Authors: Alex McDaniel, Marco Ajello, Christopher M. Karwin, Mattia Di Mauro, Alex Drlica-Wagner, and Miguel Sànchez-Conde
First Author’s Institution: Department of Physics and Astronomy, Clemson University, Clemson, SC, 29631, USA
Status: Published in Phys. Rev. D [closed access]
As of this article’s publication date, no one knows what dark matter is. (Here’s hoping that statement will be necessary within Astrobites’ lifetime.) Dark matter is essentially invisible, because it doesn’t interact with light or electromagnetically with ordinary matter. We only know it exists due to the gravitational effect it has on visible matter around it. However, the strange properties that make it so elusive are also clues to understand its fundamental nature. Based on these properties, astronomers have theorized a number of dark matter candidates, ranging from primordial black holes to undiscovered particles.
The authors of today’s paper searched for signatures of theorized dark matter particles, Weakly Interacting Massive Particles (WIMPs), in nearby systems known to be dominated by dark matter, small galaxies that orbit the Milky Way called dwarf spheroidal satellite galaxies.
Why WIMPs?
WIMPs are hypothetical particles that could exist as a supersymmetric extension of the Standard Model. They are a promising dark matter candidate because as theorized they don’t absorb light, emit light, or interact strongly with other particles – all properties that match what we know about dark matter. Additionally, the theory is testable because if they exist, they should create gamma-rays which we could actually detect. When WIMP particles and antiparticles collide, they annihilate and leave behind elementary particles like quarks and leptons. These particles can then decay, which emits gamma-rays. Dark matter WIMP annihilation may also resolve the Galactic center excess (GCE), the unexplained surplus of gamma-ray radiation from the center of the Milky Way.
Many WIMP annihilation models, or “channels” exist, but today’s paper explores the two most likely to produce the GCE: the \(b\)\(\bar{b}\ \) channel, which results in a bottom quark and a bottom antiquark, and the \(\tau^{+} \)\(\tau^{-} \) channel, which results in a positive tau lepton and a tau lepton.
Dark matter dominates dwarfs
Today’s authors chose to search for WIMPs in dwarf spheroidal galaxies (dSphs) orbiting the Milky Way because they are nearby, dwarf galaxies tend to have high proportions of dark matter, and they rarely emit contaminating gamma-rays from other astrophysical processes, like supernovae or pulsars. Using 14 years of data from the Fermi-Large Area Telescope (LAT) gamma-ray detector, the authors analyzed 50 dSphs with no nearby gamma-ray sources.
Time to put on your statistics cap…
The dark matter mass for each galaxy was inferred from its stellar kinematics, then converted to a J-factor, which describes the dark matter density. From this, they calculated the gamma-ray flux expected from both annihilation models based on the J-factor, as a function of possible WIMP particle rest masses, cross sections, and energies. Finally, using the observed gamma-ray energy distribution from the Fermi-LAT data, they created likelihood functions for each galaxy to measure how well each annihilation model describes the observed data.
Likelihood functions were also calculated for “blank field” regions of the sky, which were used to calibrate the null hypothesis for the test statistic. Without getting too deep into the statistics, the test statistic compares a galaxy’s likelihood function with the null likelihood function to measure the difference between their distributions – a high test statistic indicates that a galaxy’s observed gamma-ray flux was more likely to have come from the WIMP annihilation model than the background.
As are shown by the colored lines in Figure 2, seven dSphs had excess gamma-ray detection above 2\(\sigma \) (standard deviations) for both annihilation models, but that’s not quite high enough to claim a “true” detection. Typically, astronomers might accept 3\(\sigma \) as proof of detection, but they aren’t satisfied until they’ve reached around 5\(\sigma \).
The authors then carried out a combined likelihood analysis by summing the likelihood functions of their dSphs. This was done for three subsets of the total sample, which varied by how much they trusted the J-factor measurement and lack of contamination for each galaxy. Figure 3 shows that there was no excess detection for the “Measured” subset (the most trustworthy), but the “Benchmark” (slightly less trustworthy) and “Inclusive” (least trustworthy) subsets reveal a marginal excess detection of 2-3\(\sigma \) for both annihilation models. These results predict dark matter WIMP masses of 150 – 230 GeV for the \(b\)\(\bar{b}\ \) channel, and 30 – 50 GeV for the \(\tau^{+} \)\(\tau^{-} \) channel.
Ten years of time will tell
The lack of a significant detection means the jury is still out on whether WIMPs exist or constitute dark matter. However, they could come back with an answer in a decade or so with a larger sample size. The authors estimated that with 10 more years of Fermi-LAT data and the new dSphs that are discovered in the meantime, their ~2\(\sigma \) detection could could jump to ~4\(\sigma \), if it is real.
Astrobite edited by Bình Nguyễn
Featured image credit: Jillian Sparks, Final Project for Introductory Astronomy: Dr. Jeremy Bailin, University of Alabama
