Dark matter hurricane: Measuring the S1 stream with dark matter detectors

The recently discovered S1 stream passes through the Solar neighborhood on a low inclination, counterrotating orbit. The progenitor of S1 is a dwarf galaxy with a total mass comparable to the present-day Fornax dwarf spheroidal, so the stream is expected to have a significant DM component. We compute the effects of the S1 stream on WIMP and axion detectors as a function of the density of its unmeasured dark component. In WIMP detectors the S1 stream supplies more high energy nuclear recoils so will marginally improve DM detection prospects. We find that even if S1 comprises less than 10% of the local density, multiton xenon WIMP detectors can distinguish the S1 stream from the bulk halo in the relatively narrow mass range between 5 and 25 GeV. In directional WIMP detectors such as CYGNUS, S1 increases DM detection prospects more substantially since it enhances the anisotropy of the WIMP signal. Finally, we show that axion haloscopes possess by far the greatest potential sensitivity to the S1 stream if its dark matter component is sufficiently cold. Once the axion mass has been discovered, the distinctive velocity distribution of S1 can easily be extracted from the axion power spectrum.

Figure 1

The S1 stream in Galactic coordinates with ($X$, $Y$) defining the Galactic plane and $Z$ the height above the disk. This view is partial as it is limited by the footprint of the SDSS- Gaia dataset, whilst the S1 stream extends well beyond the footprint. The arrows show the total Galactocentric velocity of the S1 stars. The Sun and the Sun’s motion are marked as a star and a magenta arrow. Notice that the Sun lies in the path of the counter-rotating S1 stars. A 2 kpc radius sphere and a grey plane are crude representation of the Galactic bulge and the Galactic plane to give a sense of S1’s size and morphology. A triad of velocity vectors (scale of $300\mathrm{km}{\mathrm{s}}^{-1}$) is marked in the bottom of each panel to illustrate the velocity scale.

Figure 2

The S1 stars projected into the ($Y$, $Z$) and ($X$, $Y$) planes. The stream is seen sideways-on (left) and face-on (right) in the two projections. The Sun’s velocity is marked as a yellow arrow, whilst the position of the Sun is indicated by the grey crosshair. S1 has modest inclination with respect to the Galactic plane, but it is broad ($\sim 2\mathrm{kpc}$) as befits its dwarf galaxy origin. The 34 S1 stars plotted here were found in searches through the comparatively local SDSS- Gaia dataset [23] shown as a grey distribution, but the full extent and morphology of the stream awaits searches through the more extensive Gaia Data Release 2.

Figure 3

Laboratory frame speed distributions for the SHM (green) and $\mathrm{SHM}+\mathrm{S}1$ (red) models. The shaded region delimits the range taken by the speed distribution modulated over one year. In the $\mathrm{SHM}+\mathrm{S}1$ model we have assumed that the stream comprises 10% of ${\rho }_0$.

Figure 4

Differential xenon recoil spectra as a function of nuclear recoil energy for DM with a mass 6, 20 and 100 GeV. For each mass we show both the recoil distribution for the SHM (green) and the SHM with a 10% contribution from S1 (red). The cross section in each case is chosen for illustrative purposes and lie near to current exclusion limits. Except for the 20 GeV spectra, the red and green lines are nearly indistinguishable. In blue we show the main nuclear recoil backgrounds. These include the four neutrino backgrounds (${}^8\mathrm{B}$, $hep$, diffuse supernova and atmospheric) as well as the detector and environmental background in LZ, labeled “Expt.,” which we take as a proxy for all xenon detectors. The spectra displayed here do not include the effects of energy resolution and detection efficiency.

Figure 5

Stream discovery limits for left: LZ ($5.6\mathrm{ton}\times 1000\mathrm{day}$ exposure) and right: DARWIN (200 ton-year exposure). The colored shaded regions indicate the values of WIMP masses and cross sections required for the median of each experiment to detect S1 at $3\sigma$ for a given density. The colors from light to dark indicate density fractions from $0.01{\rho }_0$ to ${\rho }_0$. The upper grey boundary in each panel shows the cross sections already excluded by experiment. The lower dashed line shows the neutrino floor for xenon. If S1 comprises less than 10% of the local density these detectors can distinguish the S1 stream from the bulk halo in the narrow mass range between 5 and 25 GeV.

Figure 6

Main: Annual modulation amplitude in $\mathrm{d}R/\mathrm{d}{E}_r$ as a function of recoil energy. The green lines show modulation under the SHM only model, the red lines after the inclusion of a 10% S1 component. Inset: The modulation amplitude of the total rate $R$ for both models as a function of time.

Figure 7

Mollweide projection in galactic coordinates of the value of the double differential angular recoil rate as a function of the inverse of the recoil direction $-\widehat{\mathbf{q}}$ and as a function of energy (5, 10, and 20 keV from left to right). We assume a 20 GeV WIMP and include the angular recoil spectra from both He and ${\mathrm{SF}}_6$ gas with pressures of 740 and 20 torr respectively. The upper three panels show the angular distribution of the SHM model, the middle panels show the distribution after the inclusion of a 10% contribution from S1, while the lower three panels show the same distribution as the middle three but after a smearing by an angular resolution of 30°. In each panel we indicate the direction of ${\mathbf{v}}_{\mathrm{lab}}$ with a white star.

Figure 8

SI (top) and SD (bottom) stream discovery limits as in Fig. 5 but for the CYGNUS-1000 (left) and CYGNUS-100k (right) experiments. In this case we show the neutrino floor for a ${}^{19}\mathrm{F}$ target. In the SI case we can exclude lower masses due to the presence of helium recoils. In the SD case the S1 stream can be distinguished from the smooth halo at much higher masses because the SD WIMP-proton cross section is less well constrained.

Figure 9

Relative change in the sensitivity to the axion-photon coupling for the $\mathrm{SHM}+\mathrm{S}1$ model relative to the SHM alone. We make the comparison as a function of the density fraction of S1 and the dispersion of the stream. The dispersion implied by the stellar stream is shown as the green curve; it is equivalent to a 1-dimensional dispersion of $46\mathrm{km}{\mathrm{s}}^{-1}$. Expressed in this way, ratios greater than one have enhanced sensitivity relative to a stream-less halo, while below one the sensitivity is suppressed. Since axions are easiest to detect when their overall spectrum is sharper, we see that the colder streams give the largest enhancements.

Figure 10

The enhancement in the sensitivity of three experiments: ABRACADABRA, ADMX and MADMAX, due to the presence of the S1 stream. In each case we assume projections of these setups to their final QCD sensitive configuration and scan time. For clarity we have shown the result for a very cold analogue of S1 with a stream dispersion of $5\mathrm{km}{\mathrm{s}}^{-1}$. We also show the already excluded regions of this space from the helioscope CAST [205], and through high energy astrophysical observation (Refs. [206, 207, 208] labelled Fermi, H.E.S.S. and $\mathrm{SN}\text{−}\gamma$ respectively).

Figure 11

Axion power spectra as a function of frequency (shifted by the axion mass $m_a$) and time in months over the year. Left: the signal for the $\mathrm{SHM}+\mathrm{S}1$ model with the dispersion velocity of the stellar stream, equivalent to a 1-d dispersion of $46\mathrm{km}{\mathrm{s}}^{-1}$ Right: the signal for a much colder S1 with a 1-d dispersion of $20\mathrm{km}{\mathrm{s}}^{-1}$. The power has been calculated for an ADMX-like experiment and a DFSZ axion. We assume the power spectra have been obtained for integration times of 0.28 seconds and then averaged over each month. We show Gaussian fluctuations due to thermal white noise for illustrative purposes. For comparison we show the 90% contribution from the SHM as a red dashed line.