The high entropy of the UHECR arrival direction distribution favors a light composition

arXiv:2608.17801 · astro-ph.HE · Submitted 2026-08-18 · Read on arXiv

Listen

Radio episode about this paper

Transcript

Introduction to the show: ident: Astrophysics Radio. Generated commentary on the latest astrophysics papers.

Vera: We're looking at the paper: "The high entropy of the UHECR arrival direction distribution favors a light composition".

Paper summary: Vera: The core thesis here is that heavy nuclei give a larger anisotropy signal due to their shorter propagation distances, even though they get deflected more by magnetic fields.

Jocelyn: But then they introduce this new measure of anisotropy called entropy of the arrival direction distribution which is much less sensitive to the configuration of the galactic magnetic field.

Subrahmanyan: That entropy measure is what has strong discriminating power between heavy and light composition models, which is a key finding for us in cosmology.

Vera: They then look at data from Auger above three point two times ten to nineteen electron volts and find that the correlation with large scale structure on those large angular scales is weak.

Jocelyn: And to mask that LSS signal, they suggest a very low source density, specifically s naught less than ten to negative four megaparsecs cubed per cubic megaparsec.

Subrahmanyan: They rule out a higher source density of ten to negative two Mpc cubed per cubic megaparsec at ninety-nine percent confidence level.

Vera: The high entropy they found is consistent with proton models for a low source density, and inconsistent with heavy nuclei models at ninety-six percent confidence level when the source density is ten to negative five megaparsecs cubed per cubic megaparsec or greater.

Conclusion: Vera: So, to wrap up, the paper "The high entropy of the UHECR arrival direction distribution favors a light composition" points toward a light composition for ultra-high-energy cosmic rays.

Jocelyn: This conclusion stems from analyzing how the arrival directions are spread out, using an entropy measure that bypasses some of the magnetic field uncertainties.

Subrahmanyan: What this means for us is that if we see this entropy signal, it strongly suggests a light composition and requires a very low source density to be consistent with the data.

Vera: They also found tension between their model predictions for the dipole amplitude at energies above three point two times ten to nineteen electron volts and what we actually observe, which is consistent with isotropy even for a low source density.

Jocelyn: So, while they see hints of LSS correlation in heavy nuclei models due to short distances, the overall entropy measurement favors protons under specific conditions.

Subrahmanyan: This work provides a way to use these arrival direction statistics to constrain both the composition and the source distribution of UHECRs across different energy regimes.

Department of Particle Physics and Astrophysics, Weizmann Institute of Science, Rehovot, Israel

astro-ph.HE

Submitted: 2026-08-18

Updated: 2026-11-17

Comments: Revised after receiving many useful responses. Submitted to ApJ

Code: https://github.com/Nimrod-S/crtools

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Importance score: 78/100

The gist: The high entropy of the UHECR arrival direction distribution favors a light composition We analyze the constraints on the composition of ultra-high-energy cosmic rays (UHECRs)…

Key concepts

Entropy of Arrival Distribution
This is a mathematical measure applied to UHECR arrival maps. A high entropy indicates a more uniform or isotropic distribution of arrivals across the sky. The paper uses this metric to distinguish between models based on whether cosmic rays are light (like protons) or heavy (like nuclei).
Propagation Energy Loss Model
The authors developed a new semi-analytical method to model how UHECR nuclei lose energy as they travel through space. This method is faster than detailed simulations and helps determine the effective distance traveled by the cosmic rays, which is crucial for understanding their observed energy.
Large-Scale Structure (LSS) Anisotropy
The LSS refers to the distribution of matter in the universe, like galaxy clusters. Cosmic ray arrival directions can show patterns related to this structure. The paper investigates how the anisotropy signal from LSS is affected by whether cosmic rays are protons or heavy nuclei, considering magnetic field deflections.
Source Density ($s_0$)
This parameter represents the density of cosmic ray sources in space per cubic megaparsec. The analysis shows that to explain the observed data, a very low source density is required (around $10^{-4} ext{ Mpc}^{-3}$). A higher source density leads to inconsistencies with the observed high entropy.

Terminology

Summary

The high entropy of the UHECR arrival direction distribution favors a light composition We analyze the constraints on the composition of ultra-high-energy cosmic rays (UHECRs), and on the density and distribution of their sources, that may be inferred from their arrival-direction distribution, using a novel semi-analytical description of the propagation energy loss of atomic nuclei (A > 4) with energy > 2 × 1019 eV, that allows generating UHECR arrival maps faster than by using detailed propagation simulations and yields insights to the impact of propagation energy loss We show that the anisotropy of the UHECR arrival direction distribution due to the large-scale structure (LSS) of matter distribution is larger for heavy nuclei composition compared to protons, despite their larger deflections by magnetic fields, due to their shorter propagation distance and weaker dependence of rigidity on observed energy Identifying the LSS anisotropy signal is hampered for heavy nuclei due to their large deflections by the uncertain Galactic magnetic field (GMF) We introduce a new measure of anisotropy, an ”entropy” of the arrival-direction distribution, that is largely independent of the GMF configuration and has strong discriminating power between heavy- and light-composition models Analyzing the public > 3. p. 02 × 1019 eV Auger data, we show that the correlation with the LSS on large angular scales is weak and requires a low source density, s0 ≤ 10−4Mpc−3, to allow masking the LSS signature by ”cosmic-variance” (s0 = 10−2Mpc−3 is ruled out at > 99% confidence level (CL)) The high entropy of the distribution is consistent with proton models and inconsistent with heavy nuclei models at > 96% CL for s0 ≥ 10−5Mpc−3

Modeling Propagation and Energy Loss

The paper presents a novel semianalytical description of the propagation energy loss of atomic nuclei (A > 4) with energy > 2 × 1019 eV, which allows generating UHECR arrival maps faster than by using detailed propagation simulations and yields insights to the characteristics of the impact of propagation energy loss For A >> 4, the mass loss is approximated as a continuous process described by an equation involving redshift, proper distance, and mean free paths The effective MFP is defined as λ−1 = ∑i niλ−1i, which simplifies the expression to dA/dz = −dx/dz 1/λ(A, γ, z) The solution for energy loss is given by ln As − ln Ao = λ(γ)−1 ˜d(z), where ˜d(z) is the ”effective propagation distance” This expression can also be written as ˜d = ln (Es/Eo)λ(Es/(Asmpc 2)), so instead of depending on our approximation for λ(γ), we can instead directly plug in the known λ(A, γ) values

Analyzing Magnetic Deflections

The study addresses magnetic deflections by focusing on observables that are not sensitive to the exact GMF model used The coherent component's effect is handled by calculating deflection lenses Lm,i(l, b) (where m indexes the GMF model and i indexes the energy bin) which will be used later when we simulate UHECR arrival maps The random component of the magnetic field's impact is estimated using formulas derived from Waxman & Miralda-Escudé 1996 The RMS of the deflection angle can be estimated to be θ ≈ 5.9◦ (fV 0.05 D 100 Mpc l 100 kpc)B nG (R3 EV−1), which is not negligible but is not big enough to completely erase anisotropy signals The EGMF deflections are concluded to be unimportant compared to the GMF deflections

LSS Correlation and Anisotropy Measures

The paper derives test statistics to compare the data against different models, including composition and source bias The correlation to large scale structure is quantified by Tp = ∑Xi N(Xi) ln Φp(Xi)/∑Xi N(Xi) The entropy of the arrival map is calculated as S = −∑xi N˜(xi) log N˜(xi), where ˜ is the normalized, smeared arrival map The effective entropy SE = ∑Xi S(Xi) (46) which is the sum of the entropies calculated separately for each of the large angular bins described in Table 3 These measures are used to differentiate between protons and nuclei, and they are robust with respect to uncertainties in the GMF structure

Results and Conclusions

The results for an Auger exposure show that the data are consistent with an isotropic source distribution, implying a low source density is required to mask the LSS correlation signal A source density of s0 = 10−2 Mpc−3 is ruled out with a p-value of < 0.1% even for the relatively isotropic proton model The high entropy of the available data implies a light composition with ∼ 95% confidence for a low, 10−4 Mpc−3, source density, and ∼ 99% for a high, 10−2 Mpc−3, source density The model predictions for the amplitude of the dipole at E > 32 EeV are in tension with the observed dipole, which is consistent with isotropy, even for a low source density Heavy nuclei models are inconsistent with the data at 99% CL, while proton models are inconsistent with the data at lower CL, 94% and 92% for source densities of 10−2 Mpc−3 and 10−4 Mpc−3 respectively

Comparison to Previous Work

The study addresses earlier work by noting that the test statistic (TS) defined is calculated for UHECR protons with proton flux attenuation, then applied to data that might actually be UHECR nuclei The paper concludes that a heavy composition should yield a less isotropic distribution, but the large magnetic deflections of nuclei lower their LSS correlation signal to be comparable to protons The high entropy of the available data implies a light composition with ∼ 95% confidence for a low, 10−4 Mpc−3, source density, and ∼ 99% for a high, 10−2 Mpc−3, source density The model predictions for the amplitude of the dipole at E > 32 EeV are in tension with the observed dipole, which is consistent with isotropy, even for a low source density

Summary and Conclusions

The paper has described a novel method for simulating UHECR arrival maps, including a semi-analytic method for propagating the spectrum of UHECR nuclei that is detailed in § 2 The expression we derived for the relation between the source distance, source energy and observed energy, given in Eq. (14), matches simulations well as seen in Fig. p. 4 3 and allows us to calculate important quantities like the rigidity distribution (Fig. p. 4 4) and the source distance distribution (Fig. p. 6 5) of nuclei UHECRs We neglected EGMF deflections and chose our observables to be as insensitive as possible to the exact configuration of the GMF All of our results confirm that the data are highly isotropic The high entropy of the available data implies a light composition with ∼ 95% confidence for a low, 10−4 Mpc−3, source density, and ∼ 99% for a high, 10−2 Mpc−3, source density The physical reason for this difference between compositions is that nuclei must ultimately come from much closer distances (as seen in Fig. p. 0 5) and therefore from many fewer sources The high entropy of the available data implies a light composition with ∼ 95% confidence for a low, 10−4 Mpc−3, source density, and ∼ 99% for a high, 10−2 Mpc−3, source density The physical reason for this difference between compositions is that nuclei must ultimately come from much closer distances (as seen in Fig. p. 0 p. 0 5) and therefore from many fewer sources The high entropy of the available data implies a light composition with ∼ 95% confidence for a low, 10−4 Mpc−3, source density, and ∼ 99% for a high, 10−2 Mpc−3, source density The physical reason for this difference between compositions is that nuclei must ultimately come from much closer distances (as seen in Fig. 5) and therefore from many fewer sources

How the Training Signal is Built

The paper details how test statistics are built to compare the data against different models, including composition and source bias For a given set of parameters p and energy bin we can calculate the mean arrival map Φp(x) by integrating Eq.

Green — the paper says it, and the page is printed. Red — no passage carries it, so treat that sentence as unverified. The same check runs on your own text: check a summary against its source, $5.

Improvements for AI systems

  1. Improved UHECR arrival map simulation capabilities by integrating a novel semi-analytical description of propagation energy loss for atomic nuclei (A > 4) that allows generating maps faster than by using detailed propagation simulations.

  2. Enhanced compositional discrimination between heavy and light nuclei models through the introduction of a new measure called entropy of the arrival-direction distribution, which is largely independent of the GMF configuration and has strong discriminating power between heavy- and light-composition models.

  3. Refined constraints on source density by demonstrating that for heavy nuclei models, both low source density and GMF deflections are required to account for the observed T value with s0 equal to 10 minus 4 Mpc minus 3, whereas in the absence of such deflections, s0 less than or equal to 10 minus 5 Mpc minus 3 would be required.

  4. More robust magnetic field analysis by focusing on observables that are not sensitive to exact GMF model uncertainties, specifically using very large angular scales or general properties like entropy and inter-energy correlation statistics that are not affected significantly by deflections.

  5. Better source localization and bias inference by developing a method where one can skew the distribution we get from the much more complete 2MRS by choosing a different b, trusting that the sources follow the LSS anyway, instead of relying on direct matter density maps.

Sources

Related papers