Effective J -factors for dwarf spheroidal with velocity-dependent annihilation

Kimberly K. Boddy,1 Jason Kumar,2 Andrew B. Pace,3, 4 Jack Runburg,2 and Louis E. Strigari5

1Department of Physics & Astronomy, Johns Hopkins University, Baltimore, Maryland 21218, USA 2Department of Physics & Astronomy, University of Hawaii, Honolulu, Hawaii 96822, USA 3Department of Physics and Astronomy, Mitchell Institute for Fundamental Physics and Astronomy, Texas A&M University, College Station, Texas 77843, USA 4McWilliams Center for Cosmology, Carnegie Mellon University, 5000 Forbes Avenue, Pittsburgh, Pennsylvania 15213, USA 5Department of Physics and Astronomy, Mitchell Institute for Fundamental Physics and Astronomy, Texas A&M University, College Station, TX 77843, USA

We calculate the effective J-factors, which determine the strength of indirect detection signals from dark matter annihilation, for 25 dwarf spheroidal galaxies (dSphs). We consider several well-motivated assumptions for the relative velocity dependence of the dark matter annihilation 2 4 cross section: σAv: s-wave (velocity independent), p-wave (σAv v ), d-wave (σAv v ), and ∝ ∝ Sommerfeld-enhancement in the Coulomb limit (σAv 1/v). As a result we provide the largest and most updated sample of J-factors for velocity-dependent∝ annihilation models. For each scenario, we use Fermi-LAT gamma-ray data to constrain the annihilation cross section. Due to the assumptions made in our gamma-ray data analysis, our bounds are comparable to previous bounds on both the p-wave and Sommerfeld-enhanced cross sections using dSphs. Our bounds on the d-wave cross section are the first such bounds using indirect detection data.

I. INTRODUCTION as for p-wave annihilation without assuming a Maxwell- Boltzmann distribution. One of the most promising strategies for the indirect detection of dark matter (DM) is the search for gamma We use a Navarro-Frenk-White (NFW) density pro- rays arising from DM annihilation in dwarf spheroidal file for the dSph halos and assume that the DM velocity galaxies (dSphs). These targets are especially useful be- distribution is related to the density profile by the Ed- cause they have large dark-to-luminous mass ratios, large dington inversion formula [11]. Under the approximation expected DM annihilation rates, and no standard astro- that a dSph spans a small angular size (which is well jus- physical sources of gamma rays. tified for all dSphs we consider), we employ previous work The flux of gamma rays arising from DM annihilation that has determined the effective J-factor in terms of the depends on the properties of the astrophysical source scale density, scale radius, and distance to the halo for through the J-factor. Under the standard assumption all annihilation models we consider [6]. We then estimate of velocity-independent DM annihilation, the J-factor is these parameters by fitting the associated velocity disper- determined by the DM density profile of the dSph. If, sion to stellar data and present results for the effective however, the annihilation cross section is velocity de- J-factors, integrated over various angular cones. pendent, the calculation of the J-factor must account for this velocity dependence by incorporating the full DM velocity distribution [1–8]. Previous works have Finally, we use the MADHAT code [12] to perform a estimated these effective J-factors for some dSphs, us- stacked analysis of Fermi-LAT gamma-ray data [13] for ing a variety of techniques, under the assumptions of these targets. We obtain bounds on the DM annihila- Sommerfeld-enhanced DM annihilation in the Coulomb tion cross section for each of the annihilation models we arXiv:1909.13197v4 [astro-ph.CO] 31 Jul 2020 limit (σAv 1/v)[3,5,7,9, 10] and p-wave annihilation consider. Limits on Sommerfeld-enhanced annihilation 2 ∝ (σAv v )[4,9]. ∝ and p-wave annihilation have been previously obtained In this work, we calculate the effective J-factors using a smaller set of dSphs, with effective J-factors de- for 25 dSphs of the Milky Way (MW), under well- termined using different methodologies [9, 14]. motivated annihilation models: s-wave, p-wave, d-wave, and Sommerfeld-enhancement in the Coulomb limit. We present the first effective J-factor analysis conducted for This paper is organized as follows. In Sec.II, we de- many of these dSphs under certain annihilation scenar- termine the effective J-factors for our set of 25 dSphs, ios. In particular, for Sagittarius II, we perform the first describing our methodology in detail and comparing to J-factor analysis for any annihilation model. Moreover, previous results. In Sec.III, we use these effective J- we are the first to our knowledge to calculate effective factors, along with Fermi-LAT data, to set bounds on the J-factors of any dSph for d-wave annihilation, as well DM annihilation cross section. We conclude in Sec.IV. 2

II. EFFECTIVE J-FACTORS through an interaction that respects Minimal Fla- vor Violation. In this case, annihilation from an s-wave state is chirality-suppressed, and the p-wave We express the DM annihilation cross section as σAv = initial state is forbidden by symmetry of the wave (σAv)0S(v/c), where (σAv)0 is a quantity independent of the relative velocity v of the annihilating particles. The function [17, 18]. differential photon flux arising from DM annihilation in Following Ref. [8], we assume that the DM velocity any astrophysical target is distribution is a function of only two dimensionful pa- rameters: the scale radius r and the scale density ρ . d2Φ (σ v) dN s s A 0 Furthermore, we take the limit θ 1, where θ r /D = 2 JS(Ω) , (1) 0 0 s dΩ dEγ 8πmX dEγ is the characteristic angular scale of the target.≡ Under these assumptions, the effective J-factor for a given an- where dN/dE is the photon spectrum produced per an- γ nihilation model parameter n may thus be written as nihilation and mX is the DM particle mass. We have as- sumed that the DM particle is its own antiparticle. The  2 n/2 ˜ 2 4πGN ρsrs ˜ ˜ effective J-factor, J (Ω), encodes the information about JS(n)(θ) = 2ρ rs JS(n)(θ) , (3) S s c2 the DM distribution in the target. For a target with a ˜ ˜ central potential and DM particles on isotropic orbits, where θ θ/θ0 and J˜S(θ) is the scale-free angular dis- the DM velocity distribution f(r, vp) is simply a function tribution≡ that depends only on n and on the functional of the distance from the center of the target and the ve- form of the velocity distribution, but not on the param- locity of the DM particle [11]. With this simplification, eters ρs, rs, or D. Deviations from this result scale as the effective J-factor is (θ2), which is negligible for the dSphs we consider. O 0 Z ∞ Z Z Therefore, to determine the effective J-factor for any 3 3 JS(θ) = d` d v1 d v2 S( v1 v2 /c) dSph, it is only necessary to know the halo parame- 0 | − | ters (ρs, rs, and D) and scale-free angular distribution f [r(`, θ), v1] f [r(`, θ), v2] , (2) ˜ ˜ × JS(n)(θ). The latter has been previously calculated for −1 −2 where ` is the distance along the line of sight and θ is an NFW density profile ρ(r) = ρs(r/rs) (1 + r/rs) the angle between the light of sight and the line to the and for all values of n discussed above [8]. We make the center of the target. The radial distance from the halo same assumptions as Ref. [8] about the DM velocity dis- center can be recast via r2(`, θ) = `2 + D2 2`D cos θ, tribution, which is related to the density profile through where D is the distance to the center of the− target. the Eddington inversion formula. Using these results, we We consider DM annihilation models of the form are able to determine the various effective J-factors for S(v/c) = (v/c)n, for integer n. In particular, we focus individual dSphs if we know ρs, rs, and D. on the following possible scenarios: In the following subsections, we describe our procedure of determining these parameters and present the resulting (i) n = 1: Sommerfeld-enhanced annihilation in effective J-factors for specific dSphs. the Coulomb− limit [15, 16]. If the annihila- tion proceeds through a heavy mediator, then −1 (σv)0(2παX )(v/c) , where αX is the DM self cou- A. Halo parameters pling. We fix αX = 1/2π. We use the halo parameter analysis, originally pre- (ii) n = 0: s-wave velocity-independent annihilation. sented in Ref. [19], which calculated J-factors for 41 This scenario is the one that is usually considered. dSphs. We consider the subset of 22 dSphs that have (iii) n = 2: p-wave annihilation. This scenario is rel- confidently measured velocity dispersions and are MW evant if DM is a Majorana fermion, which annihi- satellites. The general methodology for determining halo lates to Standard Model fermion/antifermion pairs parameters in dSphs is through a spherical Jeans anal- through an interaction that respects Minimal Flavor ysis [20–22]. The analysis involves solving the spheri- Violation. In this case, annihilation from an s-wave cal Jeans equation (which relates the velocity dispersion, initial state is chirality-suppressed. As another ex- stellar anisotropy, and gravitational potential) for a set ample, this scenario is relevant if DM is a fermion of halo parameters, projecting it into the line-of-sight di- (Dirac or Majorana) that annihilates through a rection, and comparing the line-of-sight dispersion to ob- scalar current coupling, regardless of the final state served stellar kinematics. For our spherical Jeans anal- particles; in this case, the matrix element is only ysis, we assume a Plummer distribution for the stellar non-vanishing if the DM initial state is p-wave (see, density [23], an NFW profile for the DM distribution [24], for example, Ref. [17]). and a constant stellar anisotropy. This analysis includes a total of seven parameters: the (iv) n = 4: d-wave annihilation. This scenario is rele- three needed to find the effective J-factor (ρs, rs, and vant if DM is instead a real scalar [18] that annihi- D) and four others [average line-of-sight velocity, half- lates to Standard Model fermion/antifermion pairs light radius (rp), ellipticity (), stellar anisotropy (β)]. 3

The half-light radius, ellipticity, and distance all contain J-factor integrated over cones with half-angles of 0.5◦ Gaussian priors based on literature measurements. For and 10.0◦, indicating that in most cases the 0.5◦ cone the halo parameters, we assume Jeffreys priors: 2 < encompasses the dSph almost entirely. −3 − log10 (rs/kpc) < 1 and 4 < log10 ρs/(M kpc ) < 14. TableI lists the median integrated effective J-factors The stellar anisotropy prior is uniform in a symmetrized with their 1σ uncertainties for each of our annihilation version: 0.95 < β˜ < 1, where β˜ = β/(2 β).1 scenarios. We provide effective J-factors integrated over To eliminate− some unphysical points in the parameter− cones with half-angles of 0.1◦, 0.2◦, 0.5◦, and 10◦3. space, we use the global density slope-anisotropy inequal- ity, γ?(r) 2β(r), where γ? is the log stellar density slope [25–28≥]. For a Plummer stellar profile and a con- C. Comparison with other approaches stant stellar anisotropy, this constraint is β < 0. We use an unbinned likelihood [22] and determine the pos- We compare our results to other results found in the terior distributions with a multimodal nested sampling literature. In Fig.3, we compare our results for the algorithm [29, 30]. We refer the reader to Ref. [19] for s-wave J-factor (integrated over a 0.5◦ cone) to those more details. found in Refs. [19, 31, 32, 44, 45]. In Fig.4, we com- We also apply the same analysis to three additional pare our results for the s-wave J-factor (integrated over dSphs: Crater II [31], Hydrus I [32], and Sagittarius II2. a 10◦ cone) to the total integrated J-factors found in The literature properties we use for our modeling are Refs. [5,9, 10]. In Fig.5, we compare our results for as follows: D = 117.5 1.1 kpc, rp = 31.1 2.5 ar- the Sommerfeld-enhanced effective J-factor (integrated ± ± ◦ cmin [Crater II, 40]; D = 27.6 0.5 kpc, rp = 7.42 0.58 over a 10 cone) to the total Sommerfeld-enhanced ef- ± ± arcmin,  = 0.21 0.11 [Hydrus I, 32]; and D = 73.1 0.9 fective J-factors found in Refs. [3,5,9, 10]. 4 In Fig.6, ± ± kpc, rp = 1.7 0.05 arcmin [Sagittarius II, 35]. For we compare our results for the p-wave effective J-factor ± Crater II and Hydrus I, we determine membership with (integrated over a 0.5◦ cone) to those in Ref. [4] and for a mixture model including a dSph and MW foreground the p-wave effective J-factor (integrated over a 10◦ cone) component (see Ref. [41] for details on the mixture to the total p-wave effective J-factor found in Ref. [9]. model). We include Gaia DR2 proper motions, which We note here one detail regarding the comparison of helps identify dSph [42, 43]. The J-factors of Hy- our results to those in Ref. [9] for the case of p-wave drus I and Crater II have been presented before [31, 32], or Sommerfeld-enhanced annihilation. In Ref. [9], the and our results are comparable to previous measure- DM velocity distribution is assumed to be Maxwell- ments. This is the first J-factor analysis of Sagittarius II. Boltzmann, with a velocity dispersion that is indepen- dent of position. In this case, the velocity and position integrals in Eq. (2) factorize, and the total effective J- B. Results factor can be written as the product of the total s-wave J-factor and a velocity integral that depends only on the Using the values of rs, ρs, and D from our nested sam- assumed velocity dispersion. In Ref. [9], this integral is pling runs, we are able to produce the posterior distribu- absorbed into the definition of the thermally averaged tions for the effective J-factor [from Eq. (3)], integrated annihilation cross section. For the purposes of compar- over a given angular region of each dSph for different ison, we have rescaled the s-wave J-factors reported in annihilation models. In Fig.1, we show the effective J- Ref. [9] by the appropriate integrals, in order to obtain factors for our set of 25 dSphs, integrated over cones with their total effective J-factors. For the case of p-wave [or opening half-angles of 0.5◦ and 10.0◦, for the scenarios of Sommerfeld-enhanced] annihilation, the rescaling factor 2 −1/2 −1 s-wave, p-wave, d-wave, and Sommerfeld-enhanced an- is 6(v0/c) [or π (v0/c) ], where the values of the nihilation. In each case, we plot the median J-factor, velocity dispersion v0 are also taken from Ref. [9]. along with 68% containment bands. Note that, in gen- eral, there is very little difference between the effective III. CONSTRAINTS ON DARK MATTER ANNIHILATION

1 Generally, β ranges between and 1. Negative and positive Having determined the effective J-factors for our set values correspond to tangential−∞ and radial orbits, respectively. of dSphs under different DM annihilation models, we This alternate parameterization uniformly favors tangential and now constrain DM annihilation to a variety of Standard radial orbits. 2 We note that the identification of Sagittarius II as a dSph versus a cluster is not yet definite. It has a very compact size and high luminosity compared to what is expected for a dSph [33, 34]. The velocity dispersion is resolved, but the mass-to-light ratio 3 The posterior distributions are available at the following website: is much lower than other ultrafaint dSphs, and the https://github.com/apace7/J-Factor-Scaling dispersion is possibly resolved [35]. Furthermore, there are other 4 All of the effective J-factors in these other works have been dwarf galaxies included in our sample whose nature is debated: rescaled to αX = 1/2π for direct comparison with our calcu- Segue 1 [36, 37] and Willman 1[38, 39]. lations. 4

25

20

15 ) 5 − cm 2

(GeV 10 J/ 10 log

5

0 Sommerfeld s-wave p-wave d-wave

0.5◦ 10◦ SclI SxtI HyiI LeoI HerI CarI HorI Seg1 DraI Wil1 FnxI BooI SgrII UMiI RetII LeoII CarII CraII TucII UMaI AquII CBerI CVenI UMaII CVenII

FIG. 1. Median effective J-factors, integrated over cones with opening half-angles of 0.5◦ (hexagons) and 10.0◦ (diamonds), along with asymmetric 68% containment bands. We show results for Sommerfeld-enhanced (green), s-wave (blue), p-wave (red), and d-wave (orange) DM annihilation.

Model final states by performing a stacked dSph analysis imately constant. This process makes it possible to ap- with Fermi-LAT gamma-ray data. We use the MADHAT ply a stacked analysis to any particle physics model in 1.0 software package [12], which is based on the model- this energy range without having to process Fermi-LAT independent analysis framework described in Ref. [14]. data for a given analysis to account for the energy de- MADHAT uses Fermi-LAT Pass 8R3 data [46], collected pendence of the detector. We simply need the gamma- over a time frame of nearly 11 years, and incorporates ray spectrum dN/dEγ for a specific annihilation channel gamma rays only in the energy range of 1-100 GeV, across to produce bounds with MADHAT, and we obtain relevant which the Fermi-LAT effective area is treated as approx- spectra from the tools described in Ref. [47]. 5

In Fig.2, we plot constraints on s-wave, p-wave, d- IV. CONCLUSIONS wave, and Sommerfeld-enhanced DM annihilation to ¯bb, + − ττ¯ ,µµ ¯ , and W W . For each channel, the solid line is There are well-motivated DM models that produce the 95% C.L. bound derived from an analysis of all 25 annihilation cross sections with power-law scalings of dSphs, setting the effective J-factors (integrated over a the relative velocity. For scenarios beyond velocity- ◦ cone with an opening half-angle of 0.5 ) to their median independent s-wave annihilation, standard calculations values, while the uncertainty band arises from adjusting of the astrophysical J-factor for indirect detection must their values upward or downward by their 1σ systematic be modified to account for any velocity dependence. Un- 5 uncertainties. der various simplifying assumptions, we can infer the For comparison, we also plot 95% C.L. bounds on an- DM halo velocity distribution from the density profile us- nihilation to ¯bb from other analyses, represented by the ing the Eddington inversion formula. With the velocity solid gray lines in Fig.2. For the s-wave scenario, we distribution at hand, we incorporate the velocity depen- show the bounds from the Fermi collaboration analy- dence of the annihilation cross section into the calculation sis [48]. These constraints are stronger than ours, but of the J-factor to produce an effective J-factor. the two results are in agreement within the level of their We have determined these effective J-factors for 25 uncertainty bands (the 95% containment bands from dwarf spheroidal galaxies, with assumed NFW halo pro- Ref. [48] are not shown) for mX . 100 GeV. At higher files, for DM annihilation that is s-wave, p-wave, d- DM masses, our s-wave bounds are considerably weaker wave, or Sommerfeld-enhanced in the Coulomb limit. We than those found in Ref. [48]. This discrepancy is likely present the first analysis that we are aware of for several due to our analysis using MADHAT, which limits the pho- dSphs under certain annihilation models. In particular, ton energy range to < 100 GeV in order to achieve a we perform the first analysis for Sagittarius II for any an- model-independent analysis framework, while the analy- nihilation model. Changing the assumed particle physics sis in Ref. [48] uses photons up to an energy of 500 GeV. model for DM annihilation can change the effective J- We also show the bounds on p-wave annihilation found factor by several orders of magnitude. in Ref. [9], recast as a bound on (σv)0. Finally, to provide We have used these effective J-factors and the MADHAT a comparison for Sommerfeld-enhanced annihilation in software package to determine bounds on DM annihila- the Coulomb limit, we obtain bounds using the effective tion in each of these scenarios with a stacked analysis J-factors found in Ref. [3], rescaled to αX = 1/2π. In of Fermi-LAT gamma-ray data from dSphs. The limits both cases, these bounds lie within our 1σ systematic on s-wave annihilation are consistent with those found uncertainty band for mχ < (100) GeV. previously in the literature. O As bounds on d-wave annihilation have not previously Although we have assumed the dSphs have an NFW been determined, we comment on the applicability of our density profile, similar methods can be used for other d-wave constraints. If DM is a real scalar that annihi- profiles, such as the generalized NFW, Burkert, Einasto, lates to a fermion/antifermion pair (φφ ff¯ ) through etc. It would be interesting to see how the choice of a dif- an interaction respecting Minimal Flavor→ Violation, then ferent profile affects the effective J-factors for non-s-wave annihilation from a p-wave initial state is exactly forbid- models of DM annihilation. Another avenue for improve- den, while annihilation from an s-wave initial state is ment in our analysis is to lift the assumption of spherical 2 symmetry. This has been explored in the s-wave J-factor chirality-suppressed by a factor (mf /mφ) . For DM of ∼ case but has not yet been extended to velocity dependent mass mX (10) TeV annihilating to muons, the sup- pression factor∼ O is (10−10), which is still larger than models [49–52]. We leave such a study to future work. the v4-suppression∼ factor O associated with d-wave DM an- Observations of systems other than dSphs may pro- nihilation in a dSph. As such, absent fine-tuning, we vide competitive or stronger limits on velocity-dependent expect d-wave annihilation to dominate s-wave annihila- annihilation. For instance, typical DM particle veloci- tion for DM much heavier than (10) TeV, in which case ties can be quite small in the early Universe, and thus the Fermi-LAT would not be theO ideal instrument to set strong constraints on Sommerfeld-enhanced DM annihi- constraints. Note that these considerations are not nec- lation arise from observations of the cosmic microwave essarily relevant for constraints on p-wave annihilation, background and measurements of light element abun- since there are scenarios in which DM annihilation from dances [53, 54]. How competitive the cosmological con- an s-wave initial state is effectively forbidden, while DM straints are compared to dSphs, however, is model- annihilation from a p-wave state is necessarily dominant. dependent: the velocity behavior of the Sommerfeld en- hancement depends on the mass of the particle mediating the dark matter self-interaction. While limits on p-wave and d-wave annihilation may be stronger from systems with larger characteristic ve- 5 We also obtained 95% C.L. bounds by stacking only dSphs with locities, such as clusters or the Galactic Center, the lim- integrated effective J-factors that are at least 15% of the largest its we have derived are robust, because the DM distri- integrated effective J-factor. Using this subset of dSphs did not butions in dSphs are directly extracted from the data. affect the limits in any significant way. Since baryons contribute a non-negligible amount to the 6

20 s-wave 12 p-wave 10− 10−

21 13 10− 10−

22 14 10− 10− ] 1 −

s 23 15 3 10− 10− [cm 0

) 24 16 10− 10− σv ( + + 25 µ µ− 17 µ µ− 10− + 10− + τ τ − τ τ − + + W W − W W − 26 18 10− b¯b 10− b¯b b¯b, Fermi-LAT [48] b¯b, Zhao et al. 2016 [9] 10 27 10 19 − 10 100 1000 − 10 100 1000

4 d-wave 24 Sommerfeld enhancement 10− 10−

5 10− 25 10−

6 10− 26 10− ] 1 10 7

− −

s 27 3 10− 8

[cm 10− 0

) 28 10− 9 σv

( 10− + 29 W W − 10− 10 + ¯ 10− µ µ− bb + + τ τ − τ τ − 30 11 + 10− + 10− W W − µ µ− b¯b b¯b, Boddy et al. 2018 [6] 10 12 10 31 − 10 100 1000 − 10 100 1000 mχ [GeV] mχ [GeV]

FIG. 2. Exclusion limits at 95% C.L. for s-wave (upper left), p-wave (upper right), d-wave (lower left) and Sommerfeld-enhanced (in the Coulomb limit, lower right) DM annihilation. We consider the following annihilation channels: ¯bb (orange),ττ ¯ (blue),µµ ¯ (red), and W +W − (green). Each central solid line is obtained using the median effective J-factors for all 25 dSphs considered, while the colored bands indicate the limits obtained by varying all effective J-factors either up or down by their 1σ uncertainty. The solid gray lines reproduce the exclusion limits for annihilation to the ¯bb channel at 95% C.L. found in Ref. [48](s-wave), Ref. [9][p-wave, expressed as a bound on (σv)0], and Ref. [14] (Sommerfeld-enhanced, rescaled to αX = 1/2π). potential of clusters and MW-like galaxies in the regions tion boost factors [58]. Future observatories, such as where the DM annihilation signal arises, they represent CTA [59], will target the Galactic Center in particular, an important systematic uncertainty that must be dealt and there are a variety of upcoming and proposed in- with in these systems. Previous studies of p-wave DM an- struments that will improve our understanding of reion- nihilation in the Galactic Center considered the increase ization. It is thus important to compare the ultimate in density and characteristic velocities near the central sensitivity that these types of observations can achieve black hole [55–57] and used these to argue for stronger in comparison to dSphs. constraints on p-wave models. There are also constraints We expect the discovery of many new dSphs from cur- on p-wave annihilation from the epoch of reionization, for rent instruments, such as DECam [e.g., 60, 61] and which the results depend on assumptions of the reioniza- Hyper-Surprime Cam [62, 63], as well as from future tion history of the intergalactic medium and the struc- observatories, such as the LSST [64–66]. If nearby ture formation prescription used to determine annihila- dSphs with large effective J-factors are found, obser- 7 vational sensitivity to DM annihilation could improve on Collider, Dark Matter and Neutrino Physics 2019, significantly. It is standard practice to estimate s- where work on this project began, for their hospitality. wave J-factors for dSphs and dSph-candidates, but we The work of JK is supported in part by DOE Grant de- have demonstrated that it is just as straightforward sc0010504. The work of ABP is supported by NSF grant to estimate the effective J-factors relevant for velocity- AST-1813881. The work of LES is supported by DOE dependent DM annihilation. Since the fundamental na- Grant de-sc0010813. ture of DM interactions is still mysterious, it is important This work has made use of data from the to use data to search for and constrain a variety of DM European Space Agency (ESA) mission Gaia annihilation models. (https://www.cosmos.esa.int/gaia), processed by the Gaia Data Processing and Analysis Consortium (DPAC, https://www.cosmos.esa.int/web/gaia/ ACKNOWLEDGMENTS dpac/consortium). Funding for the DPAC has been provided by national institutions, in particular the We are grateful to Stephen Hill for useful discussions. institutions participating in the Gaia Multilateral JK is grateful to the organizers of the Mitchell Conference Agreement.

[1] Brant Robertson and Andrew Zentner, “Dark Matter interacting dark matter in 20 dwarf spheroidal galaxies,” Annihilation Rates with Velocity-Dependent Annihila- PRD 98, 043017 (2018), arXiv:1712.03188 [astro-ph.CO]. tion Cross Sections,” Phys. Rev. D 79, 083525 (2009), [11] Lawrence M. Widrow, “Distribution Functions for Cuspy arXiv:0902.0362. Dark Matter Density Profiles,” Astrophys. J. Suppl. 131, [2] Francesc Ferrer and Daniel R. Hunter, “The impact of 39–46 (2000). the phase-space density on the indirect detection of dark [12] Kimberly K. Boddy, Stephen Hill, Jason Kumar, Pearl matter,” J. Cosmol. Astropart. Phys. 2013, 005–005 Sand ick, and Barmak Shams Es Haghi, “MAD- (2013), arXiv:1306.6586. HAT: Model-Agnostic Dark Halo Analysis Tool,” , [3] Kimberly K. Boddy, Jason Kumar, Louis E. Strigari, arXiv:1910.02890 (2019), arXiv:1910.02890 [hep-ph]. and Mei-Yu Wang, “Sommerfeld-enhanced J-factors for [13] W. B. Atwood et al. (Fermi-LAT), “The Large dwarf spheroidal galaxies,” Phys. Rev. D 95, 123008 Area Telescope on the Fermi Gamma-ray Space Tele- (2017), arXiv:1702.00408. scope Mission,” Astrophys. J. 697, 1071–1102 (2009), [4] Yi Zhao, Xiao-Jun Bi, Peng-Fei Yin, and Xinmin Zhang, arXiv:0902.1089 [astro-ph.IM]. “Constraint on the velocity dependent dark matter anni- [14] Kimberly K. Boddy, Jason Kumar, Danny Marfatia, and hilation cross section from gamma-ray and kinematic ob- Pearl Sandick, “Model-independent constraints on dark servations of ultrafaint dwarf galaxies,” PRD 97, 063013 matter annihilation in dwarf spheroidal galaxies,” Phys. (2018), arXiv:1711.04696 [astro-ph.HE]. Rev. D 97, 095031 (2018), arXiv:1802.03826. [5] Mihael Petac, Piero Ullio, and Mauro Valli, “On [15] Nima Arkani-Hamed, Douglas P. Finkbeiner, Tracy R. velocity-dependent dark matter annihilations in dwarf Slatyer, and Neal Weiner, “A Theory of Dark Matter,” satellites,” J. Cosmol. Astropart. Phys. 2018, 039–039 Phys. Rev. D 79, 015014 (2009), arXiv:0810.0713. (2018), arXiv:1804.05052. [16] Jonathan L. Feng, Manoj Kaplinghat, and Hai- [6] Kimberly K. Boddy, Jason Kumar, and Louis E. Stri- Bo Yu, “Sommerfeld Enhancements for Thermal Relic gari, “The Effective J-Factor of the Galactic Center for Dark Matter,” Phys. Rev. D 82, 083525 (2010), Velocity-Dependent Dark Matter Annihilation,” Phys. arXiv:1005.4678. Rev. D 98, 063012 (2018), arXiv:1805.08379. [17] Jason Kumar and Danny Marfatia, “Matrix element [7] Thomas Lacroix, Martin Stref, and Julien Lavalle, analyses of dark matter scattering and annihilation,” “Anatomy of Eddington-like inversion methods in the Phys. Rev. D 88, 014035 (2013), arXiv:1305.1611. context of dark matter searches,” J. Cosmol. Astropart. [18] Federica Giacchino, Laura Lopez-Honorez, and Michel Phys. 2018, 040–040 (2018), arXiv:1805.02403. H. G. Tytgat, “Scalar Dark Matter Models with Signif- [8] Kimberly K. Boddy, Jason Kumar, Jack Runburg, and icant Internal Bremsstrahlung,” J. Cosmol. Astropart. Louis E. Strigari, “Angular distribution of gamma-ray Phys. 2013, 025–025 (2013), arXiv:1307.6480. emission from velocity-dependent dark matter annihi- [19] Andrew B. Pace and Louis E. Strigari, “Scaling relations lation in subhalos,” Phys. Rev. D 100, 063019 (2019), for dark matter annihilation and decay profiles in dwarf arXiv:1905.03431. spheroidal galaxies,” MNRAS 482, 3480–3496 (2019), [9] Yi Zhao, Xiao-Jun Bi, Huan-Yu Jia, Peng-Fei Yin, and arXiv:1802.06811 [astro-ph.GA]. Feng-Rong Zhu, “Constraint on the velocity dependent [20] Louis E. Strigari, Savvas M. Koushiappas, James S. dark matter annihilation cross section from Fermi-LAT Bullock, Manoj Kaplinghat, Joshua D. Simon, Marla observations of dwarf galaxies,” PRD 93, 083513 (2016), Geha, and Beth Willman, “The Most Dark-Matter- arXiv:1601.02181 [astro-ph.HE]. dominated Galaxies: Predicted Gamma-Ray Signals [10] Sebastian Bergstr¨om, Riccardo Catena, Andrea Chi- from the Faintest Milky Way Dwarfs,” ApJ 678, 614– appo, Jan Conrad, Bj¨ornEurenius, Magdalena Eriksson, 620 (2008), arXiv:0709.1510 [astro-ph]. Michael H¨ogberg, Susanna Larsson, Emelie Olsson, An- [21] V. Bonnivard, C. Combet, M. Daniel, S. Funk, dreas Unger, and Rikard Wadman, “J -factors for self- A. Geringer-Sameth, J. A. Hinton, D. Maurin, J. I. Read, 8

S. Sarkar, M. G. Walker, and M. I. Wilkinson, “Dark and D. Zaritsky, “A Deeper Look at the New Milky Way matter annihilation and decay in dwarf spheroidal galax- Satellites: Sagittarius II, Reticulum II, Phoenix II, and ies: the classical and ultrafaint dSphs,” MNRAS 453, Tucana III,” ApJ 863, 25 (2018), arXiv:1804.08627. 849–867 (2015), arXiv:1504.02048 [astro-ph.HE]. [35] N. Longeard, N. Martin, E. Starkenburg, R. A. Ibata, [22] A. Geringer-Sameth, S. M. Koushiappas, and M. Walker, M. L. M. Collins, B. P. M. Laevens, D. Mackey, R. M. “Dwarf Annihilation and Decay Emission Pro- Rich, D. S. Aguado, A. Arentsen, P. Jablonka, J. I. Gon- files for Dark Matter Experiments,” ApJ 801, 74 (2015), zalez Hernandez, J. F. Navarro, and R. Sanchez-Janssen, arXiv:1408.0002. “The Pristine Dwarf-Galaxy survey - II. In-depth obser- [23] Henry C. Plummer, “On the problem of distribution in vational study of the faint Milky Way satellite Sagittarius globular star clusters,” Mon. Not. Roy. Astron. Soc. 71, II,” arXiv e-prints (2019), arXiv:1902.02780. 460–470 (1911). [36] M. Niederste-Ostholt, V. Belokurov, N. W. Evans, [24] J. F. Navarro, C. S. Frenk, and S. D. M. White, “A G. Gilmore, R. F. G. Wyse, and J. E. Norris, “The Universal Density Profile from Hierarchical Clustering,” origin of Segue 1,” Monthly Notices of the Royal Astro- ApJ 490, 493–508 (1997), astro-ph/9611107. nomical Society 398, 1771–1781 (2009), arXiv:0906.3669 [25] N. W. Evans, J. An, and M. G. Walker, “Cores and [astro-ph.GA]. cusps in the dwarf spheroidals,” Monthly Notices of [37] J. D. Simon, M. Geha, Q. E. Minor, G. D. Martinez, the Royal Astronomical Society 393, L50–L54 (2009), E. N. Kirby, J. S. Bullock, M. Kaplinghat, L. E. Strigari, arXiv:0811.1488 [astro-ph]. B. Willman, P. I. Choi, E. J. Tollerud, and J. Wolf, [26] Jin H. An and N. Wyn Evans, “A Theorem on Cen- “A Complete Spectroscopic Survey of the Milky Way tral Velocity Dispersions,” ApJ 701, 1500–1505 (2009), Satellite Segue 1: The Darkest Galaxy,” ApJ 733, 46–+ arXiv:0906.3673 [astro-ph.GA]. (2011), arXiv:1007.4198 [astro-ph.GA]. [27] Luca Ciotti and Lucia Morganti, “How general is the [38] M. H. Siegel, M. D. Shetrone, and M. Irwin, “Trimming global density slope-anisotropy inequality?” Monthly No- Down the Willman 1 dSph,” The Astronomical Journal tices of the Royal Astronomical Society 408, 1070–1074 135, 2084–2094 (2008), arXiv:0803.2489. (2010), arXiv:1006.2344 [astro-ph.CO]. [39] B. Willman, M. Geha, J. Strader, L. E. Strigari, J. D. [28] V. Bonnivard, C. Combet, D. Maurin, and M. G. Simon, E. Kirby, N. Ho, and A. Warres, “Willman 1 - A Walker, “Spherical Jeans analysis for dark matter indi- Probable Dwarf Galaxy with an Irregular Kinematic Dis- rect detection in dwarf spheroidal galaxies - impact of tribution,” The Astronomical Journal 142, 128 (2011), physical parameters and triaxiality,” Monthly Notices of arXiv:1007.3499. the Royal Astronomical Society 446, 3002–3021 (2015), [40] G. Torrealba, S. E. Koposov, V. Belokurov, and M. Ir- arXiv:1407.7822 [astro-ph.HE]. win, “The feeble giant. Discovery of a large and diffuse [29] F. Feroz and M. P. Hobson, “Multimodal nested sam- Milky Way dwarf galaxy in the of Crater,” pling: an efficient and robust alternative to Markov Monthly Notices of the Royal Astronomical Society 459, Chain Monte Carlo methods for astronomical data anal- 2370–2378 (2016), arXiv:1601.07178 [astro-ph.GA]. yses,” MNRAS 384, 449–463 (2008), arXiv:0704.3704. [41] Andrew B. Pace, Manoj Kaplinghat, Evan Kirby, [30] F. Feroz, M. P. Hobson, and M. Bridges, “MULTINEST: Joshua D. Simon, Erik Tollerud, Ricardo R. Mu˜noz, an efficient and robust Bayesian inference tool for cos- Patrick Cˆot´e, S. G. Djorgovski, and Marla Geha, mology and particle physics,” MNRAS 398, 1601–1614 “Multiple chemodynamic stellar populations of the Ursa (2009), arXiv:0809.3437. Minor dwarf spheroidal galaxy,” Monthly Notices of [31] N. Caldwell, M. G. Walker, M. Mateo, E. W. Olszewski, the Royal Astronomical Society 495, 3022–3040 (2020), S. Koposov, V. Belokurov, G. Torrealba, A. Geringer- arXiv:2002.09503 [astro-ph.GA]. Sameth, and C. I. Johnson, “Crater 2: An Ex- [42] Gaia Collaboration, A. G. A. Brown, A. Vallenari, tremely Cold Dark Matter Halo,” ApJ 839, 20 (2017), T. Prusti, J. H. J. de Bruijne, C. Babusiaux, C. A. L. arXiv:1612.06398. Bailer-Jones, M. Biermann, D. W. Evans, L. Eyer, and [32] S. E. Koposov, M. G. Walker, V. Belokurov, A. R. Casey, et al., “Gaia Data Release 2. Summary of the con- A. Geringer-Sameth, D. Mackey, G. Da Costa, D. Erkal, tents and survey properties,” AAP 616, A1 (2018), P. Jethwa, M. Mateo, E. W. Olszewski, and J. I. Bailey, arXiv:1804.09365. “Snake in the Clouds: a new nearby dwarf galaxy in the [43] Gaia Collaboration, A. Helmi, F. van Leeuwen, P. J. Magellanic bridge*,” MNRAS 479, 5343–5361 (2018), McMillan, D. Massari, T. Antoja, A. C. Robin, L. Lin- arXiv:1804.06430. degren, U. Bastian, F. Arenou, and et al., “Gaia Data [33] Benjamin P. M. Laevens, Nicolas F. Martin, Edouard J. Release 2. Kinematics of globular clusters and dwarf Bernard, Edward F. Schlafly, Branimir Sesar, Hans- galaxies around the Milky Way,” AAP 616, A12 (2018), Walter Rix, Eric F. Bell, Annette M. N. Ferguson, arXiv:1804.09381. Colin T. Slater, William E. Sweeney, Rosemary F. G. [44] Alex Geringer-Sameth, Savvas M. Koushiappas, and Wyse, Avon P. Huxor, William S. Burgett, Kenneth C. Matthew G. Walker, “A Comprehensive Search for Dark Chambers, Peter W. Draper, Klaus A. Hodapp, Nicholas Matter Annihilation in Dwarf Galaxies,” Phys. Rev. D Kaiser, Eugene A. Magnier, Nigel Metcalfe, John L. 91, 083535 (2015), arXiv:1410.2242. Tonry, Richard J. Wainscoat, and Christopher Waters, [45] Matthew G. Walker, Mario Mateo, Edward W. Ol- “Sagittarius II, Draco II and Laevens 3: Three New szewski, Sergey Koposov, Vasily Belokurov, Prashin Milky Way Satellites Discovered in the Pan-STARRS 1 Jethwa, David L. Nidever, Vincent Bonnivard, III Bailey, 3π Survey,” ApJ 813, 44 (2015), arXiv:1507.07564 [astro- John I., Eric F. Bell, and Sarah R. Loebman, “Magel- ph.GA]. lan/M2FS Spectroscopy of Tucana 2 and Grus 1,” ApJ [34] B. Mutlu-Pakdil, D. J. Sand, J. L. Carlin, K. Spekkens, 819, 53 (2016), arXiv:1511.06296 [astro-ph.GA]. N. Caldwell, D. Crnojevi´c,A. K. Hughes, B. Willman, [46] P. Bruel, T. H. Burnett, S. W. Digel, G. Johannes- 9

son, N. Omodei, and M. Wood (Fermi-LAT), “Fermi- Dark Energy Survey Data,” ApJ 807, 50 (2015), LAT improved Pass 8 event selection,” in 8th Interna- arXiv:1503.02584. tional Fermi Symposium: Celebrating 10 Year of Fermi [61] A. Drlica-Wagner et al. (The DES Collaboration), “Eight Baltimore, Maryland, USA, October 14-19, 2018 (2018) Ultra-faint Galaxy Candidates Discovered in Year Two arXiv:1810.11394 [astro-ph.IM]. of the Dark Energy Survey,” ApJ 813, 109 (2015), [47] Marco Cirelli, Gennaro Corcella, Andi Hektor, Gert arXiv:1508.03622. Hutsi, Mario Kadastik, Paolo Panci, Martti Raidal, Fil- [62] Daisuke Homma, Masashi Chiba, Sakurako Okamoto, ippo Sala, and Alessandro Strumia, “PPPC 4 DM ID: A Yutaka Komiyama, Masayuki Tanaka, Mikito Tanaka, Poor Particle Physicist Cookbook for Dark Matter In- Miho N. Ishigaki, Kohei Hayashi, Nobuo Arimoto, direct Detection,” JCAP 1103, 051 (2011), [Erratum: Jos´eA. Garmilla, Robert H. Lupton, Michael A. Strauss, JCAP1210,E01(2012)], arXiv:1012.4515 [hep-ph]. Satoshi Miyazaki, Shiang-Yu Wang, and Hitoshi Mu- [48] Fermi-LAT Collaboration, “Searching for Dark Matter rayama, “Searches for new Milky Way satellites from the Annihilation from Milky Way Dwarf Spheroidal Galaxies first two years of data of the Subaru/Hyper Suprime-Cam with Six Years of Fermi-LAT Data,” Phys. Rev. Lett. survey: Discovery of Cetus III,” PASJ 70, S18 (2018), 115, 231301 (2015), arXiv:1503.02641. arXiv:1704.05977 [astro-ph.GA]. [49] N. W. Evans, J. L. Sanders, and A. Geringer-Sameth, [63] Daisuke Homma, Masashi Chiba, Yutaka Komiyama, “Simple J-factors and D-factors for indirect dark matter Masayuki Tanaka, Sakurako Okamoto, Mikito Tanaka, detection,” PRD 93, 103512 (2016), arXiv:1604.05599. Miho N. Ishigaki, Kohei Hayashi, Nobuo Arimoto, [50] J. L. Sanders, N. W. Evans, A. Geringer-Sameth, Scott G. Carlsten, Robert H. Lupton, Michael A. Strauss, and W. Dehnen, “Indirect dark matter detection for Satoshi Miyazaki, Gabriel Torrealba, Shiang-Yu Wang, flattened dwarf galaxies,” PRD 94, 063521 (2016), and Hitoshi Murayama, “Bo¨otes.IV. A new Milky Way arXiv:1604.05493. satellite discovered in the Subaru Hyper Suprime-Cam [51] K. Hayashi, K. Ichikawa, S. Matsumoto, M. Ibe, M. N. Survey and implications for the missing satellite prob- Ishigaki, and H. Sugai, “Dark matter annihilation and lem,” PASJ , 91 (2019), arXiv:1906.07332 [astro-ph.GA]. decay from non-spherical dark halos in galactic dwarf [64] Z.ˇ Ivezi´c,J. A. Tyson, E. Acosta, R. Allsman, S. F. An- satellites,” Monthly Notices of the Royal Astronomical derson, J. Andrew, J. R. P. Angel, T. S. Axelrod, J. D. Society 461, 2914–2928 (2016), arXiv:1603.08046. Barr, A. C. Becker, et al., “Lsst: from science drivers [52] Niki Klop, Fabio Zandanel, Kohei Hayashi, and to reference design and anticipated data products,” ApJ Shin’ichiro Ando, “Impact of axisymmetric mass mod- (2008), arXiv:0805.2366v4. els for dwarf spheroidal galaxies on indirect dark mat- [65] J. R. Hargis, B. Willman, and A. H. G. Peter, “Too ter searches,” PRD 95, 123012 (2017), arXiv:1609.03509 Many, Too Few, or Just Right? The Predicted Number [astro-ph.CO]. and Distribution of Milky Way Dwarf Galaxies,” ApJL [53] Douglas P. Finkbeiner, Lisa Goodenough, Tracy R. 795, L13 (2014), arXiv:1407.4470. Slatyer, Mark Vogelsberger, and Neal Weiner, [66] Oliver Newton, Marius Cautun, Adrian Jenkins, Car- “Consistent Scenarios for Cosmic-Ray Excesses from los S. Frenk, and John C. Helly, “The total satellite Sommerfeld-Enhanced Dark Matter Annihilation,” population of the Milky Way,” MNRAS 479, 2853–2870 JCAP 1105, 002 (2011), arXiv:1011.3082 [hep-ph]. (2018), arXiv:1708.04247 [astro-ph.GA]. [54] Junji Hisano, Masahiro Kawasaki, Kazunori Kohri, Takeo Moroi, Kazunori Nakayama, and Toyokazu Sekiguchi, “Cosmological constraints on dark matter models with velocity-dependent annihilation cross sec- tion,” Phys. Rev. D 83, 123511 (2011), arXiv:1102.4658. [55] Jessie Shelton, Stuart L. Shapiro, and Brian D. Fields, “Black hole window into p-wave dark matter annihilation,” Phys. Rev. Lett. 115, 231302 (2015), arXiv:1506.04143 [astro-ph.HE]. [56] Pearl Sandick, Kuver Sinha, and Takahiro Yamamoto, “Black Holes, Dark Matter Spikes, and Constraints on Simplified Models with t-Channel Mediators,” Phys. Rev. D98, 035004 (2018), arXiv:1701.00067 [hep-ph]. [57] Christian Johnson, Regina Caputo, Chris Karwin, Si- mona Murgia, Steve Ritz, and Jessie Shelton, “Search for gamma-ray emission from p-wave dark matter annihi- lation in the Galactic Center,” Phys. Rev. D99, 103007 (2019), arXiv:1904.06261 [astro-ph.HE]. [58] Hongwan Liu, Tracy R. Slatyer, and Jess Zavala, “Con- tributions to cosmic reionization from dark matter an- nihilation and decay,” Phys. Rev. D94, 063507 (2016), arXiv:1604.02457 [astro-ph.CO]. [59] B. S. Acharya et al. (CTA Consortium), Science with the Cherenkov Telescope Array (WSP, 2018) arXiv:1709.07997 [astro-ph.IM]. [60] K. Bechtol et al. (The DES Collaboration), “Eight New Milky Way Companions Discovered in First-year 44 23 12 33 09 39 28 18 12 10 39 51 25 09 15 30 70 05 53 09 39 23 32 12 39 48 24 12 35 11 41 29 20 13 10 38 59 28 10 16 32 16 06 39 10 44 22 33 12 42 ...... 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 ◦

+0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +1 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − 10 10 70 68 75 12 25 03 46 34 08 36 84 42 34 81 03 53 03 88 71 21 30 68 85 07 82 ...... 22 22 21 22 22 23 23 20 23 22 21 23 23 21 22 23 22 22 23 22 23 22 23 23 23 42 21 12 31 10 38 24 21 12 11 39 48 21 08 15 25 80 07 53 15 41 22 30 15 38 47 22 13 32 12 39 24 24 13 10 37 57 23 10 16 26 24 08 40 15 43 22 33 13 38 ...... 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 ◦ +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +1 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − 5 . 0 62 54 69 05 20 89 30 20 97 34 79 34 14 77 03 36 95 84 59 13 16 61 66 05 72 ...... 22 22 21 22 22 22 23 20 22 22 21 23 23 21 22 23 21 22 23 22 23 22 23 23 23 42 23 15 31 13 39 25 24 14 12 39 47 24 11 15 26 87 09 54 20 42 25 32 18 40 48 24 17 32 15 41 25 26 15 11 38 58 25 13 16 28 31 10 43 19 45 25 36 15 39 ...... 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 ◦ +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +1 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − Sommerfeld 2 . 0 54 41 62 96 13 78 16 06 85 30 72 27 00 70 03 23 87 77 50 03 03 53 51 99 63 ...... 22 22 21 21 22 22 23 20 22 22 21 23 23 21 22 23 21 22 23 22 23 22 23 22 23 . 43 25 18 31 16 41 26 25 15 13 41 48 26 13 17 28 92 11 58 24 44 28 33 21 41 48 26 21 33 18 43 27 28 18 12 39 58 28 17 16 30 39 12 47 21 47 28 39 16 43 ...... 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 ◦ +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +1 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − 1 . 0 45 28 53 87 04 66 03 92 72 22 65 20 86 62 02 10 77 67 40 92 91 44 37 92 54 ...... 22 22 21 21 22 22 23 19 22 22 21 23 22 21 22 23 21 22 23 21 22 22 23 22 23 54 63 . . 56 17 54 46 26 51 62 46 10 31 03 61 47 23 66 37 16 50 26 44 76 39 19 32 86 42 04 52 50 93 75 75 10 68 39 85 85 26 13 26 30 61 66 36 64 29 47 66 ...... 0 1 1 0 1 0 1 1 0 0 1 2 1 0 0 1 1 0 1 0 1 0 1 0 1 ◦ − +0 − +1 +1 − +1 − +1 − +1 − +0 − +1 − +1 − +0 − +0 − +1 − +2 − +1 − +0 − +0 − +2 − +0 − +2 − +0 − +1 − +1 − +1 − +0 − +2 − 10 54 . 70 29 45 16 37 73 50 50 16 38 67 42 00 19 75 65 67 42 53 27 25 73 01 35 ...... 2 2 2 1 2 1 1 3 3 2 0 3 2 2 1 2 2 1 1 3 2 4 2 3 0 − 46 50 . . 47 96 51 36 23 34 32 36 09 28 96 24 45 23 40 14 44 22 23 73 10 19 25 40 63 99 82 31 40 48 19 49 10 48 03 16 72 26 40 21 68 34 36 02 15 24 86 84 ...... 0 1 0 0 1 0 1 1 0 0 1 1 1 0 0 1 0 1 0 1 0 1 0 1 1 ◦ +0 − 5 − − − − − − − − − − − − − − − − − − − − − − − +1 +0 +0 +1 +0 +1 +1 +0 +0 +1 +2 +1 +0 +0 +1 +0 +1 +0 +1 +1 +1 +0 +1 +1 − . 84 0 . 58 01 41 04 34 49 16 32 14 34 60 00 97 19 47 61 57 32 23 22 83 71 94 10 ...... 2 2 2 1 2 1 1 3 3 2 0 3 2 1 1 2 2 1 1 3 2 3 2 2 0 − -wave 46 50 48 02 . . . . d 34 78 43 22 18 23 07 26 10 20 80 99 36 23 16 10 38 18 11 69 94 17 12 52 83 60 17 27 30 03 35 10 32 84 00 54 26 17 13 36 21 24 78 01 19 54 ...... 0 1 1 0 0 1 0 1 1 0 0 1 1 0 0 0 1 0 1 0 1 0 0 0 1 ◦ +0 − +2 − 2 +1 − +0 − +0 − +1 − +0 − +1 − +1 − +0 − +0 − +1 − +1 − +1 − +0 − +0 − +1 − +0 − +1 − +0 − +1 − +0 − +1 − +0 − +1 − . 22 13 0 . . 37 65 27 81 20 21 77 01 04 21 42 58 84 18 12 45 45 08 91 09 44 59 78 ...... 3 0 2 1 1 1 1 1 2 3 2 0 3 1 1 1 2 2 1 1 2 2 3 2 2 − − 46 61 49 20 . . . . 25 68 35 09 15 15 92 21 12 12 67 84 27 23 99 08 36 18 04 63 87 17 00 44 74 45 08 18 18 94 27 12 18 71 90 40 24 04 09 15 19 18 65 93 18 33 ...... 0 1 https://github.com/apace7/J-Factor-Scaling 1 0 0 1 0 1 0 0 0 1 1 0 0 0 0 0 1 0 1 0 0 0 1 ◦ +0 − +2 − 1 +1 − +0 − +0 − +1 − +0 − +1 − +0 − +0 − +0 − +1 − +1 − +0 − +0 − +0 − +1 − +0 − +1 − +0 − +1 − +0 − +0 − +0 − +1 − . 60 39 0 . . 11 30 07 54 98 91 40 68 86 05 23 20 65 16 78 21 29 78 57 88 07 39 58 ...... 3 0 2 1 1 1 0 0 2 2 1 0 3 1 1 1 1 2 1 0 2 1 3 2 2 − − -wave, and Sommerfeld-enhanced annihilation for cones with opening half-angles 76 05 05 30 09 39 03 09 99 93 14 89 08 08 08 55 93 29 15 47 09 59 21 15 13 08 18 60 27 38 40 82 34 97 16 37 91 27 18 07 17 ...... 63 03 29 42 10 53 20 42 33 ...... d ...... 0 1 1 0 0 1 1 0 0 0 0 0 1 1 1 0 0 0 0 0 0 0 0 1 0 ◦ +0 − +1 − +1 − +0 − +0 − +1 − +1 − +0 − +1 − +1 − +0 − +1 − +1 − +1 − +1 − +0 − +0 − +1 − +0 − +0 − +1 − +0 − +0 − +1 − +0 − 10 24 06 29 13 05 39 58 55 24 90 68 14 51 86 32 26 38 95 54 43 78 72 33 83 56 ...... 9 9 9 6 8 9 9 8 9 10 10 11 11 10 11 10 10 11 11 10 11 10 10 10 10 -wave, p 78 21 09 30 71 08 82 69 14 82 99 85 02 51 73 25 34 10 92 71 15 11 09 68 13 80 96 57 58 31 88 14 90 34 88 25 18 14 13 ...... 60 45 87 21 94 36 96 40 20 60 16 ...... 0 0 0 1 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 1 0 ◦ +0 − +0 − +1 − +0 − +0 − +0 − +0 − +1 − +0 − +0 − +1 − +0 − +0 − +0 − 5 +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +1 − +0 − . 0 96 97 29 51 96 51 65 05 02 19 37 59 21 20 -wave, 98 33 82 50 83 14 72 69 33 62 45 ...... s 9 9 9 9 9 6 8 9 9 8 9 10 10 11 10 10 11 10 11 10 10 10 10 10 10 -wave 61 15 18 06 76 61 14 72 90 66 01 45 61 17 25 06 85 63 14 87 03 65 78 45 46 25 77 12 82 35 10 82 53 19 18 27 14 ...... 49 30 76 13 83 39 10 84 55 25 20 77 15 ...... p ) for ...... 0 0 1 0 0 0 0 0 0 0 1 0 5 0 0 0 0 0 0 0 0 0 0 0 1 0 ◦ +0 − +0 − +1 − +0 − +0 − +0 − +0 − +0 − +1 − +0 − +0 − +0 − 2 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +1 − +0 − . 0 63 70 12 36 68 17 53 90 19 28 08 05 68 20 62 37 57 81 93 61 83 58 33 42 21 ...... cm 9 9 9 9 9 5 9 8 9 9 9 8 9 2 10 10 11 10 10 11 10 10 10 10 10 10 53 14 07 07 74 58 17 64 54 15 19 08 82 61 17 74 87 42 21 70 12 77 38 13 78 46 14 18 56 39 99 18 41 ...... 98 45 23 71 13 81 41 12 77 49 16 20 56 95 72 20 41 ...... GeV 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 ◦ +0 − +0 − +1 − +0 − +0 − +0 − +0 − +0 − 1 +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +1 − +0 − +0 − +0 − J/ . ( 0 32 41 93 14 40 84 35 73 99 38 01 39 16 29 48 78 46 51 41 30 99 20 89 93 85 ...... 10 9 9 9 9 9 9 5 9 8 9 9 9 9 8 9 8 9 10 10 10 10 10 10 10 10 62 38 17 53 10 59 50 23 16 09 54 80 47 12 15 53 79 05 68 10 56 32 49 11 52 72 45 24 55 11 66 57 25 20 10 58 87 57 19 16 64 23 05 63 12 65 37 54 11 73 ...... 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 ◦ +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +1 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − 10 48 39 49 92 88 57 25 56 96 11 35 28 93 64 66 16 48 63 12 87 13 40 72 80 46 ...... 18 18 17 17 17 18 19 15 18 18 17 19 18 17 17 19 17 18 19 17 19 18 19 18 19 58 28 15 47 09 52 36 25 12 10 52 70 31 11 15 38 91 05 68 11 50 28 39 11 46 65 30 16 48 09 54 36 27 12 10 51 80 32 13 16 38 36 05 48 12 56 28 41 12 52 ...... 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 ◦ +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +1 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − 5 . 0 39 19 43 82 83 37 01 35 82 09 29 17 65 61 66 94 35 58 00 75 93 33 44 76 36 ...... The posterior distributions are available at the following website: ◦ 18 18 17 17 17 18 19 15 18 18 17 19 18 17 17 18 17 18 19 17 18 18 19 18 19 -factors presented as log -wave 57 27 14 43 11 50 32 28 13 11 51 65 28 10 15 32 02 07 67 17 51 28 39 16 43 63 29 15 44 13 52 32 30 13 11 49 75 30 10 15 34 47 08 48 18 55 28 42 14 45 J ...... s 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 ◦ +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +1 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − +0 − 2 . , and 10 0 26 99 33 68 72 19 81 13 63 02 20 05 42 52 65 73 22 47 87 58 74 21 21 68 24 ◦ ...... 18 17 17 17 17 18 18 15 18 18 17 19 18 17 17 18 17 18 18 17 18 18 19 18 19 , 0.5 56 28 17 41 14 51 32 29 14 13 52 62 29 11 16 32 09 10 68 22 53 29 40 21 45 62 30 18 42 17 53 31 33 16 12 49 75 31 13 16 34 60 11 53 22 57 31 45 17 44 ◦ ...... 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 ◦ +0 +0 +0 +0 +0 +0 +0 +0 +0 +0 +0 +0 +0 +0 +0 +0 +1 +0 +0 +0 +0 +0 +0 +0 +0 − − − − − − − − − − − − − − − − − − − − − − − − − 1 . , 0.2 0 12 79 19 53 57 02 61 91 43 91 09 92 21 38 63 54 07 33 73 40 56 07 00 56 11 ...... ◦ 18 17 17 17 17 18 18 14 18 17 17 18 18 17 17 18 17 18 18 17 18 18 19 18 19 TABLE I. Integrated of 0.1 Galaxy Aquarius II Bo¨otesI Canes Venatici I Canes Venatici II Carina I Carina II Coma Berenices I Crater II Draco I Fornax I Hercules I Horologium I Hydrus I Leo I Leo II Reticulum II Sagittarius II Sculptor I Segue 1 Sextans I Tucana II Ursa Major I Ursa Major II Ursa Minor I Willman 1 11

Comparison of J-factors at 0.5◦ for s-wave annihilation

20

19 ) 5

− 18 cm 2 (GeV J/ 10

log 17

16

This analysis Geringer-Sameth et al. 2015 [44] Pace et al. 2019 [19] Walker et al. 2016 [45] Caldwell et al. 2017 [31] 15 Koposov et al. 2018 [32] SclI SxtI HyiI LeoI HerI CarI HorI Seg1 DraI Wil1 BooI FnxI SgrII UMiI RetII LeoII CarII CraII TucII UMaI AquII CBerI CVenI UMaII CVenII

FIG. 3. J-factors for s-wave DM annihilation integrated over a 0.5◦ cone from this analysis and others. 12

Comparison of total integrated J-factors for s-wave annihilation

20

19 ) 5 − cm 2 18 (GeV J/ 10 log

17

16

This analysis Bergstrom et al. 2018 [10] Petac et al. 2018 [5] Zhao et al. 2016 [9] SclI SxtI HyiI LeoI HerI CarI HorI Seg1 DraI Wil1 BooI FnxI SgrII UMiI RetII LeoII CarII CraII TucII UMaI AquII CBerI CVenI UMaII CVenII

FIG. 4. J-factors for s-wave DM annihilation integrated over a 10◦ cone from this analysis and total J-factors from other analyses. 13

Comparison of total integrated J-factors for Sommerfeld-enhanced annihilation 25

24 )

5 23 − cm 2 (GeV J/ 10

log 22

21

This analysis Bergstrom et al. 2018 [10] Petac et al. 2018 [5] Boddy et al. 2017 [3] Zhao et al. 2016 [9] 20 SclI SxtI HyiI LeoI HerI CarI HorI Seg1 DraI Wil1 BooI FnxI SgrII UMiI RetII LeoII CarII CraII TucII UMaI AquII CBerI CVenI UMaII CVenII

FIG. 5. Effective J-factors for Sommerfeld-enhanced DM annihilation integrated over a 10◦ cone from this analysis and total J-factors from other analyses. The total effective J-factors from other analyses have been rescaled to αX = 1/2π for comparison. 14

Comparison of integrated J-factors for p-wave annihilation

13

12

11 ) 5

− 10 cm 2 (GeV J/ 9 10 log

8

7

This analysis, 0.5◦ 6 Zhao et al. 2018 [4], 0.5◦ This analysis, 10◦ Zhao et al. 2016 [9], total SclI SxtI HyiI LeoI HerI CarI HorI Seg1 DraI Wil1 BooI FnxI SgrII UMiI RetII LeoII CarII CraII TucII UMaI AquII CBerI CVenI UMaII CVenII

FIG. 6. Effective J-factors for p-wave DM annihilation, integrated over a cone of 0.5◦ (hexagons) or 10.0◦ (diamonds), or the total integrated effective J-factor (diamonds), as indicated in the legend.