Galactic Center GeV Excess (GCE)

• Galactic Center GeV Excess (GCE) is a distinct ~2 GeV gamma-ray signal with spherical symmetry centered on Sgr A* and correlated with the stellar bulge.
• Spectral and spatial analyses using log-parabola fits and gNFW profiles reveal a sharp rise below 2 GeV, a plateau at 1–3 GeV, and a rapid cutoff above 10 GeV.
• Advanced template fitting and photon-count statistics suggest the GCE may originate from unresolved millisecond pulsars or cosmic-ray processes rather than a dominant dark matter component.

The Galactic Center GeV Excess (GCE) denotes a statistically significant surplus of γ-ray emission peaking at ∼1–3 GeV discovered with the Fermi Large Area Telescope (LAT) in the region surrounding the Galactic Center (GC). Its spectrum, morphology, and overall luminosity have generated extensive debate regarding its physical origin, with candidate explanations ranging from annihilation of weakly interacting massive particle (WIMP) dark matter (DM), to unresolved populations of millisecond pulsars (MSPs), and non-steady-state injection of cosmic-ray electrons or protons. Over the past decade, increasingly sophisticated analyses—spanning spectral, morphological, and photon-count statistical domains—have refined the empirical properties of the GCE and continually reshaped the landscape of preferred interpretations.

1. Spectral Properties and Observational Robustness

The GCE is characterized by a sharply peaked spectral energy distribution (SED) at $E_\gamma \sim 2\,\mathrm{GeV}$, with a rapid rise below this energy, a plateau at $1$–$3\,\mathrm{GeV}$, and a steep cutoff above $10\,\mathrm{GeV}$ in most analyses. Empirically, it is well described by either a log-parabola,

$$\frac{d\Phi}{dE} = N_0 \left(\frac{E}{E_0}\right)^{-\alpha - \beta \ln(E/E_0)},$$

with best-fit parameters $$N_0 \sim 6.5 \times 10^{-8}\,\mathrm{GeV}^{-1}\,\mathrm{cm}^{-2}\,\mathrm{s}^{-1}\,\mathrm{sr}^{-1}$$, $\alpha \sim 2.0$, and $\beta \sim 0.27$ ($E_0 = 1$ GeV) (Mauro, 2021), or by the $b\bar{b}$ prompt spectrum for $m_X \sim 50$ GeV WIMP annihilation with

$$m_X = (49^{+6}_{-7})\,\mathrm{GeV}, \qquad \langle \sigma v \rangle = (1.4^{+0.3}_{-0.2})\times 10^{-26}\,\mathrm{cm}^3/\mathrm{s}.$$

Systematic uncertainties—the dominant source of error—stem primarily from interstellar emission (IEM) modeling, influencing normalization and spatial profiles at the $\sim 30$–$60\%$ level (Mauro, 2021). Nonetheless, the GCE remains robustly detected across all plausible IEM realizations, with test-statistics $TS \gtrsim 600$ in a variety of analyses (Zhou et al., 2014).

High-energy tails extending to $E \sim 50$ GeV have been observed, with mild evidence for spectral hardening or changing morphology at these energies, further constraining proposed emission mechanisms (Linden et al., 2016).

2. Spatial Morphology: Sphericity, Profile, and Bulge-Tracing Models

Morphologically, early studies found the GCE distributed with approximate spherical symmetry about Sgr A*, showing less than $\sim 10$–$20\%$ deviation from isotropy in all angular bins. The radial surface brightness follows a power-law consistent with a generalized Navarro–Frenk–White (gNFW) density profile: $$\rho(r) \propto r^{-\gamma}, \qquad \gamma = 1.20 \pm 0.05,$$ implying $J$-factors $J(1^\circ$–$$10^\circ) \approx (1.5 \pm 0.3)\times 10^{23}\,\mathrm{GeV}^2\,\mathrm{cm}^{-5}$$ (Hooper, 2022). The fitted centroid coincides with $(l, b) = (0,0)$ at high precision.

Crucially, extended template analyses that allow for more complex stellar morphologies—such as the boxy/X-shaped bulge and nuclear bulge—demonstrate that the addition of these templates to Poisson-likelihood fits is strongly preferred over spherical DM templates, with improvements in log-likelihood corresponding to statistical significances $\gtrsim 16\sigma$ (Macias et al., 2016, Manconi et al., 5 Nov 2025). Once the bulge templates are included, spherically symmetric DM components are required only at $\lesssim 2 \sigma$ in most studies (Macias et al., 2016, Manconi et al., 5 Nov 2025).

Adaptive template fitting with skyFACT, in combination with 1pPDF pixel-count statistics, confirms that the GCE at both low and high energies closely follows the stellar bulge distribution rather than a spherical or NFW-like annihilation profile (Manconi et al., 5 Nov 2025, Manconi et al., 7 Feb 2024).

3. Photon-Count Statistics, Point Source Constraints, and Small-Scale Structure

The origin of the GCE—diffuse versus unresolved point sources—has been rigorously investigated using non-Poissonian template fitting (NPTF), wavelet studies, and 1pPDF analysis. Initial studies reported strong preference to decompose the entire excess into a population of unresolved MSP-like point sources based on the pixel-count distribution's excess variance (Bartels et al., 2015). However, injection tests showed these findings to be artifacts of insufficiently flexible background modeling (especially north/south asymmetry): injected smooth DM signals were misattributed as point sources with 100% efficiency (Leane et al., 2020, Hooper, 2022). Masking of all 4FGL sources and the use of up-to-date diffuse templates generally eliminates the statistical preference for subthreshold point sources contributing the GCE in the key analysis regions.

Quantitatively, upper limits now restrict the fraction of the GCE from sources brighter than $F_\mathrm{min} = 1 \times 10^{-10}$ ph/cm$^2$/s ($E > 1$ GeV) to $f_{PS} < 30\%$ at 95% CL (Hooper, 2022). At fainter fluxes, 1pPDF analyses resolve a population of faint point sources down to $S \sim 10^{-12}$ ph cm$^{-2}$ s$^{-1}$, with an excess (over Fermi 4FGL catalog sources) at these levels consistent with the numbers expected from a large population of dim bulge MSPs, but not capable of accounting for the entire GCE in most templates without invoking an unrealistically large and faint population (Manconi et al., 7 Feb 2024, Manconi et al., 5 Nov 2025). The source-count function in the regime $10^{-12} \leq S \leq 10^{-11}$ ph cm$^{-2}$ s$^{-1}$ follows $dN/dS \propto S^{-2.0\pm0.2}$ (Manconi et al., 7 Feb 2024).

Wavelet methods detect photon-clustering patterns consistent with a bulge MSP population at high statistical significance ($\sim10\sigma$), but this result is sensitive to unresolved systematics, and cannot by itself distinguish between truly discrete and extremely low-luminosity or diffuse emission (Bartels et al., 2015).

4. Astrophysical and Particle Physics Interpretations

4.1. Dark Matter Anniliation Models

The GCE spectrum, intensity, and morphological properties are in notable agreement with the expectations for thermal WIMPs annihilating via the $b\bar{b}$ channel for $m_X \sim 50$ GeV and $\langle \sigma v \rangle \sim 1.4 \times 10^{-26}$ cm$^3$/s. The annihilation flux is given by: $$\frac{d\Phi}{dE}(E, \Delta\Omega) = \frac{1}{4\pi} \frac{\langle \sigma v \rangle}{2 m_X^2}\frac{dN_\gamma}{dE} \times J(\Delta\Omega),$$ with $J$ the standard line-of-sight integral of $\rho^2$ (Hooper, 2022).

Nevertheless, once the bulge templates are included, the space for a dominant DM component is severely constrained, and the best-fit DM $J$-factor is no longer consistent with the inferred spatial morphology (Manconi et al., 5 Nov 2025). Extracted 95% CL cross-section upper limits for $b\bar{b}$ annihilation are at or below the canonical thermal value for $m_\chi \lesssim 300$ GeV (assuming cuspy NFW profiles), with thermal-relic WIMPs excluded as a dominant source for most plausible DM density profiles (Manconi et al., 5 Nov 2025).

4.2. Millisecond Pulsar and Other Astrophysical Models

Comprehensive template analyses converge on a preference for the GCE tracking the Galactic bulge's stellar mass distribution, compatible with a centrally concentrated, old stellar population such as MSPs (Manconi et al., 5 Nov 2025, Macias et al., 2016, Manconi et al., 7 Feb 2024). The combined MSP prompt γ-ray emission spectrum and the inverse-Compton emission from MSP-injected $e^\pm$ pairs can naturally reproduce the GCE's entire SED, including the high-energy tail above $10$ GeV (Linden et al., 2016, Manconi et al., 7 Feb 2024, Macias et al., 2016). The required bulge MSP population is estimated at $N_\mathrm{MSP} \gtrsim 10^{4-5}$, with luminosity functions significantly dimmer than those measured in globular clusters or the local disk (Hooper, 2022). Gravitational wave searches for the expected bulge MSP ensemble with next-generation detectors may test this scenario in the near future (Lei et al., 19 Nov 2025).

Alternative models assign the GCE to leptonic or hadronic cosmic-ray outbursts in the central $\sim1$ kpc within the past $10^5$–$10^7$ yr, which can replicate both spectrum and angular extension (Petrovic et al., 2014, Cholis et al., 2015). Bipolar morphologies have been claimed in some studies, suggestive of non-spherical, time-dependent injection events (Yang et al., 2016), though these are not universally found.

5. Methodological Advances and Systematic Uncertainties

The primary systematic limiting further progress is the modeling of spatial and spectral components of the Galactic diffuse γ-ray background. IEM mismodeling can induce uncertainties of order $10$–$30\%$ in the recovered GCE's normalization and inner profile slope $\gamma$ (Mauro, 2021, Mauro, 2020). Robust conclusions require template flexibility (e.g., skyFACT nuisance parametrization of gas/ICS templates (Manconi et al., 5 Nov 2025)), pixel-count statistics (1pPDF) for point source flux thresholds well below catalog limits (Manconi et al., 7 Feb 2024), and explicit simulation-based validation of extraction pipelines.

Compact region-of-interest selection ($|b|>2^\circ$, $r\sim 1^\circ$–$10^\circ$), analysis in annuli, quadrant symmetry tests, and quadrant-by-quadrant profile-likelihood scans are employed to minimize the influence of local background mismodeling. No energy-dependent shape evolution beyond $<10\%$ in the best-fit $\gamma$ parameter is found between $0.6$ and $30$ GeV (Mauro, 2021). The centroid of the GCE remains coincident with Sgr A* at sub-degree accuracy.

6. Current Status and Outlook

The current consensus from high-dimensional template fitting, photon statistics, and spectral-morphological analysis is that the GCE is a robust, spherically symmetric (to $\sim 10\%$) γ-ray excess centered at the GC, peaking at $E\sim2$ GeV, spatially correlated with the bulge and nuclear stellar populations. No statistically significant evidence remains for sub-degree unresolved clustering beyond established 4FGL point sources, nor for a dominant smooth, spherically symmetric (NFW-like) dark matter template when the stellar bulge is properly modeled (Manconi et al., 5 Nov 2025, Manconi et al., 7 Feb 2024, Macias et al., 2016, Hooper, 2022). The GCE high-energy tail is fully consistent with a combination of prompt and IC emission from a bulge MSP population, with the faint-point-source population detected by 1pPDF falling within theoretical expectations.

Direct detection, high-resolution radio and γ-ray surveys, multi-wavelength studies of GC electromagnetic and gravitational-wave emission, and improved modeling of Galactic diffuse backgrounds will further clarify the partition of the GCE into dark matter versus stellar components. The interpretation of the GCE as a signal of WIMP annihilation is strongly constrained—if not excluded—for all plausible Galactic halo profiles, whereas the bulge MSP scenario remains viable and predictive (Manconi et al., 5 Nov 2025).

---

References

• "The Status of the Galactic Center Gamma-Ray Excess" (Hooper, 2022)
• "The characteristics of the Galactic center excess measured with 11 years of Fermi-LAT data" (Mauro, 2021)
• "Stellar-like Galactic center excess challenges particle dark matter" (Manconi et al., 5 Nov 2025)
• "The Galactic center excess at the highest energies: morphology and photon-count statistics" (Manconi et al., 7 Feb 2024)
• "Galactic Bulge Preferred Over Dark Matter for the Galactic Center Gamma-Ray Excess" (Macias et al., 2016)
• "On the GeV excess in the diffuse γ-ray emission towards the Galactic Center" (Yang et al., 2016)
• "Strong support for the millisecond pulsar origin of the Galactic center GeV excess" (Bartels et al., 2015)
• "The Enigmatic Galactic Center Excess: Spurious Point Sources and Signal Mismodeling" (Leane et al., 2020)
• "Interpreting the galactic center gamma-ray excess in the NMSSM" (Cao et al., 2015)
• "The High-Energy Tail of the Galactic Center Gamma-Ray Excess" (Linden et al., 2016)
• "The Galactic Center GeV Excess from a Series of Leptonic Cosmic-Ray Outbursts" (Cholis et al., 2015)
• "GeV excess and phenomenological astrophysics modeling" (Huang et al., 2015)
• "Discovery of a New Galactic Center Excess Consistent with Upscattered Starlight" (Abazajian et al., 2014)
• "Galactic Center gamma-ray 'excess' from an active past of the Galactic Centre?" (Petrovic et al., 2014)
• "The Spatial Morphology of the Secondary Emission in the Galactic Center Gamma-Ray Excess" (Lacroix et al., 2015)
• "How Bright in Gravitational Waves are Millisecond Pulsars for the Galactic Center GeV Gamma-Ray Excess? A Systematic Study" (Lei et al., 19 Nov 2025)