A Testable Candidate Taxonomy and Ranked Nearby Source List
Kevin L. Brown
August 2026
Ultra-high-energy cosmic rays (UHECRs) reach energies above (10^{18}) electron volts and, in rare events, exceed (10^{20}) eV. Despite decades of observation, their astrophysical origins remain unresolved. Their charged nature prevents simple point-back source identification because Galactic and extragalactic magnetic fields deflect their trajectories, while energy losses and nuclear breakup further restrict which sources can contribute at the highest energies.
This paper reframes the UHECR source problem as a finite classification-and-ranking problem.
Rather than stopping at broad environments such as jets, shocks, radio lobes, starburst winds, compact-object transients, or accretion shocks, the paper defines a closed six-type taxonomy of astrophysical source candidates and then identifies specific nearby systems that can be tested against present and future observations. The goal is to move from asking what kinds of environments could theoretically accelerate UHECRs to asking which actual systems remain viable after confinement, acceleration, energy losses, escape, source history, distance, composition, and magnetic propagation are considered together.
The six candidate types are:
Type A — Radio-loud AGN and radio galaxies
Type B — Starburst and superwind galaxies
Type C — Composite AGN-starburst systems
Type D — Compact-object and explosive transients
Type E — Galaxy-cluster and accretion-shock systems
Type F — Local irregular or satellite anomalies
Within the persistent nearby-source sample, Centaurus A and Fornax A are placed in the highest-priority tier as Type A radio-galaxy and lobe-reservoir candidates. NGC 4945 is classified as a Type C composite AGN-starburst discriminator, while M83, NGC 253, and M82 form the Type B starburst/superwind comparison set. The Large Magellanic Cloud is retained as a Type F anomaly candidate because of its proximity to reported directional clustering among very-high-energy events. M87/Virgo A is used as a Type A control: it possesses apparently favorable accelerator properties but lacks a comparably strong UHECR excess, making it valuable for testing whether the ranking framework genuinely discriminates between powerful sources.
Types D and E remain part of the complete taxonomy even though they are not represented by a single persistent object in the initial ranking. Gamma-ray bursts, tidal-disruption events, magnetars, mergers, galaxy clusters, and accretion-shock reservoirs instead require transient, population-level, or extended-source analyses.
The paper does not claim that any listed object has been identified as a UHECR source, nor does it assign calibrated numerical probabilities. The proposed ranking is a preregistration scaffold for future testing. A fitted model must ultimately outperform isotropy and simpler luminosity-, distance-, or source-class baselines on held-out UHECR data while remaining compatible with spectrum, composition, magnetic deflection, source energetics, and multimessenger constraints.
The central contribution is therefore a bounded and falsifiable search architecture: a finite source taxonomy, a named candidate hierarchy, explicit controls, and clear conditions under which the hypothesis should be rejected.