Possible evidence for dark matter annihilations from the excess microwave emission around the center of the Galaxy seen by the Wilkinson Microwave Anisotropy Probe

The Wilkinson Microwave Anisotropy Probe (WMAP) experiment has revealed an excess of microwave emission from the region around the center of our Galaxy. It has been suggested that this signal, known as the “WMAP haze,” could be synchrotron emission from relativistic electrons and positrons generated in dark matter annihilations. In this article, we revisit this possibility. We find that the angular distribution of the WMAP haze matches the prediction for dark matter annihilations with a cusped density profile, $\rho (r)\propto r^{-1.2}$ in the inner kiloparsecs. Comparing the intensity in different WMAP frequency bands, we find that a wide range of possible weakly interacting massive particle (WIMP) annihilation modes are consistent with the spectrum of the haze for a WIMP with a mass in the 100 GeV to multi-TeV range. Most interestingly, we find that to generate the observed intensity of the haze, the dark matter annihilation cross section is required to be approximately equal to the value needed for a thermal relic, $\sigma v\sim 3\times {10}^{-26}{\mathrm{cm}}^3/\mathrm{s}$. No boost factors are required. If dark matter annihilations are in fact responsible for the WMAP haze, and the slope of the halo profile continues into the inner Galaxy, GLAST is expected to detect gamma rays from the dark matter annihilations in the galactic center if the WIMP mass is less than several hundred GeV.

Figure 1

The specific intensity observed by WMAP as a function of the angle from the galactic center (GC). The frames correspond to the five WMAP frequency bands. Following the treatment of Ref. 9, all WMAP pixels near the galactic plane are masked where excessive dust makes the $\mathrm{H}\alpha$-based free-free template unreliable. The chance correlation of the CMB anisotropy with the haze leads to a systematic error in the derived spectrum of the haze, effectively yielding some linear combination of the true haze spectrum and the CMB spectrum in the fit. This error is of order 5% at 23 GHz, and 50% at 61 GHz, so the spectral index derived from the $23\mathrm{GHz}/33\mathrm{GHz}$ ratio is the most meaningful. See Ref. 9 for details.

Figure 2

The specific intensity of microwave emission in the 22 GHz WMAP channel as a function of the angle from the galactic center, compared to the synchrotron emission from the annihilation products of a 100 GeV WIMP annihilating to $e^{+}{e}^{-}$. In the upper frame, our default diffusion parameters have been used. The solid line denotes the choice of a NFW halo profile, while the dashed line is the result from a profile with a somewhat steeper inner slope, $\rho (r)\propto r^{-1.2}$. In the lower frame, we have used a NFW profile with our default propagation parameters (solid line), and with a smaller diffusion zone of $L=2\mathrm{kpc}$ (dashed line), and a longer energy loss time of $\tau (1\mathrm{GeV})=4\times {10}^{15}\mathrm{s}$ (dotted line).

Figure 3

Upper frame: The ratio of intensities in the 22 and 33 GHz WMAP frequency bands of synchrotron from dark matter annihilation products, as a function of the WIMP mass, for several possible annihilation modes. The intensities were averaged between 6° and 15°. The horizontal dashed lines reflect the $2\sigma$ statistical range measured by WMAP. Lower frame: The WIMP annihilation cross section (times boost factor) required to produce the intensity of the WMAP haze. For a wide range of masses and annihilation modes, the cross section required is within a factor of approximately two of the value required of a $s$-wave thermal relic, $\sigma v\sim 3\times {10}^{-26}{\mathrm{cm}}^3/\mathrm{s}$. In each frame, the contours denote the following annihilation modes: $e^{+}{e}^{-}$ (solid lines), ${\mu }^{+}{\mu }^{-}$ [gray (blue) dot-dashed line], ${\tau }^{+}{\tau }^{-}$ (dot-dashed line), $W^{+}{W}^{-}$ (dashed line), $ZZ$ gray (blue) dashed line], and $b\overline{b}$ (dotted line).