Flattened Darkness: Stellar Streams Outline Invisible Halos

‹›

Dark matter — the invisible foundation of cosmic architecture — makes up 85% of all mass, yet eludes direct observation. In the 1920s, Edwin Hubble discovered the expansion of the universe and the realm of galaxies; by the 1930s, Fritz Zwicky suspected hidden mass in their clusters. The decisive evidence was gathered in the 1970s by Vera Rubin, studying galaxy rotation. Yet it remains a mystery what shape dark matter halos around galaxies take: spherical, elongated, or squashed. The answer is crucial for theories of the Big Bang and understanding the role of dark energy. Now a new approach uses stellar streams — long trails of stars torn from dwarf satellites by tidal forces — as natural probes, like serpentine loops that trace ghostly swirls in space, following every curve of the gravitational landscape.

Unlike work with streams in the Milky Way, where a three-dimensional view is accessible, for extragalactic systems scientists had only their projections on the sky. Using the STRRINGS catalog of 32 streams, they applied a method of direct modeling: for each, they created virtual stellar tails in trial gravitational potentials, assuming an axisymmetric model for the dark matter halo. By comparing the observed trajectory with the model, they estimated the flattening parameter q via Bayesian inference. To account for systematic errors, an additional dispersion term, self-tuned by the data, was added to the likelihood function — a technique akin to methods used in gravitational lensing, where image shape reveals mass distribution. Just as the curve of a melting track in the snow can reveal the contours of an object hidden beneath, so too can we reconstruct the shape of invisible mass.

The results were striking. Selecting a 'golden' sample of 17 streams, where the data are most free of systematics, astronomers computed the population mean flattening: μ_q ≈ 0.72 with an intrinsic scatter σ_q ≈ 0.34. In other words, a typical dark matter halo turned out to be squashed like a pumpkin, not elongated like a melon, and certainly not perfectly round — the spherical case (q=1) is excluded at the ~1.5σ level. This verdict echoes predictions from hydrodynamical simulations like IllustrisTNG and Auriga, where baryonic physics — gas, stars, feedback — compresses and rounds out the initially triaxial dark halos formed in collisionless calculations. The natural damping of primordial asymmetry, much like entropy increases in an isolated system, leaves its imprint on the modern shape. Additionally, an intriguing trend was noticed: squashed halos often align with the disk of the galaxy, as expected from theory.

This work is just a prelude. Upcoming sky surveys like Euclid and Rubin/LSST will multiply the number of known extragalactic streams by orders of magnitude, turning this statistical method into a routine tool of cosmology. The next steps are already clear: explicitly accounting for the stellar disk, freeing the position of the stream progenitor, and transitioning to triaxial potentials. Incorporating information on stream width and surface brightness will further tighten the constraints. In the long run, this will enable mapping the distribution of dark matter in the local universe with unprecedented detail, testing alternative theories of gravity and refining the standard cosmological model. Thus, ghostly arcs of stellar trails become a cartographic grid cast upon an invisible continent.