Late time decay of very heavy dark matter is considered as one of the possible explanations for diffuse PeV neutrinos observed in IceCube. We consider implications of multimessenger constraints, and show that proposed models are marginally consistent with the diffuse $\gamma$-ray background data. Critical tests are possible by a detailed analysis and identification of the sub-TeV isotropic diffuse $\gamma$-ray data observed by Fermi and future observations of sub-PeV $\gamma$ rays by observatories like HAWC or Tibet $\mathrm{AS}+\mathrm{MD}$. In addition, with several-year observations by next-generation telescopes such as IceCube-Gen2, muon neutrino searches for nearby dark matter halos such as the Virgo cluster should allow us to rule out or support the dark matter models, independently of $\gamma$-ray and anisotropy tests.

In the indirect detection of dark matter through its annihilation products, the signals depend on the square of the dark matter density, making precise knowledge of the distribution of dark matter in the Universe critical for robust predictions. Many studies have focused on regions where the dark matter density is greatest, e.g., the galactic center, as well as on the cosmic signal arising from all halos in the Universe. We focus on the signal arising from the whole Milky Way halo; this is less sensitive to uncertainties in the dark matter distribution, and especially for flatter profiles, this halo signal is larger than the cosmic signal. We illustrate this by considering a dark matter model in which the principal annihilation products are neutrinos. Since neutrinos are the least detectable standard model particles, a limit on their flux conservatively bounds the dark matter total self-annihilation cross section from above. By using the Milky Way halo signal, we show that previous constraints using the cosmic signal can be improved on by 1–2 orders of magnitude; dedicated experimental analyses should be able to improve both by an additional 1–2 orders of magnitude.