Strong constraints on dark matter annihilation in Ursa Major III/UNIONS 1

Very recent work has identified a new satellite galaxy, Ursa Major III/UNIONS I, which is the faintest such system ever observed. Dynamical considerations indicate that if the system is in equilibrium, it is likely to be highly dark matter dominated. This, in combination with its proximity, predicts that it may be the preeminent dwarf spheroidal galaxy target for dark matter indirect detection searches. We utilize 15 years of Fermi-LAT data to search for $\gamma$-ray emission from Ursa Major III. Finding no excess, we set strong constraints on dark matter annihilation. Intriguingly, if the high J-factor of Ursa Major III is confirmed, standard thermal dark matter annihilation to $b\overline{b}$ final states would be ruled out for dark matter masses up to 4 TeV. The discovery of Ursa Major III, combined with recent tentative measurements of other high J-factor systems, suggests the exciting possibility that near-future data could produce transformative constraints on thermal dark matter.

Figure 1

Upper limits on the annihilation cross section from Ursa Major III (black solid), evaluated at the tentative J-factor of ${10}^{21}{\mathrm{GeV}}^2{\mathrm{cm}}^{-5}$ [33], with both 0.6 dex error bands motivated by Ref. [37] (dark gray), and the 16–84% error bands estimated by Ref. [34] (light gray), assuming a point-source morphology. The yellow dash-dot line represents the upper limits of a 2D Gaussian distribution with a spatial width of 0.5°, centered on the reported location of Ursa Major III. Considering Ursa Major III an extended source results in the limits $\sim 60\%$ weaker compared to the point-source consideration and comparable to the NFW profile consideration (limits not shown for plot clarity). We also show recent dSph joint-likelihood limits [19] (dashed blue), a fit to the Galactic Center excess [38] (green region), and the thermal annihilation cross section [6] (dotted purple). If the J-factor of Ursa Major III is confirmed, it would dominate indirect detection constraints, ruling out thermal WIMPs up to $\sim 2–4\mathrm{TeV}$.

Figure 2

Residual TS map of the $10°\times 10°$ RoI centered on Ursa Major III (central green cross). Positions of sources removed in the data reduction process are shown as white crosses. The green circle shows 1° distance from the central point. We report no statistically significant excess in 15 years of Fermi-LAT data in either the point-source or extended-source analysis at the position of Ursa Major III.

Figure 3

The upper limit on the spectral energy distribution and the log-likelihood profile of Ursa Major III between 500 MeV and 500 GeV under the assumption of a point source morphology. The arrows correspond to the flux upper limits at the 95% C.L.

Figure 4

Upper limits on the annihilation cross section from Ursa Major III (black) as a function of dark matter mass, evaluated at the tentative J-factor value of ${10}^{21}{\mathrm{GeV}}^2{\mathrm{cm}}^{-5}$ [33] and assuming a point-source model for Ursa Major III. We consider three annihilation channels: $b\overline{b}$ (left), ${\tau }^{+}{\tau }^{-}$ (center), and $W^{+}{W}^{-}$ (right). Also shown are 0.6 dex error bands motivated by Ref. [37] (dark gray) and the 16–84% error bands estimated by Ref. [33] (light gray). The canonical thermal annihilation cross section is shown in purple [6]. The excluded WIMP masses reach up to $\sim 4\mathrm{TeV}$, $\sim 1\mathrm{TeV}$, and $\sim 2\mathrm{TeV}$ for $b\overline{b}$, ${\tau }^{+}{\tau }^{-}$, and $W^{+}{W}^{-}$ annihilation channels, respectively.