Stellar motions sharpen the estimate of M87’s black hole mass

The black hole at the center of the giant elliptical galaxy M87 has been weighed through stars, gas and the glowing material around it. A new stellar-motion analysis puts its mass at 5.38 billion times that of the Sun, closely matching the team’s earlier Keck-only result while narrowing the statistical uncertainty.

Emily R. Liepold, Chung-Pei Ma, Refa M. Al-Amri and Jonelle L. Walsh report the result in an arXiv preprint. Their approach combines JWST’s detailed view of the inner galaxy with Keck’s wider coverage. It does not photograph the black hole or track individual stars around the event horizon: it infers mass from the collective motions encoded in starlight.

Two views of the same gravitational system

The authors combine infrared spectra from JWST’s NIRSpec instrument with optical spectra from the Keck Cosmic Web Imager. Together, the measurements span projected distances from roughly 10 parsecs to 10 kiloparsecs from M87’s center. This broad coverage helps constrain the gravitational contributions of the black hole, stars and dark matter.

Absorption features in the spectra carry information about the distribution of stellar speeds along our line of sight. The researchers fit stellar templates to these features, measuring both the spread of velocities and departures from a simple bell-shaped distribution. Across 618 spatial bins, they obtain 4,944 kinematic measurements for their dynamical models.

In the innermost JWST measurements, the velocity dispersion—the spread of stellar velocities—reaches about 500 kilometers per second. The JWST and Keck measurements agree in their overlapping region, despite using different instruments, wavelengths and template libraries. That agreement supports the authors’ interpretation, although it cannot by itself eliminate shared modeling uncertainties.

Removing an instrumental imitation

Reaching inward required dealing with a misleading signal. Near M87’s bright active nucleus, NIRSpec spectra contain quasiperiodic “wiggles” caused by how the instrument samples and reconstructs the light. These are detector-related artifacts, not fluctuations in the galaxy.

The team developed software called hyphen to predict how the wiggle frequency changes with wavelength from the paths that dispersed light follows across the detector. It then fits and subtracts the artifact in individual exposures before combining them. Constraining the correction through detector geometry is intended to reduce the risk of removing real stellar absorption features.

The correction does not make every central spectrum usable. The authors exclude spectra dominated by the active nucleus and others with intermediate or anomalous features. Their innermost stellar measurements therefore remain far outside the event horizon, at approximately 30,000 Schwarzschild radii. The work probes the black hole’s gravitational influence, not horizon-scale physics directly.

Precision depends on the model

Using the TriOS code, the researchers assemble libraries of possible stellar orbits and adjust their weights to reproduce the measured motions and galaxy light distribution. Their models allow M87 to be triaxial, with three unequal principal axes, rather than forcing rotational symmetry. This matters because the galaxy’s observed stellar motions are misaligned with its apparent major axis.

The reported mass is 5.38 billion solar masses, with a statistical uncertainty of 0.13 billion and a separate distance-related uncertainty of 0.22 billion. The statistical range represents a 68% credible interval within the adopted modeling framework. According to the authors, it is two to three times narrower than in their previous Keck-only study.

That precision is not a complete inventory of possible error. The models adopt a particular radial variation in stellar mass-to-light ratio, informed by earlier stellar-population work, and a specified dark-matter density profile. The paper does not jointly fit spatially resolved stellar populations and motions. The authors also find that choices in interpolating the model-fitting results introduce uncertainties approximately half the statistical uncertainty.

An independent constraint, not a final verdict

For comparison with the Event Horizon Telescope, the team reports a mass-to-distance ratio expressed as an angular gravitational radius of 3.16 ± 0.08 microarcseconds. Unlike the mass alone, this quantity avoids the separate uncertainty in M87’s distance. Its central value is lower than the EHT collaboration’s combined multi-epoch estimate, but the reported 90% credible intervals overlap.

The authors propose using their stellar-dynamical result as prior information in future EHT analyses. That could help constrain interpretations of the emitting plasma and black hole spin; this study does not itself establish either.

This article is based on the supplied extracted paper text, the original versioned arXiv landing page for 2611.05486v1 and metadata derived from that page. The work is a preprint; those materials do not establish peer review. The extraction omits images and may distort equations or layout, and the findings have not been independently verified here.