Preprint: JWST and Keck stellar motions measure M87’s black hole at 5.38 billion suns
A new arXiv preprint reports one of the tightest independent measurements yet for the supermassive black hole in galaxy M87: 5.38 billion times the mass of the sun. The estimate, derived by combining stellar-motion data from NASA’s James Webb Space Telescope and the Keck Observatory in Hawaii, also overlaps with published Event Horizon Telescope constraints, giving astronomers a stronger cross-check between two very different ways of weighing the same black hole.
The study, “The M87 Black Hole Mass from JWST and Keck Stellar Kinematics Spanning 10 pc to 10 kpc,” was posted to arXiv this week and has not yet been peer-reviewed. The paper is by Emily R. Liepold, Chung-Pei Ma, Refa M. Al-Amri and Jonelle L. Walsh of the University of California, Berkeley, and Texas A&M University. M87’s black hole is unusually important because it was the target of the Event Horizon Telescope’s first black hole image, released in 2019.
The authors report a black hole mass of (M_{\mathrm{BH}} = (5.38 \pm 0.13 \pm 0.22) \times 10^9 M_\odot). The first uncertainty is statistical, while the second is a systematic uncertainty dominated by the distance to M87. They also express the result as a mass-to-distance measure, (\theta_g = GM/(c^2D) = 3.16 \pm 0.08) microarcseconds. That is lower than the central value from a cited multi-epoch Event Horizon Telescope analysis, which found (3.62^{+0.41}{-0.34}) microarcseconds and a mass of (6.16^{+0.71}{-0.57} \times 10^9) solar masses. But the new paper says the 90% credible intervals overlap, so the results are consistent within their published ranges rather than contradictory.
That agreement matters because the methods are so different. The Event Horizon Telescope infers black hole properties from horizon-scale imaging, while the new study relies on stellar dynamics — the motions of stars responding to the black hole’s gravity. To do that, the team combined JWST NIRSpec observations of M87’s central 3-by-3-arcsecond region with a wider-field Keck KCWI mosaic, building a continuous two-dimensional map of stellar motions from projected radii of about 10 parsecs to 10 kiloparsecs. They then fit those data with triaxial Schwarzschild stellar-orbit models, using roughly 5,000 kinematic constraints across 618 spatial bins.
The authors said the JWST spectra in the innermost region were affected by quasiperiodic “wiggle” artifacts, and they developed a correction tool called hyphen to clean the data before combining them with the Keck measurements.
The paper says the new mass is consistent with the team’s earlier KCWI-only result, but the added JWST data reduced the statistical uncertainty by about a factor of 2 to 3. That shift is important because it changes what limits the measurement: the biggest remaining uncertainty is no longer the stellar-motion fit itself, but the distance to M87.
For readers, that leaves two takeaways. First, this is still a preprint, not a peer-reviewed paper. Second, the result strengthens an important methodological check. A galaxy-scale estimate based on stellar motions and a horizon-scale estimate based on Event Horizon Telescope imaging are not identical, but they do overlap within uncertainties — a useful sign that both techniques are converging on the same answer for one of astronomy’s most closely watched black holes.