Astronomy · Active galaxies and central black holes

Images, obscuration and viewing angle

Compare a declared isotropic-axis selection model with real obscuration evidence, then follow interferometric measurements through image reconstruction and emission-scale calibration.

Continue in the approved chapter order. Necessary vocabulary, quantities and assumptions are explained in the lesson.

Back to the chapter

After reading, you can

Calculate a clear-view fraction under an explicit sharp-opening and isotropic-orientation model.
Explain how polarized scattered light can reveal otherwise hidden nuclear emission.
Distinguish measured interferometric data, reconstructed emission, a shadow and a causal event horizon.

A hidden component can leave indirect evidence

Gas and dust can change which nuclear components are directly visible. Absorption and scattering depend on wavelength, material and geometry. Obscuration can occur near the nucleus or in larger host-galaxy structures, so one smooth ring of dust is not a complete model for every active galaxy.

Polarization measurements separate additional properties of received light. In the historical study of NGC 1068, Antonucci and Miller found broad emission-line features in polarized flux and interpreted them through scattering of otherwise hidden nuclear radiation. This provides a target-specific inference rather than a rule that every polarized nucleus has identical structure.

A direct total spectrum can be diluted by stellar light, while scattered light can bring a different view of emitting material into the sightline. The observations and calibration therefore matter alongside the proposed obscuring geometry.

Hickox and Alexander also discuss Type 2 systems that can intrinsically lack a broad-line region. Orientation-based unification explains important relationships, but an observed category cannot always be assigned solely by viewing angle. The following exercise isolates geometry under much stricter assumptions.

Equal angle intervals do not contain equal solid angles

For an original ideal model, assume each nucleus has an unoriented symmetry axis and two opposite, perfectly clear polar openings. Let θ0 = 45° be the stipulated half-opening angle around either pole. Outside those openings the direct view is classified blocked by this rule alone.

Assume isotropic axis directions and a completely orientation-independent selection of objects. Fold the angle i between the observer and the nearer pole into 0° ≤ i ≤ 90°. The normalized probability density is sin i when the integration angle is expressed in radians.

The solid angle of one cone with half-angle θ0 is 2π(1 − cosθ0). Two opposite cones occupy 4π(1 − cosθ0) out of total 4π. Their geometric clear fraction is therefore F(θ0) = 1 − cosθ0.

At 45°, F = 1 − 1/√2 ≈ 0.292893219, with blocked fraction approximately 0.707106781. Taking 45/90 = 1/2 would incorrectly assign equal weight to equal inclination intervals instead of the isotropic solid-angle measure.

A cumulative geometric fraction is not an observed census

More generally, the folded-axis cumulative fraction up to inclination i is F(i) = 1 − cos i on the stated interval. It begins at zero for zero opening, reaches about 0.292893219 at 45° and reaches one at 90°. The original plot shows that cumulative geometry.

The curve is neither a brightness profile nor a distribution measured from an AGN catalogue. It assumes the same sharp double opening for all objects, random axes and no orientation-dependent inclusion. A jet detection probability is a different selection problem.

Clumpy material, host obscuration, luminosity-dependent structure and intrinsic accretion differences can change the observational classification. Directionally enhanced emission can also make a selected sample depart from the assumed orientation distribution.

The model is useful because its assumptions make a reproducible prediction. Comparing it with a real fraction would require consistent class definitions, completeness and alternative explanations. One numerical agreement would not prove that every physical nucleus has the same torus.

Original cumulative solid-angle model F(i) = 1 − cos i for isotropic unoriented axes folded into 0–90°. The 45° opening is stipulated. Values are geometric fractions under complete orientation-independent selection, not observed Type 1/Type 2 proportions.
Folded inclination / half-opening	Cumulative geometric fraction
0°	0
45°	0.292893219
90°	1

Isotropic axes and an opening angle

Scroll sideways to see the full diagram. Its explanation is below.

Original idealized geometric probability model, not observed active-galaxy counts or classification data. Axis directions are isotropic and the two opposite directions of one unoriented axis are identified. Fold inclination i to 0 through 90 degrees, where 0 degrees means looking along either direction of the axis and 90 degrees means a perpendicular line of sight. On this folded interval the normalized inclination density is sin(i) when i is in radians, since its integral from 0 to pi/2 is one. The violet solid curve is the cumulative probability F(i) = 1 - cos(i), with degrees converted to radians before evaluation. Both axes are linear and explicitly start at zero: inclination from 0 to 90 degrees, cumulative geometric fraction from 0 to 1. The cumulative fraction is the fraction of axes with folded inclination at or below the horizontal-axis value, not the density per degree. The endpoints are F(0 degrees) = 0 and F(90 degrees) = 1. Dashed orange guides and a violet circle mark a stipulated sharp double-polar opening with half-angle 45 degrees: F(45 degrees) = 1 - cos(45 degrees) = 0.2928932188, with complementary geometric fraction 0.7071067812. This is the two polar caps divided by the full sphere, with axis folding counted once; 45 degrees therefore does not represent one half of isotropic axes. The complete table also gives complementary fractions 1, 0.7071067812 and 0 at inclinations 0, 45 and 90 degrees. Using this geometric opening fraction as a selected-sample fraction additionally requires complete selection independent of orientation. It is not a measured Type 1 or Type 2 fraction, an actual smooth torus model, a jet probability, or a universal opening boundary. Clumpy material, host obscuration, intrinsic source differences and orientation-dependent selection are outside the declared geometry.

Reconstruction connects sparse measurements to structure

Very long baseline interferometry combines signals recorded by separated radio telescopes. A baseline is the projected separation between two stations for a target direction. The measured correlations, often described as complex visibilities, constrain spatial patterns in the source brightness.

An array samples some of these spatial measurements; it is not a uniformly filled telescope surface as large as the largest station separation. Earth rotation changes the projected baselines, and the target’s visibility determines which station pairs can contribute.

Calibration, noise and incomplete sampling enter the reconstruction problem. Image methods and forward models propose brightness structures whose predicted measurements can be compared with the data. Tests using alternative methods, synthetic inputs and changed data selections help assess which structures are robust.

The Event Horizon Telescope’s M87 imaging paper describes that process and its validation. Its historical observing campaign is a specific data set, not a claim about present array capabilities. An educational ring drawing is not a replacement for the measured correlations or a newly observed image.

Calibrate the physical interpretation of the image

A reconstructed emission ring describes a brightness feature. A black-hole shadow describes a dark apparent region associated with photon propagation and capture in a stated spacetime and illuminating environment. The event horizon is the causal boundary itself, rather than a bright material surface.

Gravitational light bending and the location, motion and opacity of emitting plasma connect those concepts. Their sizes are not automatically identical. The EHT M87 mass analysis explicitly calibrates an observed emission diameter against model emission, rather than simply declaring it equal to a photon-ring diameter.

An inferred angular gravitational scale has the form θ_g = GM/(Dc²), where θ_g is in radians, M is mass, D is the adopted distance, G is the gravitational constant and c is light speed. Converting that scale to mass needs both the emission-model calibration and distance. The equation alone supplies neither measurement.

A careful account therefore states what the data constrain, what reconstruction features persist, and how an emission model connects them to a shadow and physical scale. The next lesson relates nuclear activity to the host galaxy, where association and transferred energy must likewise retain their selection and causal limits.

Worked examples

Compute the two-opening fraction

Why does a 45° half-opening give about 29.3 percent clear views rather than 50 percent?

One cone occupies solid angle 2π(1 − cos 45°); two opposite cones double it.
Divide by 4π for the full sphere to obtain 1 − cos 45°.
The result is 1 − 1/√2 ≈ 0.292893219 under isotropic directions and complete selection.

Isotropic directions are weighted by solid angle, not uniformly by the folded inclination number.

Keep the measured feature and the physical scale separate

What additional steps connect a reconstructed emission-ring diameter to a mass?

Test the reconstructed brightness feature against the calibrated interferometric data and alternative methods.
Use a stated emission and relativistic propagation model to infer an angular gravitational scale.
Combine that scale with an adopted distance through M = θ_g Dc²/G, retaining calibration and distance uncertainty.

Reconstruction, emission-model calibration and distance are required; the bright diameter is not simply the event-horizon diameter.

Try the reasoning

For isotropic folded axes, are equal inclination intervals automatically equally probable?
Show the answer and reasoning
No.
Their weight follows sin i in radians, or the corresponding solid-angle measure.
Is the supplied 0.292893219 an observed universal Type 1 fraction?
Show the answer and reasoning
No.
It belongs to a stipulated geometric model with complete orientation-independent selection.
Does a largest baseline mean a uniformly filled collecting surface of that diameter?
Show the answer and reasoning
No.
The array samples correlations on particular projected baselines and needs a reconstruction model.
Can emission-ring, photon-ring, shadow and event-horizon diameters be identified without a physical model?
Show the answer and reasoning
No.
Radiation, propagation, geometry and causal boundaries describe different quantities.

What to carry forward

Obscuration and scattered light can reveal different views of a nucleus.
The clear fraction of a sharp-opening model depends on solid-angle and selection assumptions.
Interferometric images and black-hole physical scales are connected through calibration, reconstruction and emission models.

References

Teaching resources and research supporting this lesson’s explanations and examples.