Part VII: Observational Techniques
Observational astrophysics transforms the faint whispers of electromagnetic radiation, gravitational waves, neutrinos, and cosmic rays into quantitative measurements that constrain physical models of the universe. Mastering telescope design, detector physics, spectroscopic analysis, interferometric techniques, and adaptive optics is essential for extracting the maximum scientific return from every photon collected. This part provides a graduate-level treatment of the instruments and methods that underpin modern astrophysical research.
1. Telescopes & Optics
Astronomical telescopes are photon-collecting machines whose performance is governed by fundamental diffraction theory, geometric aberrations, and the practical constraints of mirror fabrication and mechanical support. The primary figure of merit is the combination of collecting area, angular resolution, and field of view.
1.1 Refractors vs. Reflectors
Refracting telescopes use lenses to bring light to a focus, while reflecting telescopes use curved mirrors. Refractors dominated early astronomy but suffer from several fundamental limitations:
Chromatic Aberration
The refractive index of glass varies with wavelength (dispersion), so different colors focus at different points along the optical axis. This produces colored halos around point sources.
The focal length varies as $f(\lambda) \propto 1/(n(\lambda) - 1)$, and the longitudinal chromatic aberration scales as $\Delta f / f \approx 1/V$, where $V$ is the Abbe number of the glass. Achromatic doublets using crown and flint glass reduce but cannot eliminate this aberration.
Spherical Aberration
Spherical surfaces do not bring parallel rays to a perfect focus. Marginal rays focus closer to the lens/mirror than paraxial rays. The transverse aberration scales as $\propto h^3/f^2$, where $h$ is the ray height and $f$ the focal length. For mirrors, a parabolic surface eliminates spherical aberration for on-axis rays.
Coma
Off-axis rays from a parabolic mirror suffer from coma, producing comet-shaped images. The angular extent of coma scales as $\theta_{\text{coma}} \propto \theta / (f/D)^2$, where $\theta$ is the field angle and $f/D$ is the focal ratio. Coma limits the useful field of view of Newtonian telescopes.
Astigmatism & Field Curvature
Further off-axis, the focal surface curves (Petzval curvature) and tangential and sagittal foci separate (astigmatism). Modern wide-field correctors employ multiple aspheric elements to flatten the field and minimize astigmatism across degree-scale fields.
Reflectors avoid chromatic aberration entirely, can be made arbitrarily large (since the mirror is supported from behind), and can be given aspheric figures. All modern research telescopes above ~1 meter aperture are reflectors.
1.2 Primary Mirror Designs
The choice of mirror conic section determines the aberration properties. A conic of revolution is described by the equation $z = \frac{r^2/R}{1 + \sqrt{1 - (1+K)r^2/R^2}}$, where $R$ is the radius of curvature and $K$ is the conic constant:
| Surface | Conic K | Properties |
|---|---|---|
| Sphere | K = 0 | Easy to fabricate; large spherical aberration |
| Paraboloid | K = −1 | Perfect on-axis focus; coma off-axis (Newtonian) |
| Hyperboloid | K < −1 | Used in Ritchey-Chrétien and Cassegrain designs |
| Ellipsoid | −1 < K < 0 | Gregorian telescopes |
Ritchey-Chrétien (RC) design: Both the primary and secondary mirrors are hyperboloids. This two-mirror aplanatic system eliminates both spherical aberration and coma, providing a much wider usable field of view than classical Cassegrain designs. The HST (2.4 m), Keck (10 m), VLT (8.2 m), and the upcoming ELT (39 m) all employ RC or modified RC optical prescriptions.
The effective focal length of a two-mirror system is $f_{\text{eff}} = f_1 M$, where $f_1$ is the primary focal length and $M = f_2/(f_2 - d)$ is the magnification of the secondary, with $d$ the mirror separation.
1.3 Diffraction Limit & Angular Resolution
A circular aperture of diameter $D$ produces an Airy diffraction pattern. The angular radius of the first dark ring defines the Rayleigh criterion:
In practical units, $\theta \approx 0.252\,\text{arcsec}\,(\lambda/1\,\mu\text{m})\,(1\,\text{m}/D)$. The Airy pattern intensity distribution is given by:
where $J_1$ is the Bessel function of the first kind. The central Airy disk contains approximately 84% of the total encircled energy. The first ring contains about 7%, the second about 3%.
Light-Gathering Power
The photon collection rate scales as $\propto D^2$. A 10 m telescope collects$(10/2.4)^2 \approx 17\times$ more light than the HST. The limiting magnitude for detection scales as $m_{\text{lim}} \propto 5\log_{10}(D)$.
Resolving Power
At $\lambda = 550$ nm, the diffraction limit is $\theta \approx 14$ mas for a 10 m telescope, but atmospheric seeing typically limits ground-based resolution to $\sim 0.5$–$2''$ without adaptive optics.
1.4 Current & Future Facilities
| Telescope | Aperture | Type | Key Features |
|---|---|---|---|
| HST | 2.4 m | RC, space | Diffraction-limited UV/optical/NIR; no atmosphere |
| JWST | 6.5 m | TMA, space | NIR/MIR (0.6–28 μm); L2 orbit; 18 hex segments |
| Keck I/II | 10 m | Segmented RC | 36 hex segments; LGS-AO; interferometric baseline |
| VLT (4×UT) | 8.2 m | Active optics | VLTI interferometry; MUSE IFU; SPHERE coronagraph |
| Subaru | 8.2 m | RC | Hyper Suprime-Cam (1.77 deg² FOV); prime focus |
| ELT | 39 m | 5-mirror, segmented | 798 segments; first light ~2028; MCAO |
| TMT | 30 m | RC, segmented | 492 segments; NFIRAOS MCAO |
| GMT | 25.4 m | 7 monolithic | Seven 8.4 m mirrors; adaptive secondary |
Interactive Simulation: Telescope Resolution & Airy Pattern
Explore how telescope diameter, wavelength, and binary star separation affect angular resolution. The simulation computes the Airy diffraction pattern, displays a 2D PSF, and evaluates whether a binary star system is resolved according to the Rayleigh criterion.
Telescope Resolution & Airy Pattern
PythonCompute and visualize the Airy diffraction pattern for circular apertures of various telescope diameters.
Click Run to execute the Python code
Code will be executed with Python 3 on the server
2. Detectors & Photometry
Charge-coupled devices (CCDs) and their successors (CMOS detectors, infrared arrays) are the workhorses of modern astronomical imaging and spectroscopy. Understanding detector noise properties is essential for optimizing observations and correctly estimating measurement uncertainties.
2.1 CCD Physics
A CCD is a silicon-based semiconductor detector in which incident photons generate electron-hole pairs via the photoelectric effect. The key performance parameters are:
Quantum Efficiency (QE)
The fraction of incident photons that produce a detected photoelectron. Modern back-illuminated CCDs achieve QE $> 90\%$ at peak wavelength (~600–700 nm). Deep-depletion CCDs extend red sensitivity to $\sim 1\,\mu$m. The QE curve determines the effective bandpass of the detector.
Read Noise
Noise introduced during charge transfer and amplification, typically $\sigma_{\text{read}} \sim 2$–$10\,e^-$/pixel for scientific CCDs, and $< 1\,e^-$ for electron-multiplying CCDs (EMCCDs). Read noise is independent of exposure time and adds in quadrature per pixel per readout.
Dark Current
Thermal generation of electron-hole pairs, typically $\sim 0.001$–$0.01\,e^-$/pixel/s at operating temperatures of $-100^\circ$C to $-120^\circ$C. The dark current approximately halves for every 7°C decrease in temperature:$I_{\text{dark}} \propto T^{3/2} \exp(-E_g/2k_BT)$.
Gain & Full Well
The gain $g$ (e$^-$/ADU) converts between detected photoelectrons and digital counts. Full well capacity (typically $\sim 10^5$ e$^-$) sets the saturation limit. Dynamic range $= \text{FWC}/\sigma_{\text{read}}$, typically $\sim 10^4$–$10^5$.
2.2 Signal-to-Noise Ratio (SNR)
The fundamental equation governing the quality of any astronomical measurement is the CCD equation:
where $N_* = F_* \cdot A \cdot \text{QE} \cdot t$ is the total source count,$n_{\text{pix}}$ is the number of pixels in the extraction aperture,$N_{\text{sky}}$ is the sky background per pixel,$N_{\text{dark}}$ is the dark current per pixel, and$N_{\text{read}}$ is the read noise per pixel.
Source-Limited Regime
When $N_* \gg n_{\text{pix}}(N_{\text{sky}} + N_{\text{dark}} + N_{\text{read}}^2)$:$\text{SNR} \approx \sqrt{N_*} \propto \sqrt{t}$. Bright source photon noise dominates.
Background-Limited
When sky dominates: $\text{SNR} \approx N_* / \sqrt{n_{\text{pix}} N_{\text{sky}}} \propto \sqrt{t}$. This is the typical regime for faint-object photometry. Darker sites yield higher SNR.
Read-Noise-Limited
When $N_{\text{read}}^2$ dominates: $\text{SNR} \propto t$. This regime applies to very short exposures or very faint sources. Multiple shorter exposures accumulate read noise faster than a single long exposure.
2.3 Photometric Techniques
Photometry is the measurement of flux from astronomical sources. Two principal approaches are used:
Aperture Photometry
Sum all pixel values within a circular aperture of radius $r$, subtract the sky estimated from an annulus at larger radius. The optimal aperture radius balances including more source flux against adding more sky noise. Typically $r \approx 1.5$–$3 \times \text{FWHM}$of the PSF. Growth curves (flux vs. aperture radius) quantify the encircled energy fraction.
PSF Photometry
Fit a model point-spread function (PSF) to each source, allowing simultaneous fitting in crowded fields where apertures would overlap. DAOPHOT (Stetson 1987) pioneered this technique. The PSF model can be empirical (derived from bright isolated stars) or analytic (Gaussian, Moffat profile). PSF photometry is essential for globular clusters, galactic bulge fields, and other crowded environments.
2.4 Photometric Systems
The Johnson-Cousins UBVRI system defines broadband filters centered at approximately 365, 440, 550, 640, and 790 nm. The magnitude system is defined as:
The Sloan ugriz system uses five non-overlapping filters optimized for CCD response. The AB magnitude system is defined such that $m_{\text{AB}} = -2.5\log_{10}(F_\nu) - 48.60$, where $F_\nu$is in erg s$^{-1}$ cm$^{-2}$ Hz$^{-1}$. A source with$F_\nu = 3631$ Jy has $m_{\text{AB}} = 0$ in all bands.
Color indices such as $B - V$ measure the slope of the spectral energy distribution. For blackbodies, color correlates with temperature: hotter stars have more negative $B - V$. Bolometric corrections $BC = m_{\text{bol}} - m_V$ account for flux outside the observed bandpass and depend on spectral type.
2.5 Atmospheric Extinction
Earth's atmosphere absorbs and scatters light, dimming sources by an amount that depends on zenith angle. The observed magnitude is related to the above-atmosphere magnitude by:
where $k$ is the extinction coefficient (magnitudes per airmass) and $X$ is the airmass. For a plane-parallel atmosphere, $X = \sec z$ where $z$ is the zenith angle. More accurately, $X \approx \sec z - 0.0018167(\sec z - 1) - 0.002875(\sec z - 1)^2 - 0.0008083(\sec z - 1)^3$(Kasten & Young 1989). Typical extinction coefficients at excellent sites:
| Band | U | B | V | R | I |
|---|---|---|---|---|---|
| k (mag/airmass) | 0.50 | 0.25 | 0.15 | 0.10 | 0.07 |
Interactive Simulation: CCD Signal-to-Noise Calculator
Compute the SNR as a function of exposure time for a source of given magnitude, accounting for sky background, read noise, and dark current. The simulation also shows the noise budget breakdown and the limiting magnitude for a fixed exposure.
CCD Signal-to-Noise Calculator
PythonCompute SNR vs exposure time, noise budget, and limiting magnitude for a CCD observation.
Click Run to execute the Python code
Code will be executed with Python 3 on the server
3. Spectroscopy
Spectroscopy — decomposing light into its constituent wavelengths — is the most powerful tool in astrophysics. It reveals chemical composition, temperature, density, magnetic fields, kinematics, and redshift. The design of astronomical spectrographs involves fundamental tradeoffs between spectral resolution, wavelength coverage, and throughput.
3.1 Dispersive Elements: Prisms & Gratings
A diffraction grating with groove spacing $d$obeys the grating equation:
where $m$ is the diffraction order, $\alpha$ is the angle of incidence, and $\beta$ is the diffraction angle. The angular dispersion is $d\beta/d\lambda = m/(d\cos\beta)$. Higher orders give greater dispersion but suffer from order overlap, requiring order-sorting filters or cross-dispersion.
Prisms disperse via Snell's law and have angular dispersion $d\theta/d\lambda = (t/d)(dn/d\lambda)$, where $t$ is the base thickness. Prisms produce non-linear dispersion (higher in the blue) and no order overlap, but lower resolving power than gratings. Grisms (grating + prism) combine advantages of both.
3.2 Spectral Resolving Power
The resolving power of a spectrograph quantifies its ability to separate closely spaced wavelengths:
where $N$ is the total number of illuminated grooves and $m$ is the order. The corresponding velocity resolution is $\Delta v = c/R$. For example:
| Resolution Class | R | Δv | Applications |
|---|---|---|---|
| Low | 100–1,000 | 300–3000 km/s | Redshift surveys, SED fitting, faint objects |
| Medium | 1,000–10,000 | 30–300 km/s | Stellar classification, galaxy kinematics |
| High | 10,000–100,000 | 3–30 km/s | Stellar abundances, exoplanet RV |
| Very High | >100,000 | <3 km/s | Asteroseismology, ISM structure |
3.3 Echelle Spectrographs
Echelle gratings operate at high order ($m \sim 30$–$200$) with a coarse groove spacing ($d \sim 30$–$300\,\mu$m), giving very high resolving power. The blaze angle is steep ($\theta_B \sim 63^\circ$–$76^\circ$). A cross-disperser (prism or low-order grating) separates the overlapping orders onto a 2D detector, producing an “echellogram” with each order as a horizontal stripe.
Examples: HARPS (R = 115,000 at ESO 3.6 m), ESPRESSO (R up to 190,000 at VLT), HIRES (Keck), and the planned ANDES for the ELT. These instruments achieve radial velocity precisions of ~0.3–1 m/s, enabling detection of Earth-mass exoplanets via the radial velocity method.
3.4 Integral Field Spectroscopy (IFU)
An Integral Field Unit obtains a spectrum at every spatial position within a 2D field, producing a 3D datacube with axes (x, y, λ). Three main technologies exist:
Lenslet Arrays
Microlens arrays sample the focal plane and feed individual spectra to the detector. Used in OSIRIS (Keck), SINFONI (VLT).
Fiber Bundles
Optical fibers rearrange a 2D field into a pseudo-slit. Examples: VIRUS (HET), MaNGA (SDSS-IV) with 127–1423 fibers per IFU.
Image Slicers
A stack of thin mirrors slices the field into strips, rearranged into a pseudo-slit. Used in MUSE (VLT), NIRSpec (JWST). MUSE covers 1′×1′ with 90,000 spectra.
3.5 Doppler Measurements & Radial Velocities
The radial velocity of a source is measured from the Doppler shift of spectral lines:
The precision of radial velocity measurements depends on spectral resolution, SNR, the number and sharpness of spectral lines, and wavelength calibration stability. The photon-limited RV precision scales as $\sigma_{v} \propto 1/(R \cdot \text{SNR} \cdot \sqrt{N_{\text{lines}}})$. Modern techniques use simultaneous calibration with thorium-argon lamps, iodine cells, laser frequency combs, or Fabry-Pérot etalons to achieve sub-m/s precision.
3.6 Spectral Line Profiles & Broadening
The observed profile of a spectral line is the convolution of several broadening mechanisms:
Natural broadening: The intrinsic Lorentzian width from the finite lifetime of the excited state, via the uncertainty principle:$\Delta\nu_{\text{nat}} = \gamma/(2\pi)$, where $\gamma = 1/\tau$is the damping constant. Typically $\Delta\lambda \sim 10^{-4}$ Å.
Thermal (Doppler) broadening: The Gaussian line width from thermal motions of atoms: $\Delta\lambda_D = (\lambda/c)\sqrt{2k_BT/m_{\text{atom}}}$. For hydrogen at $T = 10^4$ K, $\Delta\lambda_D \approx 0.4$ Å at Hα.
Pressure (collisional) broadening: Perturbation of energy levels by nearby particles. Linear Stark effect (hydrogen in electric fields), quadratic Stark effect (non-hydrogenic atoms), van der Waals broadening. Produces Lorentzian wings. The combined thermal + pressure profile is a Voigt profile: $V(\nu) = \int_{-\infty}^{\infty} G(\nu')\,L(\nu - \nu')\,d\nu'$.
Rotational broadening: For a rotating star, different parts of the disk have different line-of-sight velocities. The broadening kernel depends on $v \sin i$ and limb darkening. The line profile develops a characteristic “dish” shape with width $\Delta\lambda = 2\lambda v\sin i / c$.
3.7 Curve of Growth
The curve of growth relates the equivalent width $W_\lambda$ of an absorption line to the column density $N$ of the absorbing species. It has three distinct regimes:
Linear Part
For weak lines ($\tau \ll 1$): $W_\lambda \propto N f \lambda^2$, where $f$ is the oscillator strength. The line is optically thin and$W_\lambda$ grows linearly with column density.
Flat (Saturated) Part
The line core saturates ($\tau \gg 1$) and further absorption only broadens the Doppler wings: $W_\lambda \propto \sqrt{\ln(N f \lambda^2)}$. The curve flattens and $W_\lambda$ is insensitive to $N$.
Square-Root (Damped) Part
At very high column density, damping wings (Lorentzian) dominate:$W_\lambda \propto \sqrt{N \gamma}$. This regime is seen in damped Lyman-$\alpha$ systems with $N_{\text{HI}} > 2 \times 10^{20}$ cm$^{-2}$.
4. Interferometry
Interferometry combines signals from two or more telescopes separated by a baseline to achieve angular resolution far exceeding that of any single aperture. It is the foundation of radio astronomy imaging and is increasingly applied at optical and infrared wavelengths.
4.1 Two-Element Interferometry
Consider two apertures separated by baseline $B$. For a monochromatic point source, the combined signal produces sinusoidal fringes with angular period $\lambda/B$. For an extended source with intensity distribution $I(\hat{s})$, the measured complex visibility is:
This is the van Cittert–Zernike theorem: the complex visibility is the normalized Fourier transform of the source brightness distribution. The fringe visibility amplitude is:
A point source gives $V = 1$; a resolved uniform disk of angular diameter $\theta_{\text{UD}}$ gives $V = |2J_1(\pi B \theta_{\text{UD}}/\lambda)/(\pi B \theta_{\text{UD}}/\lambda)|$, which first reaches zero when $\theta_{\text{UD}} = 1.22\lambda/B$.
4.2 Angular Resolution & Baselines
The angular resolution of an interferometer is set by the longest baseline:
The factor of 2 (vs. 1.22 for a filled aperture) reflects the fact that an interferometer samples spatial frequencies up to $B_{\max}/\lambda$, while a filled aperture of diameter $D$ samples up to $D/\lambda$. In practice, limited baseline coverage (sparse sampling of the uv-plane) degrades the synthesized beam shape.
4.3 Aperture Synthesis & the uv-Plane
Each pair of telescopes measures the visibility at a single point in the Fourier plane (the uv-plane), with coordinates$(u, v) = (B_x/\lambda, B_y/\lambda)$. As Earth rotates, each baseline traces an ellipse in the uv-plane, gradually filling in Fourier coverage. An array of $N$ telescopes provides $N(N-1)/2$ simultaneous baselines.
The dirty image is obtained by inverse Fourier transforming the sampled visibilities. The dirty beam(PSF) is the Fourier transform of the sampling function. Deconvolution algorithms such as CLEAN (Högbom 1974) iteratively remove the dirty beam pattern to recover the true sky brightness distribution.
4.4 Interferometric Facilities
| Facility | Wavelength | Max Baseline | Key Achievements |
|---|---|---|---|
| VLA | Radio (1–50 GHz) | 36 km | 27 antennas; A–D configurations; continuum & line imaging |
| ALMA | mm/sub-mm (84–950 GHz) | 16 km | 66 antennas; protoplanetary disk imaging; high-z galaxies |
| VLBI (EHT) | mm (230 GHz) | ~Earth diameter | First black hole shadow images (M87*, Sgr A*) |
| VLTI | NIR/MIR | 130 m (UT), 200 m (AT) | GRAVITY: Sgr A* orbital dynamics, exoplanet detection |
| CHARA | Optical/NIR | 331 m | Stellar surface imaging; angular diameters of nearby stars |
5. Adaptive Optics
Earth's atmosphere introduces rapidly varying wavefront distortions that blur astronomical images far beyond the diffraction limit. Adaptive optics (AO) systems measure and correct these distortions in real time, recovering near-diffraction-limited performance from ground-based telescopes.
5.1 Atmospheric Turbulence & Seeing
Turbulent mixing of air layers at different temperatures creates random variations in the refractive index. Kolmogorov turbulence theory predicts the atmospheric phase structure function:
The Fried parameter $r_0$ characterizes the coherence length of the atmosphere. It is the diameter of a circular aperture over which the wavefront RMS is approximately 1 radian. Seeing-limited resolution is $\theta_{\text{seeing}} \approx 0.98\lambda/r_0$. At a good site,$r_0 \sim 10$–20 cm at 500 nm.
Wavelength Dependence
The Fried parameter scales as $r_0 \propto \lambda^{6/5}$, so seeing improves dramatically at longer wavelengths. At $\lambda = 2.2\,\mu$m (K-band),$r_0$ is $\sim 4\times$ larger than at V-band, making AO correction much easier in the infrared.
Coherence Time & Isoplanatic Angle
The atmospheric coherence time $\tau_0 = 0.314 r_0/\bar{v}$ (where $\bar{v}$is the effective wind speed) sets the AO update rate: typically $\tau_0 \sim 3$–10 ms. The isoplanatic angle $\theta_0 \sim 0.314 r_0/\bar{h}$ (where $\bar{h}$is the effective turbulence height) limits the corrected field of view to$\sim 10''$–$30''$ at V-band.
5.2 Wavefront Sensing & Correction
An AO system consists of three key components: a wavefront sensor (WFS), a deformable mirror (DM), and a real-time control computer.
Shack-Hartmann Wavefront Sensor
A lenslet array divides the pupil into subapertures. Each lenslet forms an image of the guide star. Local wavefront tilts displace each spot from its nominal position. The spot displacements are proportional to the local wavefront gradient $\nabla\phi$. The wavefront is reconstructed by integrating the measured slopes. Typically$\sim (D/r_0)^2$ subapertures are needed.
Deformable Mirrors
A thin mirror membrane is deformed by an array of actuators to apply the conjugate of the measured wavefront error. Modern DMs have $\sim 1000$–$10000$actuators. The number of actuators must match the number of correctable modes, approximately $(D/r_0)^2$. For ELT-class telescopes at optical wavelengths, $\sim 10^4$–$10^5$ actuators are required.
Tip-Tilt Correction
The lowest-order atmospheric mode is image motion (tip-tilt), which accounts for$\sim 87\%$ of the total wavefront variance for $D \gg r_0$. A separate fast steering mirror corrects tip-tilt at rates up to several kHz.
5.3 Laser Guide Stars
Natural guide stars (NGS) bright enough for wavefront sensing are sparse on the sky, limiting sky coverage. Laser guide stars (LGS) create an artificial reference by exciting sodium atoms in the mesosphere at ~90 km altitude (sodium LGS, $\lambda = 589$ nm) or by Rayleigh backscattering at ~15–20 km.
Limitations of LGS systems include: (1) the cone effect — the LGS probes a cone rather than a cylinder of atmosphere, causing focal anisoplanatism; (2) tip-tilt indetermination — the outgoing and return LGS paths traverse the same atmosphere, so overall tip-tilt cannot be measured from the LGS alone (a faint NGS is still needed for tip-tilt); (3) spot elongation for off-axis LGS on large telescopes.
Multi-conjugate AO (MCAO) uses multiple guide stars and deformable mirrors conjugated to different altitudes, extending the corrected field to ~1′–2′. Facilities like GeMS (Gemini) and MAORY/MORFEO (ELT) implement MCAO with multiple LGS.
5.4 Strehl Ratio & Performance Metrics
The Strehl ratio quantifies AO correction quality as the ratio of the peak PSF intensity to the diffraction-limited peak:
where $\sigma_\phi^2$ is the residual wavefront variance in radians$^2$(the Maréchal approximation, valid for $\sigma_\phi < 2$ rad). A Strehl ratio of $S > 0.8$ is considered “diffraction-limited.” The wavefront error budget includes contributions from fitting error, temporal bandwidth error, measurement noise, anisoplanatism, and non-common-path aberrations:
The fitting error for $N_{\text{act}}$ actuators is $\sigma_{\text{fit}}^2 = a_F (d/r_0)^{5/3}$, where $d = D/\sqrt{N_{\text{act}}}$is the actuator pitch and $a_F \approx 0.23$ for a continuous facesheet DM. Current state-of-the-art systems achieve Strehl ratios of 0.6–0.9 in K-band and 0.2–0.5 in H-band.
6. Multi-Wavelength & Multi-Messenger Astronomy
The electromagnetic spectrum spans over 20 orders of magnitude in frequency. Each wavelength regime reveals different physical processes and populations of astrophysical sources. The emerging field of multi-messenger astronomy adds gravitational waves, neutrinos, and cosmic rays as complementary information channels, enabling a far more complete picture of the universe.
6.1 Radio Astronomy
Radio waves ($\sim 10$ MHz – $\sim 1$ THz) penetrate dust and the ionosphere (above ~10 MHz). Key observables include:
21 cm hydrogen line: The hyperfine transition of neutral hydrogen at $\nu = 1420.405$ MHz. Used to map the Milky Way's spiral structure, measure galaxy rotation curves, and (at cosmological redshifts) probe the epoch of reionization and the cosmic web via intensity mapping.
Synchrotron continuum: Relativistic electrons spiraling in magnetic fields produce broadband polarized emission. Observed from supernova remnants, AGN jets, galaxy clusters. The spectral index encodes the electron energy distribution.
Pulsars: Rotating neutron stars emit beamed radio pulses with extraordinary timing precision. Pulsar timing arrays (NANOGrav, EPTA, PPTA) detect nanohertz gravitational waves from supermassive black hole binaries via correlated timing residuals.
Major facilities: VLA, ALMA, MeerKAT, ASKAP, LOFAR, and the forthcoming SKA (Square Kilometre Array) with a collecting area approaching 1 km².
6.2 Infrared Astronomy
Infrared radiation ($\sim 1$–$300\,\mu$m) traces thermal emission from dust, cool stars, and redshifted optical/UV light from high-redshift galaxies. Atmospheric windows at J (1.25), H (1.65), K (2.2), L (3.5), M (4.8), and N (10 $\mu$m) bands permit ground-based observations; longer wavelengths require space or airborne platforms.
JWST's NIRCam and MIRI instruments operate at 0.6–28 μm, revealing the first galaxies at z > 10, characterizing exoplanet atmospheres via transit spectroscopy, and imaging protoplanetary disks. The Roman Space Telescope (2.4 m, wide-field NIR) will survey billions of galaxies for weak lensing and discover thousands of exoplanets via microlensing.
6.3 X-ray Astronomy
X-rays (0.1–100 keV) are absorbed by Earth's atmosphere and require space-based observatories. X-rays trace hot gas ($T > 10^6$ K) in galaxy clusters, accretion onto compact objects, and coronal emission from active stars. Grazing-incidence mirrors (Wolter type I) focus X-rays at angles below the critical angle for total external reflection ($\theta_c \sim 1^\circ$ at 1 keV).
Chandra
Subarcsecond angular resolution (0.5″); ACIS CCD imaging spectrometer; HETG/LETG grating spectrometers. Optimal for resolving point sources in crowded fields.
XMM-Newton
Largest collecting area (~4650 cm²); 3 co-aligned telescopes with EPIC CCD cameras and RGS gratings. Superior for faint extended emission and spectroscopy.
eROSITA
All-sky X-ray survey (0.2–10 keV); detecting ~3 million AGN and ~100,000 galaxy clusters for cosmology. Eight co-aligned Wolter-I telescopes.
6.4 Gamma-ray Astronomy
Gamma rays ($> 100$ keV) cannot be focused by conventional optics. Space-based detectors (Fermi-LAT: pair-conversion tracker, 20 MeV–300 GeV) survey the entire sky. Ground-based Imaging Atmospheric Cherenkov Telescopes (IACTs) detect the Cherenkov light flashes from particle showers initiated by $> 50$ GeV photons:
Current IACTs include MAGIC (two 17 m telescopes, La Palma), H.E.S.S. (five telescopes, Namibia), and VERITAS (four 12 m telescopes, Arizona). The Cherenkov Telescope Array (CTA) will deploy ~100 telescopes across two sites (La Palma and Chile), achieving an order-of-magnitude improvement in sensitivity from 20 GeV to 300 TeV.
6.5 Neutrino Astronomy
Neutrinos interact only via the weak force, making them unique probes of dense environments opaque to photons — stellar cores, supernovae, AGN jets, and the early universe.
IceCube
One km$^3$ of Antarctic ice instrumented with 5160 optical sensors detecting Cherenkov light from neutrino-induced muons and cascades. Detected the first high-energy astrophysical neutrinos (2013) and identified the blazar TXS 0506+056 as a neutrino source (2018).
Super-Kamiokande
50,000 tonnes of ultrapure water with ~11,000 PMTs. Detected neutrinos from SN 1987A (24 events), confirming core-collapse supernova theory. Discovered neutrino oscillations (1998 Nobel Prize). Successor Hyper-Kamiokande (260,000 tonnes) under construction.
6.6 Cosmic Ray Detection
Ultra-high-energy cosmic rays (UHECRs, $E > 10^{18}$ eV) produce extensive air showers containing billions of secondary particles. The Pierre Auger Observatory (3000 km$^2$, Argentina) combines surface water-Cherenkov detectors with fluorescence telescopes to measure shower profiles and determine primary energy and composition. The Telescope Array (700 km$^2$, Utah) provides complementary northern-hemisphere coverage.
The cosmic ray spectrum follows a broken power law: $dN/dE \propto E^{-\gamma}$with $\gamma \approx 2.7$ below the “knee” ($\sim 3 \times 10^{15}$ eV), $\gamma \approx 3.1$ between knee and “ankle” ($\sim 5 \times 10^{18}$ eV), and a suppression above$\sim 5 \times 10^{19}$ eV (GZK cutoff from pion photoproduction on the CMB).
6.7 Key Space Missions
| Mission | Band | Primary Science |
|---|---|---|
| JWST | NIR/MIR | First galaxies, exoplanet atmospheres, star formation |
| Gaia | Optical | Astrometry of ~2 billion stars; 3D Milky Way map |
| Euclid | VIS/NIR | Weak lensing & galaxy clustering; dark energy |
| Roman | NIR wide-field | Dark energy, microlensing exoplanets, galaxy surveys |
| LISA (2030s) | GW (mHz) | SMBH mergers, compact binaries, EMRIs |
| Einstein Probe | Soft X-ray | X-ray transients, TDEs, GW counterparts |
Summary: Key Equations of Observational Astrophysics
Rayleigh Criterion
CCD Equation
Grating Equation
Visibility
Strehl Ratio
Atmospheric Extinction
Interferometric Resolution
Fried Parameter Scaling