Limits on GeV-scale WIMP Annihilation in Dwarf Spheroidals with IceCube DeepCore

Abstract

Dark matter is approximately five times more abundant than baryonic matter in the universe, but its physical nature continues to elude physicists. One potential candidate for dark matter is a weakly-interacting massive particle (WIMP), which is predicted by various extensions to the Standard Model (SM) of particle physics. After becoming gravitationally bound in cosmic structures, WIMPs can self-annihilate and produce SM particles including neutrinos, which are observable by detectors like IceCube. We present a search for neutrinos from low-mass $(\leq 300 \, \mathrm{GeV})$ WIMP annihilation in dwarf spheroidal galaxies with over seven years of IceCube livetime. We find no statistically significant evidence of neutrinos produced by WIMP annihilation, and therefore set upper limits on the velocity-averaged annihilation cross section $\left<σv\right>$. Our strongest upper limits at the 90\% confidence level are $\mathcal{O}\!\left(10^{-22} \, \mathrm{{cm}^{3} \, s^{-1}}\right)$ for WIMP annihilation directly into neutrino-antineutrino pairs. For our least sensitive channel, the corresponding limits are $\mathcal{O}\!\left(10^{-20} \, \mathrm{{cm}^{3} \, s^{-1}}\right)$, which is an improvement of over two orders of magnitude over previous IceCube limits from dwarf galaxies at the upper end of our mass range.

Figures (12)

• Figure 1: WIMP annihilation spectra for WIMP masses of $10 \, \mathrm{GeV}$ (upper left), $100 \, \mathrm{GeV}$ (upper right), $150 \, \mathrm{GeV}$ (lower left), and $300 \, \mathrm{GeV}$ (lower right). Note the physicality constraint for the $W^{+}W^{-}$ channel ($m_{\chi} \geq 90 \, \mathrm{GeV}$), and the production of the low-energy tail in the $\nu\bar{\nu}$ spectra in the bottom row.
• Figure 2: The locations of dSphs (red crosses) plotted over a skymap of one year of neutrino data with randomized right ascensions of events. The colorbar is events per pixel and the pixel sizes correspond to $N_{\text{side}} = 32$Gorski2005.
• Figure 3: Upper limits at 90% confidence level (solid lines), with comparison to IceCube 2013 Aartsen2013 and Fermi 2017 Albert2017 results (dashed and dotted lines, respectively).
• Figure 4: Upper limits for individual channels (solid lines) with median sensitivities (dashed lines), $\pm 1\sigma$ and $2\sigma$ Brazil bands (shaded regions), and upper limits under systematic uncertainties (dotted lines).
• Figure 5: Upper limits for $b\bar{b}$ (top) and $\tau^{+}\tau^{-}$ (bottom). Results from this work are shown in black in the same format as Fig. \ref{['fig:upper_limits_bands']}. Comparisons are made to previous results from Fermi+MAGIC 2016 MAGIC2016, Fermi 2017 Albert2017, H.E.S.S 2020 Abdallah2020, IceCube 2013 Aartsen2013, and MAGIC 2022 Acciari2022.