Could Binary Stars Help Trace the Milky Way’s Dark Matter?

How can we measure something we can’t see? When it comes to dark matter, astronomers are always finding new ways to track it down.

On a Hunt for Dark Matter

Galaxy with a big halo [made of dark matter]surrounding it

Illustration of a galaxy and its dark matter halo (shown in blue). [ESO/L. Calçada]

The motion of each star in our galaxy reflects the combined gravitational influence of all the stars, gas, dust, and dark matter in the Milky Way. In theory, we should be able to separate out the effect of the dark matter, giving us a sense of how it’s distributed throughout the galaxy. In practice, though, this is a tricky thing to measure!

Previous work has attempted to suss out the Milky Way’s gravitational field — and, by extension, its dark matter distribution — by measuring tiny shifts in the timing of the signals from extremely dense, rapidly spinning stellar remnants called pulsars. However, pulsars are relatively rare, leading astronomers to search for ways to discern the Milky Way’s dark matter distribution by keeping a close eye on some of the most common stars in the galaxy.

Measuring Midpoints

Sukanya Chakrabarti (Institute for Advanced Study and Rochester Institute of Technology) and collaborators explored the possibility of using binary stars as a probe of the Milky Way’s gravitational field. This technique hinges on making careful measurements of eclipsing binaries — those in which the stars repeatedly pass in front of each other as seen from our perspective. The team proposed that it’s possible to tease out the tiny nudge of the Milky Way’s overall gravitational field by measuring changes in the timing of the eclipse midpoint, when one star is perfectly centered on the other.

plot of time shift in seconds as a function of the binary period in days

The shift in the timing of the eclipse midpoint over the course of a decade due to individual physical mechanisms. The binary system is taken to have nearly circular orbits with an eccentricity of 0.01. The spins are assumed to be unsynchronized, which means that the effect of tidal decay is an upper limit. [Adapted from Chakrabarti et al. 2022]

However, the effect of the Milky Way’s pull is small — shifting the eclipse midpoint by just 0.1 second over the course of a decade — and other factors might also impact the eclipse timing: exoplanets tug on their parent stars, relativistic effects slowly alter elongated orbits, and stars in tight binary systems draw closer together over time. To determine if there are systems in which these other effects are small compared to the effect of the Milky Way’s pull, Chakrabarti and collaborators analyzed a sample of ~800 eclipsing binary systems. They estimated that more than 400 of these systems have orbits that are circular enough and periods that are long enough — lessening relativistic and tidal effects — for the Milky Way signal to be discernible, and they found that exoplanets don’t influence the timing of the eclipse midpoint much at all. The eclipse timing of ~200 of these systems can be measured to within 0.5 second, with some within 0.1 second.

Capable Spacecraft

plot of a simulated eclipse

Example of a simulated eclipse for KIC 4144236, demonstrating that Hubble can measure the eclipse timing to within 0.1 second. Click to enlarge. [Chakrabarti et al. 2022]

While this work by Chakrabarti and coauthors demonstrates the feasibility of this technique, putting it into practice will require patience. The authors demonstrated that the Hubble Space Telescope is able to determine eclipse timings to within 0.1 second for some systems, and JWST and the Nancy Grace Roman Space Telescope will make even more exacting measurements. Luckily, the shift due to the Milky Way’s overall gravitational pull grows over time; the longer we look, the better our measurements and our understanding of the Milky Way’s dark matter distribution will be.


“Eclipse Timing the Milky Way’s Gravitational Potential,” Sukanya Chakrabarti et al 2022 ApJL 928 L17. doi:10.3847/2041-8213/ac5c43