Cosmic signals from the hidden sector

Cosmologically long-lived, composite states arise as natural dark matter candidates in theories with a strongly interacting hidden sector at a scale of 10–100 TeV. Light axionlike states, with masses in the 1 MeV–10 GeV range, are also generic, and can decay via Higgs couplings to light standard model particles. Such a scenario is well motivated in the context of very low-energy supersymmetry breaking, where ubiquitous cosmological problems associated with the gravitino are avoided. We investigate the astrophysical and collider signatures of this scenario, assuming that dark matter decays into the axionlike states via dimension six operators, and we present an illustrative model exhibiting these features. We conclude that the recent data from PAMELA, FERMI, and H.E.S.S. points to this setup as a compelling paradigm for dark matter. This has important implications for future diffuse gamma ray measurements and collider searches.

Figure 1

A schematic depiction of the setup.

Figure 2

A schematic picture for the constraints on the $m_a\mathrm{\text{−}}{f}_a$ plane, with the shaded region corresponding to the excluded region. Note that the actual limits on $f_a$ have $O(1)$ uncertainties, as explained in the text. The dominant $a$ decay mode for a given value of $m_a$ is also depicted.

Figure 3

Cascade decays of dark matter $\varphi$ through an axionlike state $a$. Here, ${\varphi }^{\prime }$ is an unstable state in the supersymmetry breaking sector, and ${\ell }^{\pm }$ ($\ell =e$, $\mu$, $\tau$) are standard model leptons.

Figure 4

Regions of best fit (at 68% C.L.) to the PAMELA and FERMI data for dark matter mass $m_{\mathrm{DM}}$ and lifetime ${\tau }_{\mathrm{DM}}$, in the case of direct (solid), 1-step (dashed), and 2-step (dotted) decays into $e^{+}{e}^{-}$, ${\mu }^{+}{\mu }^{-}$, and ${\tau }^{+}{\tau }^{-}$. The best-fit values of $m_{\mathrm{DM}}$ and ${\tau }_{\mathrm{DM}}$ are indicated by the crosses, and are displayed inset in units of TeV and ${10}^{26}\sec$, respectively. Direct decays into $e^{+}{e}^{-}$ does not give a good fit. The case of ${\pi }^{+}{\pi }^{-}{\pi }^0$ is similar to that of ${\tau }^{+}{\tau }^{-}$.

Figure 5

The predicted $e^{\pm }$ fluxes compared to the PAMELA and FERMI data for 1-step cascade decays into $e^{+}{e}^{-}$, ${\mu }^{+}{\mu }^{-}$, and ${\tau }^{+}{\tau }^{-}$. In each case, the mass and lifetime of dark matter are chosen at the best-fit point indicated in Fig. 4, with the background (dotted) and FERMI energy-normalization marginalized as described in the text. We overlay the H.E.S.S. data with energy rescaled in the range $\pm 15\%$ to best match the theory. Note that due to considerable uncertainty in the background fluxes at H.E.S.S. energies, direct comparison of predicted fluxes with the H.E.S.S. data may be misleading.

Figure 6

The same as Fig. 5 for 2-step cascade decays into $e^{+}{e}^{-}$, ${\mu }^{+}{\mu }^{-}$, and ${\tau }^{+}{\tau }^{-}$.

Figure 7

The high energy diffuse $\gamma$-ray spectrum away from the galactic plane, averaged over galactic latitudes above 10° assuming an NFW profile. Shown are the best-fit parameters for 1-step to ${\tau }^{+}{\tau }^{-}$ (dashed), 1-step to ${\pi }^{+}{\pi }^{-}{\pi }^0$ (dot-dashed), and 2-step to ${\mu }^{+}{\mu }^{-}$ (dotted) dark matter decay modes. The $\gamma$-rays are due to ${\pi }^0$ decay in the case of 1-step to ${\tau }^{+}{\tau }^{-}$ and ${\pi }^{+}{\pi }^{-}{\pi }^0$, and to FSR in the case of 2-step to ${\mu }^{+}{\mu }^{-}$. The ${\tau }^{+}{\tau }^{-}$ and ${\pi }^{+}{\pi }^{-}{\pi }^0$ modes produce a bump in the flux clearly distinguishable from the background.