Identification of a Large-Scale Diffuse Gamma-Ray Structure in the Southern Galactic Hemisphere

The paper identifies and characterizes a large-scale diffuse gamma-ray structure in the Southern Galactic Hemisphere using 17 years of Fermi-LAT data (August 4, 2008, to October 20, 2025), covering the energy range 200 MeV to 500 GeV with the CLEAN event class. An energy-dependent likelihood analysis was performed, including alternative Galactic diffuse-emission models (an ensemble of 64 GALPROP templates), isotropic emission, the Fermi bubbles, and resolved 4FGL sources. The analysis reveals an extended excess that persists across all tested background models and spans tens of degrees.

The excess broadly follows the X-ray-defined southern eROSITA Bubble (eB) region, while also overlapping the projected southern extension of Loop I. Template fits favor a filled eB-like morphology over the adopted Wolleben Loop I shell geometry, with a benchmark Akaike information criterion difference of AICeB − AICLoop I = −78.47, favoring the eB-like morphology. The template test statistics are TS = 881.46 for the Wolleben Loop I template and TS = 959.93 for the eB template, both relative to the same baseline model. Over the full 64-model ensemble, the mean template TS values are ⟨TSLoop I ⟩ = 763.0 and ⟨TSeB ⟩ = 932.2, indicating the fit strongly favors an additional large-scale component in the southern high-latitude sky. This makes the structure a plausible gamma-ray counterpart of the southern eB, although Loop-I-related or other localized foreground emission cannot be excluded.

Under the eB template, the southern component is fainter and softer than the northern large-scale component, with an integrated luminosity lower by a factor of about seven (the total luminosity ratio between the northern and southern structures is approximately 7.2), broadly consistent with the eROSITA-bubble asymmetry (estimated at approximately 3–6 times the southern intensity based on the luminosity profile integrated over the ±50◦ latitude range). The southern structure exhibits a softer spectrum compared to its northern counterpart, with the spectral index notably larger (softer) across the observed energy range.

For the southern component in the eB-template situation, the integrated photon flux over the unmasked portion of the spatial template used in the likelihood analysis is 1.0 × 10−7 ph cm−2 s−1. The inferred gamma-ray luminosity is Lγ ≈ 2.4 × 1033 erg s−1 assuming a local origin at d = 400 pc, and Lγ ≈ 9.7 × 1035 erg s−1 for a Galactic-scale distance of d = 8 kpc. Under the same eB model framework, the northern component exhibits a significantly higher luminosity, with Lγ ≈ 1.8 × 1034 erg s−1 at 0.4 kpc and Lγ ≈ 7.1 × 1036 erg s−1 at 8 kpc. For comparison, adopting the Loop I model leads to systematically higher gamma-ray luminosities: the northern component yields Lγ ≈ 2.3 × 1034 erg s−1 (0.4 kpc) and 9.2 × 1036 erg s−1 (8 kpc), while the southern component gives 4.5 × 1033 erg s−1 and 1.8 × 1036 erg s−1, respectively.

Spectral fitting was performed using a log-parabola model, and numerical modeling with the naima package considered both leptonic (inverse-Compton with a broken power-law electron distribution, with low-energy index fixed to γ1 = 1.5) and hadronic (pp with power-law or exponentially cutoff power-law proton distributions) origins, under two distance scenarios: local (d = 400 pc) and Galactic Center (d = 8 kpc). For the adopted BPL electron model, the fitted high-energy indices are γ2 = 3.53 and 3.55 for the local and Galactic Center Loop I cases, respectively, and 4.75 and 4.45 for the corresponding eB cases. The eB-template SED therefore requires a softer high-energy electron spectrum.

If interpreted as Galactic-scale outflow emission, its faint, soft spectrum may indicate aged particles and/or distributed reacceleration in the outer bubble. A hadronic interpretation is energetically demanding: for the Galactic-scale eB interpretation with a halo gas density of nH ≃ 10−3 cm−3, the southern eB-like component requires a cosmic-ray proton energy of Wp ∼ 1055 (10−3 cm−3 /nH)(d/8 kpc)2 erg, and assuming the same effective target density in both hemispheres with the approximately seven times more luminous northern counterpart, the total proton energy for the full bipolar system approaches Wp ∼ 1056 (10−3 cm−3 /nH)(d/8 kpc)2 erg. Assuming a characteristic cosmic-ray acceleration efficiency of ηCR ∼ 0.1, the required mechanical energy approaches Emech ∼ 1057 (10−3 cm−3 /nH)(0.1/ηCR)(d/8 kpc)2 erg, substantially higher than the ∼ 1056 erg energy scale estimated for the eROSITA bubbles. In contrast, a leptonic inverse-Compton scenario is more economical but requires rapid transport and/or local reacceleration of high-energy electrons. Because the eB hadronic ECPL fit prefers a relatively low proton cutoff, Ec,p ∼ 0.1 TeV, the associated TeV neutrino signal would be strongly suppressed.

The paper concludes that the southern structure represents a plausible gamma-ray counterpart of the southern eB, helping address the long-standing north-south asymmetry in the high-latitude gamma-ray sky. While the spatial alignment and large-scale morphology are naturally interpreted within a Galactic-scale outflow scenario, the current data do not yet allow for a decisive rejection of localized interpretations.

Improvements to AI Systems:

1. Multi-Template Morphological Model Selection: Implement an AI framework that automatically fits and compares multiple spatial templates (e.g., filled bubble vs. shell) using information criteria (AIC) and test statistics (TS). The system can learn to distinguish between competing morphological hypotheses for diffuse astrophysical structures, reducing human bias in model selection.

2. Energy-Dependent Background Ensemble Analysis: Develop an AI pipeline that systematically varies Galactic diffuse emission models (e.g., an ensemble of 64 GALPROP templates) and performs joint likelihood fitting across energy bins. This improves robustness against systematic uncertainties in background modeling, enabling the detection of faint, extended sources that persist across all model variations.

3. Automated Asymmetry Quantification: Create an AI tool that quantifies north-south or east-west asymmetries in large-scale diffuse emission (e.g., luminosity ratios, spectral index differences) directly from photon count maps. This can flag physically significant asymmetries in real-time, aiding in the identification of Galactic outflows or bubbles.

4. Spectral SED Fitting with Physical Models: Integrate a module that automatically fits spectral energy distributions (SEDs) using both leptonic (inverse-Compton with broken power-law electron distributions) and hadronic (pp with power-law or cutoff power-law proton distributions) models. The AI can output best-fit parameters (e.g., electron/proton indices, cutoff energies) and compute derived quantities like total cosmic-ray energy and required mechanical energy, given user-specified distance and gas density.

5. Distance-Scenario Inference: Build an AI system that, given a detected diffuse structure, automatically evaluates multiple distance assumptions (local vs. Galactic-scale) and computes corresponding luminosities, energy requirements, and feasibility of physical interpretations (e.g., aged particles vs. reacceleration). This helps constrain the origin of the emission without manual recalculation.

6. Neutrino Predictability Screening: Add a feature that, based on fitted hadronic models (e.g., proton cutoff energy), automatically predicts the associated TeV neutrino flux and flags whether it would be detectable by current or future neutrino observatories. This optimizes follow-up observation strategies.

7. Cross-Wavelength Spatial Correlation: Implement a cross-correlation algorithm that aligns gamma-ray excess maps with X-ray (eROSITA), radio, or other wavelength data. The AI can automatically compute overlap fractions and statistical significance, suggesting multi-wavelength counterparts for newly discovered structures.

8. Uncertainty-Aware Component Separation: Develop a Bayesian component-separation method that simultaneously models point sources, isotropic emission, and multiple diffuse components, providing posterior distributions for each component’s flux and morphology. This yields more reliable error bars and helps disentangle overlapping structures (e.g., eB vs. Loop I).

What the Improved AI System Can Do:

• Automatically detect and characterize faint, extended gamma-ray structures with robust statistical significance, even in the presence of strong background uncertainty.

• Distinguish between competing morphological models (e.g., filled vs. shell) and provide quantitative evidence for the best fit.

• Rapidly compute physical parameters (luminosity, cosmic-ray energy, spectral indices) under multiple distance and emission-model scenarios, without manual intervention.

• Flag astrophysically interesting asymmetries and cross-wavelength counterparts, enabling faster follow-up studies.

• Predict observational signatures (e.g., neutrino flux) to prioritize multi-messenger observations.

• Reduce human bias in model selection and background treatment, leading to more reproducible and objective analyses of diffuse emission in the gamma-ray sky.

We identify and characterize a large-scale diffuse gamma-ray structure in the Southern Galactic Hemisphere using 17 yr of Fermi-LAT data. An energy-dependent likelihood analysis, including alternative Galactic diffuse-emission models, isotropic emission, the Fermi bubbles, and resolved 4FGL sources, reveals an extended excess that persists across the tested background models and spans tens of degrees. The excess broadly follows the X-ray-defined southern eROSITA Bubble (eB) region, while also overlapping the projected southern extension of Loop I. Template fits favor a filled eB-like morphology over the adopted Wolleben Loop I shell geometry, making the structure a plausible gamma-ray counterpart of the southern eB, although Loop-I-related or other localized foreground emission cannot be excluded. Under the eB template, the southern component is fainter and softer than the northern large-scale component, with an integrated luminosity lower by a factor of about seven, broadly consistent with the eROSITA-bubble asymmetry. If interpreted as Galactic-scale outflow emission, its faint, soft spectrum may indicate aged particles and/or distributed reacceleration in the outer bubble. A hadronic interpretation is energetically demanding, whereas a leptonic inverse-Compton scenario is more economical but requires rapid transport and/or local reacceleration of high-energy electrons.