Fermi Gamma-ray Haze via Dark Matter and Millisecond Pulsars

We study possible astrophysical and dark matter (DM) explanations for the Fermi gamma-ray haze in the Milky Way halo. As representatives of various DM models, we consider DM particles annihilating into W+W-, b-bbar, and e+e-. In the first two cases, the prompt gamma-ray emission from DM annihilations is significant or even dominant at E > 10 GeV, while inverse Compton scattering (ICS) from annihilating DM products is insignificant. For the e+e- annihilation mode, we require a boost factor of order 100 to get significant contribution to the gamma-ray haze from ICS photons. Possible astrophysical sources of high energy particles at high latitudes include type Ia supernovae (SNe) and millisecond pulsars (MSPs). Based on our current understanding of Ia SNe rates, they do not contribute significantly to gamma-ray flux in the halo of the Milky Way. As the MSP population in the stellar halo of the Milky Way is not well constrained, MSPs may be a viable source of gamma-rays at high latitudes provided that there are ~ 20 000 - 60 000 of MSPs in the Milky Way stellar halo. In this case, pulsed gamma-ray emission from MSPs can contribute to gamma-rays around few GeV’s while the ICS photons from MSP electrons and positrons may be significant at all energies in the gamma-ray haze. The plausibility of such a population of MSPs is discussed. Consistency with the Wilkinson Microwave Anisotropy Probe (WMAP) microwave haze requires that either a significant fraction of MSP spin-down energy is converted into e+e- flux or the DM annihilates predominantly into leptons with a boost factor of order 100.

💡 Research Summary

The paper addresses the origin of the diffuse gamma‑ray “haze” observed by the Fermi Large Area Telescope at high Galactic latitudes, a feature that spatially coincides with the microwave haze detected by WMAP. The authors evaluate two broad classes of explanations: annihilating dark matter (DM) and a population of millisecond pulsars (MSPs) residing in the Milky Way’s stellar halo.

First, three representative DM annihilation channels are examined: W⁺W⁻, b b̄, and e⁺e⁻. For the hadronic channels (W⁺W⁻ and b b̄) the prompt gamma‑ray emission from π⁰ decay dominates the spectrum above ∼10 GeV, providing a natural match to the hard component of the Fermi haze. In these cases the inverse‑Compton scattering (ICS) contribution from the secondary electrons and positrons is sub‑dominant. By contrast, the purely leptonic channel (e⁺e⁻) yields almost no prompt photons; the observable gamma‑rays arise almost entirely from ICS of the high‑energy e⁺e⁻ on the interstellar radiation field. To reach the measured intensity, the authors find that a boost factor of order 100 relative to the canonical thermal annihilation cross‑section is required. This boost could stem from sub‑halo clumping or from particle‑physics mechanisms that enhance the annihilation rate at low velocities.

Next, the authors assess conventional astrophysical sources. Type Ia supernovae, despite being common in old stellar populations, have an estimated rate in the halo that is too low to inject sufficient high‑energy particles; their contribution to the gamma‑ray haze is therefore negligible. The focus then shifts to MSPs, which are efficient converters of rotational spin‑down power (∼10³⁴ erg s⁻¹ per object) into relativistic electrons, positrons, and pulsed gamma‑rays. Using the known MSP luminosity function in the Galactic disk and globular clusters as a guide, the authors argue that a halo population of roughly 2 × 10⁴–6 × 10⁴ MSPs would supply a total spin‑down power of 10⁴⁰–10⁴¹ erg s⁻¹. This energy budget can simultaneously account for (i) the few‑GeV pulsed gamma‑ray component observed in the haze and (ii) a broad‑band ICS component extending from sub‑GeV to >100 GeV. Moreover, the same electron/positron population would produce synchrotron radiation in the Galactic magnetic field, naturally linking the gamma‑ray haze to the microwave haze.

To test these ideas, the authors employ GALPROP‑based propagation models. For DM, they compute both prompt and secondary (ICS) gamma‑ray spectra for each annihilation channel, adopting standard NFW or Einasto halo profiles and varying the boost factor. For MSPs, they assume a spatial distribution following the stellar halo density and a characteristic injected electron spectrum per pulsar. The resulting synthetic sky maps are compared with the observed Fermi data and the WMAP microwave residuals. The hadronic DM models reproduce the high‑energy gamma‑ray tail but under‑predict the synchrotron signal; the leptonic DM model matches the synchrotron only if the boost factor is large. The MSP scenario, without any artificial boost, can simultaneously fit both the gamma‑ray and microwave data, provided the halo contains on the order of tens of thousands of MSPs—a number that, while higher than current observational constraints, is not ruled out given the uncertainties in halo MSP demographics.

In conclusion, the paper demonstrates that the Fermi gamma‑ray haze can be explained either by dark‑matter annihilation predominantly into leptons with a substantial boost factor, or by a sizable, yet presently unconfirmed, population of millisecond pulsars in the Galactic halo. Both scenarios predict a significant inverse‑Compton component and a synchrotron counterpart, linking the gamma‑ray and microwave hazes. The authors suggest that future observations—such as deeper Fermi LAT analyses at high latitudes, next‑generation TeV instruments (CTA), and dedicated radio/X‑ray surveys for halo MSPs—will be crucial to discriminate between these possibilities and to refine our understanding of high‑energy processes in the Milky Way’s extended halo.