M–Sigma Relation¶

An empirical, approximately power-law correlation between a galaxy's central supermassive-black-hole mass and the stellar velocity dispersion of its spheroidal component, used as a demographic scaling relation whose calibration and interpretation depend on sample, morphology, measurement, and selection.

Version: v1 · 2026-08-30 · History
Domain-specific #: 2320
Origin domain: astrophysics
Subdomain: black hole and galaxy demographics
Aliases: M-sigma relation, M–σ relation, MBH–sigma relation, Black-hole-mass–velocity-dispersion relation

Core Idea¶

The M–Sigma Relation is the empirical finding that the mass M_BH of a galaxy's central supermassive black hole covaries strongly with the stellar velocity dispersion σ of the galaxy's spheroidal component. It is commonly expressed as a linear relation in logarithmic variables,

log10(M_BH / M_sun) = α + β log10(σ / σ_0) + ε,

where σ_0 is a chosen pivot, often 200 km/s, α is the normalization, β is the slope, and ε represents intrinsic scatter plus measurement effects. The original 2000 discoveries found a much tighter relation than earlier black-hole-mass correlations with bulge luminosity, though their fitted slopes differed because of samples, dispersion definitions, measurements, and regression choices.^[1]^[2]

The locked identity is dynamical or calibrated central black-hole mass + consistently defined bulge stellar velocity dispersion + population sample and measurement-error model + log-space regression with intrinsic scatter -> an empirical black-hole–host scaling relation. The relation is not one immutable equation. Published normalizations, slopes, and scatter change with galaxy selection, morphological classification, mass methodology, aperture definition, distance scale, treatment of upper limits, and regression direction. The durable abstraction is the paired observables and their conditional population relationship.

The relation matters because a black hole's sphere of direct gravitational dominance is tiny relative to its host galaxy, yet M_BH correlates with a bulge-scale dynamical property. It supplies a constraint on models of black-hole growth and galaxy assembly and a secondary way to estimate black-hole demographics when direct mass measurements are unavailable. It does not, by itself, identify which mechanism produced the covariance.^[3]

Structural Signature¶

a galaxy population — a defined sample rather than one object establishes the relation;
central massive black holes — masses refer to compact central objects, normally supermassive black holes in galactic nuclei;
a mass measurement or calibration — stellar dynamics, gas dynamics, masers, reverberation mapping, or another method supplies M_BH with uncertainties and assumptions;
a spheroidal host component — the relevant stellar dispersion belongs to a bulge or elliptical body, not indiscriminately to the galaxy's disk;
a dispersion definition — luminosity weighting, aperture, rotation treatment, and radial range specify σ;
logarithmic scaling — a power law in ordinary units becomes approximately linear in log space;
a pivot scale — σ_0 reduces covariance between fitted slope and intercept and makes normalization interpretable;
measurement error in both axes — uncertainty affects mass and dispersion, often heteroscedastically;
intrinsic scatter — galaxies need not lie exactly on a deterministic curve even if measurements were perfect;
a regression convention — forward, inverse, symmetric, or hierarchical fits answer different prediction questions;
a selection function — detectability, spatial resolution, activity, morphology, distance, and publication choices determine which galaxies enter;
a calibrated domain — the observed range and host classes bound defensible interpolation and extrapolation;
population heterogeneity — classical bulges, ellipticals, pseudobulges, barred systems, and active nuclei may occupy different distributions;
residual structure — deviations can covary with bulge mass, effective radius, morphology, or measurement method;
an inferential use — the fit estimates conditional distributions, demographic moments, or theory constraints rather than replacing observation.

Because the equation is usually used logarithmically, “scatter” is commonly quoted in dex. A scatter of s dex is multiplicative in linear mass. Reporting only a central prediction without the appropriate predictive scatter makes the relation appear more precise than it is.

What It Is Not¶

Not a direct black-hole-mass measurement. Substituting a measured σ into a fitted relation produces a scaling estimate conditioned on a calibration.
Not a universal constant of nature. α, β, and scatter are empirical and sample-dependent.
Not a statement about stellar mass dispersion. Sigma is a velocity dispersion, a measure of stellar kinematics.
Not a correlation with the whole disk. The canonical relation uses the spheroidal component; disks and pseudobulges complicate naive transfer.^[3]
Not proof of causal coevolution. Feedback, common dependence on potential depth, merger averaging, selection, and measurement covariance can all contribute.
Not identical to the black-hole–bulge-mass or black-hole–luminosity relations. These are related but distinct projections of host structure.
Not exact for every galaxy. Outliers, upper limits, intrinsic scatter, and distinct populations are scientifically meaningful.
Not safe for unrestricted extrapolation. Calibrations dominated by massive nearby galaxies may not hold for dwarf galaxies, high redshift, or unusual nuclei.
Not made unbiased by a small observed scatter. A selection favoring resolvable spheres of influence can narrow or shift the observed locus.

Scope of Application¶

The relation is used to summarize local black-hole demographics, estimate mass distributions from galaxy surveys, compare active and inactive galaxies, calibrate secondary mass estimators, test galaxy-formation simulations, and constrain feedback or merger models. A survey with spectra but no spatially resolved nuclear dynamics can map σ across many galaxies and use a chosen calibration to infer a statistical black-hole mass function. That use is defensible at the population level when calibration scatter and selection are propagated; it is weaker as a claim about one object's exact mass.

Direct dynamical mass measurements require resolving or otherwise constraining the region where the black hole measurably affects velocities. Stellar-dynamical models depend on orbital structure, mass-to-light ratio, dark matter, and spatial resolution. Gas-dynamical models depend on disk geometry and nongravitational motions. Maser dynamics can be precise but select special nuclear conditions. Reverberation mapping reaches active nuclei but relies on broad-line-region geometry and a virial scale factor. Combining methods expands coverage while introducing cross-calibration structure.

Velocity dispersion is also not measurement-free. A central aperture can be contaminated by the black hole whose mass is being correlated; a large aperture can mix bulge and disk light; rotation can be included or excluded; and barred or triaxial systems complicate interpretation. Studies must state whether σ is central, effective, luminosity-weighted within a fraction of the effective radius, or otherwise defined.

Morphology controls scope. Classical bulges and elliptical galaxies establish much of the tight canonical relation. Pseudobulges and bulgeless galaxies show that black-hole growth does not require one universal bulge history and may display weaker or displaced correlations. High-redshift application must contend with evolving populations, resolution limits, luminosity selection, and different mass estimators.

Clarity¶

M in the name means central black-hole mass, not total galaxy mass. σ means a stellar line-of-sight velocity-dispersion statistic for the relevant spheroid, not gas turbulence and not the standard deviation of black-hole masses. The hyphen or en dash denotes a relation between variables, not subtraction.

A fit of M_BH on σ minimizes or models conditional scatter appropriate to predicting mass from dispersion. The inverse fit answers a different question and inverting its coefficients does not generally recover the forward fit when there is scatter. A symmetric regression may be more appropriate for testing an underlying association. Authors can therefore derive different slopes from the same points without arithmetic error.

The black hole's sphere of influence is roughly r_infl = G M_BH / σ². A sample requiring this angular scale to be resolved selects jointly on mass, dispersion, and distance. That selection can favor overmassive black holes at a given dispersion and distort fitted scaling relations. Resolution is essential for reliable dynamics yet can become a selection variable; treating it as a neutral quality cut is unsafe.^[4]

Manages Complexity¶

Galaxy nuclei encode stellar populations, orbital distributions, dark matter, gas, dust, active accretion, and multiscale assembly. The M–Sigma Relation compresses part of that complexity into one host variable and a conditional mass distribution. It lets theory and survey data communicate through a small set of parameters while retaining intrinsic scatter as a measure of lost information.

That compression is most powerful for comparisons made with the same definitions and selection model. Convolving a galaxy velocity-dispersion function with P(M_BH | σ) can estimate a black-hole mass function, but the high-mass tail is sensitive to slope, scatter, and the abundance of high-σ galaxies. Feeding only the best-fit line through a nonlinear population calculation systematically differs from integrating the distribution.

The relation also makes residuals diagnostically useful. If residual black-hole mass correlates with bulge effective radius or stellar mass, a two-parameter “fundamental plane” may capture additional structure—or correlated errors and sample selection may manufacture the appearance. The abstraction organizes that question without presuming the answer.

Abstract Reasoning¶

If β is near four, a twofold increase in σ corresponds to roughly a sixteenfold central mass prediction before scatter.
If the calibration intercept rises while slope and pivot stay fixed, all predicted masses rise by the same logarithmic amount.
If intrinsic scatter is ignored, uncertainties in individual estimates and the high-mass population tail are understated.
If a sample preferentially includes black holes with resolvable spheres of influence, it is not a random slice of the galaxy population.
If disk light lowers or raises an aperture dispersion differently across morphologies, an apparent morphology dependence can mix physics and measurement.
If pseudobulges follow a broader or offset relation, applying an elliptical/classical-bulge calibration yields biased estimates.
If mass measurement error is correlated with the adopted dispersion, ordinary least squares can distort both slope and scatter.
If the relation evolves with redshift, a local calibration cannot be used at early epochs without an evolution term and selection model.
If two formation models predict the same mean slope but different residual correlations or scatter, the mean relation alone cannot distinguish them.
If an individual galaxy lies off the relation, that is evidence about population variation or measurement—not proof that the relation does not exist.

Knowledge Transfer¶

The exact relation transfers among samples only after harmonizing mass methods, dispersion apertures, morphology, regression target, and selection. Its structural form—an empirical log-linear population scaling with measurement error and intrinsic scatter—transfers widely across astronomy: Tully–Fisher, Faber–Jackson, fundamental-plane, mass–metallicity, and cluster scaling relations share many statistical cautions.

The transferable lesson is that a tight correlation can be simultaneously useful for prediction, informative about joint history, and insufficient for causal identification. The astrophysical identities of central black holes, bulges, and stellar velocity fields keep M–Sigma domain-specific rather than making it a second Correlation prime.

Examples¶

nearby elliptical: spatially resolved stellar dynamics supplies M_BH, and a luminosity-weighted spheroid dispersion places the galaxy in a local calibration;
maser host: precise Keplerian disk dynamics anchors mass but the host's bulge dispersion still requires consistent measurement;
active nucleus: reverberation mass and stellar dispersion test cross-calibration with inactive galaxies;
survey inference: a dispersion function is convolved with the relation and its scatter to estimate a black-hole mass function;
simulation test: modeled galaxies are sampled and measured like observations before comparing slope, normalization, and residuals;
pseudobulge: an object can be a systematic outlier under a classical-bulge calibration without lacking a central black hole;
non-example—direct orbit fit: the dynamical mass estimate itself is not the relation;
failure—point estimate: one predicted mass is reported without intrinsic scatter or calibration domain;
failure—causal shortcut: observed tightness is declared proof of one feedback mechanism.

Structural Tensions¶

predictive compactness vs. population diversity — one equation enables surveys while concealing morphology and history;
resolution quality vs. selection bias — resolving dynamical influence improves individual masses but selects the sample;
tight covariance vs. causal ambiguity — small scatter constrains formation without choosing a unique mechanism;
local calibration vs. cosmic evolution — nearby data are precise while distant populations may differ;
single-variable fit vs. multivariate host structure — sigma is powerful but not necessarily sufficient;
heterogeneous reach vs. method consistency — multiple mass techniques widen coverage and complicate calibration;
forward prediction vs. symmetric explanation — regression direction changes the statistical object;
central estimate vs. intrinsic scatter — a clean line is legible but an incomplete representation of the relation.

Structural–Framed Character¶

The M–Sigma Relation is structural as an empirical population pattern. Observation and dynamical inference determine the paired quantities, and statistical structure determines the calibrated conditional distribution. Apertures, morphology labels, sample thresholds, and regression conventions are framed research choices that must accompany any numerical coefficients.

Structural Core vs. Domain Accent¶

The structural core is paired population measurements + log-space scaling + errors and intrinsic scatter + selection-conditioned regression -> predictive covariance. The domain accent is supermassive-black-hole mass, spheroidal stellar velocity dispersion, dynamical mass measurement, galaxy morphology, and black-hole–galaxy evolution.

Correlation — M_BH and σ systematically covary without the association alone proving causation.
Power Law — the common log-linear representation corresponds to approximate power-law scaling.
Regression — fitted slope and intercept depend on the conditional question and error model.
Uncertainty — measurement errors, intrinsic scatter, and selection govern inference.
Model Calibration — local dynamical measurements anchor secondary population estimates.

The minimal prospective DAG places M–Sigma as a strict subtype of prime:correlation. Every defensible instance is a systematic covariation between variables, while Correlation spans non-astronomical pairs and non-power-law forms.

Relationships to Other Abstractions¶

Current abstraction M–Sigma Relation Domain-specific

Parents (1) — more general patterns this builds on

M–Sigma Relation is a kind of Correlation Prime

M_BH and σ systematically covary without the association alone proving causation.

Hierarchy path (1) — routes to 1 parentless root

M–Sigma Relation → Correlation

Neighborhood in Abstraction Space¶

M–Sigma Relation sits in a sparse region of the domain-specific corpus (95^th percentile for distinctiveness): few abstractions share its structure, so a faithful description tends to retrieve it precisely.

Family — Unclustered & Miscellaneous (1565 abstractions)

Nearest neighbors

Computed from structural-signature embeddings · 2026-09-08

Not to Be Confused With¶

black-hole mass as a direct measurement;
black-hole–bulge mass relation;
black-hole–bulge luminosity relation;
Faber–Jackson relation;
galaxy fundamental plane;
velocity dispersion of gas or a whole disk without definition;
one universal slope and normalization;
a deterministic law with zero scatter;
a proof of AGN feedback;
unrestricted extrapolation to every galaxy type and epoch.

References¶

[1] Laura Ferrarese and David Merritt, “A Fundamental Relation Between Supermassive Black Holes and Their Host Galaxies,” The Astrophysical Journal Letters 539 (2000), L9–L12, https://doi.org/10.1086/312838. registry ↩

[2] Karl Gebhardt et al., “A Relationship Between Nuclear Black Hole Mass and Galaxy Velocity Dispersion,” The Astrophysical Journal Letters 539 (2000), L13–L16, https://doi.org/10.1086/312840. registry ↩

[3] John Kormendy and Luis C. Ho, “Coevolution (Or Not) of Supermassive Black Holes and Host Galaxies,” Annual Review of Astronomy and Astrophysics 51 (2013), 511–653, https://doi.org/10.1146/annurev-astro-082708-101811. registry ↩a ↩b

[4] Francesco Shankar et al., “Selection bias in dynamically measured supermassive black hole samples: its consequences and the quest for the most fundamental relation,” Monthly Notices of the Royal Astronomical Society 460 (2016), 3119–3142, https://doi.org/10.1093/mnras/stw678. registry ↩

[5] Scott Tremaine et al., “The Slope of the Black Hole Mass versus Velocity Dispersion Correlation,” The Astrophysical Journal 574 (2002), 740–753, https://doi.org/10.1086/341002. registry

[6] “M–sigma relation,” Wikipedia, frozen evidence packet, https://en.wikipedia.org/wiki/M%E2%80%93sigma_relation. registry