Reading Routes and Ranges of Validity for the Core Relations

Several relations that recur throughout the main text are presented side by side here. Each relation is annotated with its physical assumptions, the place where it is first used, and the points in later text where it is easily misused; for the full formulas, see Appendix Index of Common Formulas.

The Siegert Relation

The physical picture of the Siegert relation is established in Chapter A Guided Tour: What Does a Telescope Actually Record, and Where Is This Book Taking You?, its quantum-optical background is developed in Chapter The Coherence Functions g^{(1)}, g^{(2)} and the Siegert Relation, and spatial intensity interferometry uses it in Chapter Spatial Coherence, van Cittert–Zernike, and Intensity Interferometry. When reading these chapters, work through the following steps:

1. First confirm whether the light field can be approximated as thermal light, chaotic light or a Gaussian random field. This condition allows fourth-order moments to be decomposed into combinations of second-order moments.

2. Then normalize the first-order coherence function to obtain \(g^{(1)}\) or the spatial degree of coherence \(\gamma_{12}\).

3. In ideal single-mode thermal light, the zero-delay second-order correlation excess is given by \(|g^{(1)}|^2\).

4. Once real instruments enter, write finite bandwidth, finite time response, polarization averaging, spatial modes and background as contrast factors.

Check item	How to use it when satisfied	How it goes wrong when not satisfied
Thermal-light or Gaussian-field approximation	The second-order correlation can be connected to the squared modulus of the first-order coherence.	Stimulated emission, a few emitters or a strongly non-stationary source may change \(g^{(2)}\).
Single mode or known number of modes	The bunching-peak contrast can be interpreted as physical coherence information.	Multimode averaging suppresses the peak and it is misread as an incoherent source.
Known time response	The theoretical peak can be connected to the observed peak.	ns-scale electronics dilute the fs-scale optical coherence peak to \(10^{-6}\)–\(10^{-5}\).
Background already separated	The correlation excess can be turned into the target’s $	V

When writing \(g^{(2)}-1\) later, give the coherence time, correlation bin, effective number of modes, background flux fraction and null test at the same time. A single zero-delay peak height alone is not enough to establish that it is astrophysical bunching.

The van Cittert–Zernike Theorem

The VCZ theorem links the sky brightness of a far-field incoherent source to the first-order spatial coherence. Chapter A Guided Tour: What Does a Telescope Actually Record, and Where Is This Book Taking You? gives the definition of complex visibility, Chapter Spatial Coherence, van Cittert–Zernike, and Intensity Interferometry applies it to intensity interferometry, and Chapter Observation Design, Error Budget, and Feasibility puts it into observation design. When using this relation, follow the path below:

1. Write the celestial surface brightness as a function \(I(\boldsymbol\theta)\) of angular coordinates.

2. Divide the projected baseline of the two telescopes by the wavelength to obtain the spatial frequency \(\boldsymbol u=\boldsymbol B_\perp/\lambda\).

3. Take the normalized Fourier transform of the brightness to obtain the complex visibility \(V(\boldsymbol u)\).

4. Intensity interferometry can give only \(|V|^2\) directly, so phase, mirror images and asymmetric structure require additional information.

Condition	Usage in the main text	Typical sources of breakdown
Far-field approximation	The angular structure of stars, binaries, disks and transients can be described in the \(u,v\) plane.	Near-field experiments, extreme space-baseline geometry or unmodeled wavefront curvature.
Narrowband or already-handled bandwidth	Each channel can use a single-\(\lambda\) approximation.	Broadband averaging washes out high-frequency visibility structure.
Source stable within the integration	Each epoch has one visibility model.	Outbursts, rotation, pulsation and rapid structural change mix different models.
Nonnegative brightness with a reasonable model	Phase retrieval can be constrained with priors.	$

The Atmospheric-Phase Immunity of Intensity Interferometry

Intensity interferometry is insensitive to atmospheric piston phase because it measures intensity-fluctuation correlations and does not rely on the direct coherent superposition of the two light fields. This advantage has clear boundaries:

• Transparency variations remain. Transparency changes the photon rate, and also the accidental coincidences and background normalization.

• Scintillation remains. Scintillation introduces intensity correlations on slower time scales, requiring time-shift and block checks.

• Electronic crosstalk remains. Crosstalk can produce a zero-delay structure narrower and higher than the astrophysical peak.

• Background dilution remains. Even if uncorrelated, the background can suppress the target signal through the flux fraction.

When invoking the “atmospheric-phase immunity” advantage, also report how non-phase errors are monitored: use at least a few of calibrator stars, off-source, off-band, wrong-polarization, dark field, time-shift and injection-recovery.

The Rayleigh Limit and Fisher Information

The Rayleigh criterion is an empirical imaging scale, not the information limit of every parameter-estimation problem. Chapter Quantum Estimation, the Rayleigh Limit, and SPADE Sub-Rayleigh Resolution shows the difference between direct imaging and mode measurement, and Chapter Teaching Experiments and Computational Experiments turns it into a computational experiment. When using this set of results, first check these model ingredients:

1. Specify the parameter: is it the two-point-source separation, the centroid, the flux ratio, the angular diameter, or a more complex image.

2. Specify the data: pixel intensities, mode counts, event times, or \(|V|^2\).

3. Write out the likelihood or Fisher information; do not just cite the word “super-resolution.”

4. Include background, finite photon number, mode mismatch, aberrations and centroid error.

The conclusions of SPADE apply most clearly to the case of a weak, equally bright, incoherent two-point source with a known Gaussian PSF. For a galaxy, an accretion disk, a strong-lensing arc or a source with coherent components, the source model and the measurement basis must be rewritten.

The Error Budget and Roadmap Relations

The error budget in observation design is in Chapter Observation Design, Error Budget, and Feasibility, and the readiness and milestones in the roadmap are in Chapter From White Paper to Research Plan. The error budget answers whether the target quantity can be measured; the roadmap answers at which step the project is ready to commit to the next stage.

Error term	Example	Mitigation to be recorded
Statistical error	Shot noise varying with photon number, integration time and bandwidth.	Increase area, integration time, channel accumulation or target brightness.
Calibration error	Zero-baseline contrast, time response, polarization efficiency and filter shape.	Calibrator stars, laboratory response measurements and nightly calibration.
Background error	Night-sky brightness, companion stars, continuum, dark counts.	Off-source, line/continuum separation, background windows and flux-fraction estimates.
Model error	Limb darkening, non-spherical symmetry, velocity model, radiative transfer.	Multi-model comparison, posterior predictive checks and external observational constraints.
Selection error	Triggering, weather, target sample and a posteriori selection windows.	Preregistered target criteria, global significance and failure criteria.