3NeT event at ultra-high energy and IceCube's non-observation by incorporating detailed seafloor topography into UHE neutrino sensitivity analyses. It introduces a model-independent effective parameterization with a cross-section $σ$, long-lived particle lifetime $τ$, and muon-branching ratio ${ m Br}(T o μ)$, plus a diffuse-flux normalization $Φ$, allowing both track-like and cascade-like signals via LLP production and decay. The key contribution is the concept of topographic enhancement: the azimuthally asymmetric environment around KM3NeT can boost the muon-effective area by up to a factor of $ ext{O}(3)$ in certain directions, enabling better agreement between KM3NeT and ANITA-IV while revealing remaining tension with IceCube. A joint analysis shows compatible regions in $(σ,τ)$ that lessen but do not fully remove the IceCube tension, and the authors provide code to extend this framework to other experiments, underscoring the importance of detector geography in next-generation UHE neutrino searches.

Abstract

In this article, we introduce the concept of \textit{topographic enhancement} in the context of ultra-high-energy neutrino detection by underwater neutrino telescopes. We demonstrate that the local topography around KM3NeT/ARCA can increase the detection efficiency in scenarios involving long-lived particles by up to a factor of $\sim 3$ due to the presence of an underwater mountain range in the direction of Malta. We consider a simplified model-independent approach that parametrizes the new physics able to generate both track-like and cascade-like signals in neutrino telescopes. When explaining the KM3-230213A event with a diffuse dark flux hypothesis, including its azimuthal direction--in addition to the zenith angle--provides additional constraints on the parameter space. In this effective model, the observations by KM3NeT and ANITA-IV can be simultaneously explained and the global tension with the lack of a corresponding detection in IceCube is reduced to 2.4 sigma. This work underscores the importance of incorporating topographic effects in the design and optimization of next-generation neutrino telescopes, as is done in the context of mountain-based detectors such as TAMBO. We present a numerical code which can be used to easily extend this topographical analysis to other experiments.

Figures (14)

• Figure 1: Measurements of high-energy and ultra-high-energy events compatible with a neutrino origin, illustrated through their corresponding differential flux. In orange, the four events from the four flight of ANITA ANITA:2020gmvANITA:2021xxh, consistent with a $\nu_\tau$ origin. In teal, the KM3-230213A event from KM3NeT KM3NeT:2025npi, consistent with a $\nu_\mu$ origin. In pink, purple and brown, IceCube measurements for high-energy starting events (HESE) IceCube:2020wum, Northern-Sky Tracks (NST) Abbasi:2021qfz and the Glashow resonance IceCube:2021rpz, respectively. Pink and purple shades represent the power law fits to their respective data, at 68% confidence level, and their extrapolation to higher energies. Finally, a dotted line represents the IceCube upper limit at $E\gtrsim 10\,\mathrm{PeV}$, at 90% confidence level IceCubeCollaborationSS:2025jbi. Under the Standard Model and a diffuse flux hypothesis, both KM3-230213A and ANITA-IV events are in tension with the non-observation of UHE events at IceCube.
• Figure 2: A BSM origin for the ANITA-IV and KM3-230213A events. An incoming flux of dark particles produces, with cross section $\boldsymbol{\sigma}$, secondary long-lived particles which decay after a laboratory lifetime $\boldsymbol{\tau}$. These particles can decay into muons and produce track-like topologies as in KM3-230213A (bottom), or into other SM particles able to produce electromagnetic cascades as in ANITA-IV (top). We will conservatively assume $\mathrm{Br}(\mathrm{T}\to\mathrm{casc})=1-\mathrm{Br}(\mathrm{T}\to\mu)$. While $\sigma$ and $\tau$ control the propagation, the branching ratios control the track-to-cascade ratio.
• Figure 3: Azimuthal distribution of the ARCA effective area to muons under different $(\sigma,\tau)$ parameters, overlaid on the topography surrounding the ARCA detector GEBCO_2024_Grid. The chosen elevation angle, $0.6^\circ$ above horizon, and the azimuthal angle shown in a grey band coincide with the reconstructed direction of KM3-230213A. In teal, ARCA is slightly more sensitive to $\nu_\mu$-like events from water-dominated trajectories, as explained in the main text. Analogously, in orange, ARCA is more sensitive to $\nu_\tau$-like events from rock-dominated trajectories. Here, $\mathrm{Br}(\mathrm{T}\to\mu)=1$. The incoming direction of KM3-230213A may favor a BSM interpretation with an intermediate long-lived particle.
• Figure 4: Angular dependence of the muon effective area $A_{\rm eff}^\mu$ (as defined in the main text) for the different points of the $(\sigma,\tau)$ parameter space as chosen in the rightmost plot. A common color bar for all plots shows $A_{\rm eff}^\mu$ around the horizontal plane, and the white point indicates the reconstructed direction of KM3-230213A within 68% confidence region. Long lifetimes and small cross sections favor up-going trajectories as shown in the right panels, while the SM-like scenarios shown in the left panels favor Earth-skimming trajectories. The angular distribution of ARCA events has the potential to constrain their BSM origin.
• Figure 5: Allowed $(\sigma,\tau)$ parameter space from KM3-230213A under the diffuse flux hypothesis, accounting for its angular direction and track-like topology, at 1 and 2 sigma confidence levels. On the background, the required flux normalisation to produce one event in the detector's observation time, with $\mathrm{Br}(\mathrm{T}\to\mu)=1$. Additional axes show the mean free path of N particles in Earth's crust, $\lambda_{\mathrm{crust}}$, and decay length, $c\tau$; and the $\nu_\mu$-like and $\nu_\tau$-like scenarios are shown as a black arrow and circle, respectively. KM3-230213A alone can already constrain to 1 sigma the BSM regions that can explain it.