Of course, astronomers can’t actually see these dark-matter halos, but they can see the way in which these invisible blobs can bend light. This phenomenon, known as gravitational lensing, distorts and repositions background objects from our line of sight. Gravitational lensing is pretty cool because it allows astronomers to see, for example, a galaxy that would otherwise be obscured by a closer one in front of it. Importantly, the more dark matter that’s around, the greater the observed lensing effect.

For the new study, Natarajan and her colleagues analyzed images of 11 massive galaxy clusters taken by the Hubble Space Telescope, which were supplemented by spectrographic measurements gathered by the European Southern Observatory’s Very Large Telescope (VLT). The Hubble data, in both visible and infrared light, was taken in 2011 by the telescope’s Advanced Camera for Survey and Wide Field Camera 3.

Data from Hubble and VLT data allowed the researchers to visualize the dark matter. The 3D maps, with their hills, valleys, and exaggerated stalagmites, showed the spatial distribution of the halos. The stalagmites, or peaked regions, showed the location of dark-matter halos, or subhalos in this case, associated with individual galaxies located within a cluster.

The team then took this high-fidelity data and compared it to theory-based computer simulations of clusters with similar masses and at comparable distances. The models did not match the astronomical data; the authors detected smaller lenses in the Hubble images compared to those produced by the simulations.

“There’s a feature of the real universe that we are simply not capturing in our current theoretical models,” explained Natarajan in a Yale press release. “This could signal a gap in our current understanding of the nature of dark matter and its properties, as this exquisite data has permitted us to probe the detailed distribution of dark matter on the smallest scales.”

Bob Jacobsen, a physicist as UC Berkeley who wasn’t involved in the new research, said the two maps—one produced by Hubble data and the other by current dark matter theories—look like they’re in conflict. Interestingly, he said the construction of both maps are “heavily reliant” on computation and simulations.

“This will add important pressure to improving and understanding those computational models and simulations,” explained Jacobsen in an email. “Solid measurements tend to do that. But we don’t yet know whether this is telling us something about our computations and simulations, or whether it’s telling us something fundamental about dark matter.”

Another researcher emphasized just how complex this cosmic science is.

“There are lots of missing pieces in our current understanding of dark matter,” said Esra Bulbul, an astrophysicist at Max Planck Institute for Extraterrestrial Physics who wasn’t involved with the new study. “Comparing state-of-the art simulations and testing current dark matter models with high-quality data, as presented in this work, bring us one step closer to solving this complicated puzzle.”

It’s a frustrating result, for sure, but it’s still a result. And as Jacobsen suggested, we have to get smarter about the whole thing. Solving the mystery of dark matter will require more observations of deep space and some effective number crunching. We just have to know which numbers to crunch.