On the 20 GeV excess in the Milky Way and its implications for dark matter

Recently, Tomonori Totani (University of Tokyo) published a paper proposing that a halo-like gamma-ray excess peaking around 20 GeV in the Milky Way could be explained by dark matter annihilation (publisher , arXiv ). In this post, I summarize the key points of Totani’s analysis and briefly review follow-up studies.

Note: This “~20 GeV halo-like excess” is distinct from the well-known Galactic Center GeV excess (which peaks at a few GeV and is localized toward the inner Galaxy).

20 GeV halo-like excess of the Galactic diffuse emission and implications for dark matter annihilation

Totani analyzed the Milky Way halo region (excluding the Galactic plane), where gamma rays from dark matter annihilation could be detectable with reduced astrophysical foreground contamination. In the WIMP dark matter scenario, dark matter particles can annihilate each other and produce Standard Model particles, including gamma rays.

Using 15 years of Fermi Large Area Telescope (Fermi-LAT) data, Totani reported a statistically significant excess with a spectral peak around 20 GeV. He argued that the signal is well described by an approximately spherical, halo-like component, broadly consistent with expectations from dark matter annihilation in the Milky Way halo.

Interpreting the excess as WIMP annihilation into the $b\bar{b}$ channel, he found preferred parameters of

and a thermally averaged annihilation cross section

This cross section is larger than canonical limits from dwarf spheroidal galaxies, leading to tension with existing constraints.

Follow-up studies and consistency checks

• Wang & Duan (2025) : They estimate the antiproton flux implied by Totani’s best-fit dark matter parameters and compare it with AMS-02 data. They argue that the predicted antiproton flux would overshoot the observations, and infer a much smaller allowed $\langle\sigma v\rangle$, concluding that the dark matter interpretation of the 20 GeV excess is disfavored by AMS-02 antiproton data.

Possible resolutions of the tension with dwarf spheroidal galaxies

This result is interesting enough that several groups have already followed it up (see, e.g., INSPIRE ). Roughly speaking, to address the tension with dwarf spheroidal limits, one may consider the following possibilities:

• Astrophysical uncertainties in dwarf spheroidal galaxies: Constraints from dwarfs depend on the assumed dark matter density profiles and (sometimes) velocity distributions, which can carry significant uncertainties. If the true dark matter distributions differ from the standard modeling assumptions, the inferred limits on $\langle\sigma v\rangle$ could be relaxed, potentially reducing the tension.

• Dark matter model building: The simplest WIMP picture is not the only possibility. Velocity-dependent annihilation or other mechanisms can enhance the annihilation rate in the Milky Way while remaining consistent with dwarf constraints and the relic abundance.

  • Murayama (2025) : Proposes resonant annihilation, which can enhance the annihilation cross section in the Milky Way while keeping it smaller in dwarf spheroidal galaxies.
  • Jho et al. (2025) : Considers a model with a light mediator leading to Sommerfeld enhancement (velocity-dependent annihilation), and discusses consistency with relic abundance, dwarf constraints, and the 20 GeV excess.

• Uncertainties in Galactic gamma-ray background modeling: Modeling the Galactic diffuse gamma-ray emission is complex and can introduce systematic uncertainties. If the background model is imperfect, the inferred excess (and thus the required $\langle\sigma v\rangle$) could be biased high, potentially alleviating the tension.