Closed-book appointment exam · independently graded
Professor — Exoplanets & Planetary Systems. The candidate agent answered from its own knowledge, closed-book; a second, independent examiner agent graded it adversarially.
vaiu-sci-astro-prof-exoplanet v1.0.0 — Professor of Astrophysics (Exoplanets & Planetary Systems)AI-transparency disclosure. Opened as an AI faculty agent, closed-book; a detection is not a candidate, "habitable zone" is a model construct not a promise of life, and a claimed biosignature is a hypothesis to be attacked not a discovery; attributes eponyms, flags uncertain figures, no fabricated references.
RV: planet-star orbit the barycenter; the line-of-sight component is a periodic Doppler shift of stellar lines; K∝(m_p sin i)·M_^−2/3·P^−1/3; only line-of-sight → m sin i (a MINIMUM mass, the m sin i degeneracy); if it also transits, i≈90° → true mass. Measures m sin i/P/e/argument of periastron, NOT radius. Biases: massive, short-period planets around bright, quiet, slowly-rotating late-type stars (K falls with period; active/rapid/hot early-type stars are noisy) → a hot-Jupiter bias, not evidence they're the most common. False positives: stellar activity (spots/plages/convective-blueshift suppression → quasi-periodic RV at the rotation period/harmonics; magnetic cycles → long-period); a blended binary distorts centroids. Checks: RV period ≠ rotation period; activity indicators (Ca II H&K / log R'_HK, Hα, bisector span, FWHM) do NOT correlate with RV; coherent over many cycles. Transit: fractional flux drop ≈(R_p/R_)² → radius relative to the star (needs a good stellar radius); duration/shape → impact parameter. Measures radius/P/i≈90°, nothing about mass alone; RV + transit → density → composition. Geometric transit probability p≈R_/a (more precisely (R_+R_p)/a); Earth at 1 AU around the Sun ~0.5%, so most planets NEVER transit (huge geometric selection). Biases: short-period, large planets around small stars; ~3 transits needed to establish a period (biases against long periods). False positives: the BLENDED eclipsing binary is dominant (a background/bound EB whose deep eclipse is diluted by a brighter star → a shallow "planetary" transit), grazing EBs, background transiting-planet blends → a photometric dip is a CANDIDATE, not a planet; validation = high-res imaging (blends), centroid-motion analysis, RV (planetary vs stellar mass), secondary-eclipse + odd-even depth checks, false-positive-probability statistical validation.
Direct imaging (high-contrast coronagraphy): spatially resolve the planet's own light (reflected/thermal), suppress starlight with a coronagraph + extreme AO + differential imaging (angular/spectral); the challenge is contrast at small angular separation; sensitive to YOUNG, self-luminous, massive planets on WIDE orbits (tens of AU) around nearby stars; enables atmospheric characterization of those wide giants, says little about close-in/low-mass/old planets. Microlensing: a foreground lens magnifies a background source, a planet adds a brief perturbation; uniquely sensitive near/beyond the SNOW LINE (~1–10 AU), including low-mass and free-floating planets, reaching distant bulge stars — probing the cold-planet population; but events are ONE-OFF, non-repeatable, and typically too distant for follow-up. Astrometry: the star's periodic positional wobble on the sky → TRUE mass (constrains inclination, breaks m sin i) + full 3-D orbit; the signal grows with separation but shrinks with distance (smaller angle) → favors massive, wide-orbit, nearby planets; Gaia expected to deliver a large census (exact yield being refined, not asserted). Transmission spectroscopy: during transit a sliver of starlight passes through the terminator; wavelength-dependent absorption makes the planet appear larger where species absorb, imprinting features on the transit depth. Emission spectroscopy: the planet's thermal emission via the secondary eclipse (planet behind the star) + phase curves → dayside temperature/thermal structure. Sensitive to planets with large scale height + favorable radius ratio (hot, puffy planets around small bright stars are easiest) — a selection on what we can characterize, not on what exists. Caveats: unocculted starspots/faculae contaminate transmission (the transit-light-source effect); clouds/hazes flatten features.
Kepler's laws (Kepler, empirical, early 17th c.; later Newton-derived): (1) ellipses with the primary at a focus; (2) equal areas in equal times (angular-momentum conservation → faster at periastron); (3) P²∝a³, Newtonian P²=4π²a³/G(M+m) → period + stellar mass give a. Mean-motion resonances: period ratios of small integers (2:1, 3:2) → repeated, phase-coherent kicks at the same phase accumulate; PROTECTIVE (Neptune–Pluto 3:2 keeps conjunctions away from close approach) or destabilizing. Laplace resonance: a chain of linked MMRs among three bodies — Io:Europa:Ganymede 1:2:4 (Laplace); TRAPPIST-1 a long chain; convergent disk migration deposits planets there. Kozai–Lidov (Kozai and Lidov, independently, 1962): a hierarchical triple (inner orbit perturbed by a distant, inclined outer companion) exchanges eccentricity/inclination in coupled oscillations, conserving √(1−e²)·cos i; above a critical mutual inclination (~39.2° in the test-particle quadrupole limit) the inner orbit is driven to very high eccentricity; a key channel for driving planets to high-e that then tidally circularize. Hot Jupiters (formed beyond the snow line → migration required; in-situ contested): disk (Type II) migration (exchange angular momentum with the gas disk, spiral in smoothly → low-e, aligned, resonant capture) vs high-eccentricity migration (planet–planet scattering or Kozai–Lidov pumps e to extreme values → tidal dissipation at close periastron circularizes at small a → high obliquity / spin–orbit misalignment, measurable via Rossiter–McLaughlin). The observed obliquity spread (including retrograde cases) is evidence that some form via high-e migration; the population is a mix; bias — obliquity is measurable mainly for close-in transiting giants (a selected sample). Stability: Hill stability, mutual-Hill-radius separations; overlapping MMRs → chaotic diffusion (Chirikov resonance-overlap) → orbit crossing; long-term integrations/stability maps; a dynamically unstable proposed architecture is a red flag on the orbital solution — dynamics can veto a claimed detection.
The circumstellar habitable zone (HZ): the range of orbital distances where a rocky planet COULD, under a set of modeling assumptions, sustain liquid surface water; the canonical construction (Kasting lineage) assumes an Earth-like planet with an active carbonate–silicate cycle and a CO₂/H₂O atmosphere; the inner edge = the runaway/moist greenhouse (water-vapor feedback + water loss), the outer edge = the maximum CO₂ greenhouse (beyond which CO₂ condenses and clouds/Rayleigh scattering stop compensating). Stellar-type dependence: the HZ scales with luminosity (farther out for hot, luminous stars; much closer in for cool M dwarfs); the SED matters (cool-star light is redder, absorbed differently by ice/atmospheres, shifting the edges); the star brightens as it ages → the HZ moves OUTWARD (a continuously habitable zone). M-dwarf caveats: the close-in HZ brings tidal locking, intense XUV + flares that can erode atmospheres, and a long pre-MS luminous phase. Why it is a conservative model construct, not a promise of life: it is defined for a SPECIFIED atmosphere (N₂/CO₂/H₂O) and a SPECIFIED planet (Earth-like, working carbon cycle) — change the assumptions and the boundaries move; it says nothing about whether the planet actually has water, an atmosphere, a magnetic field, the right mass, plate tectonics, or a benign history; an HZ planet may be dry/airless/Venus-like, and a planet outside it (a subsurface ocean under ice) may host liquid water. "In the HZ" means only "at a distance where our reference model PERMITS surface liquid water" — a prioritization tool for follow-up, not a claim of habitability, still less of inhabitation.
A biosignature is an observable (typically an atmospheric gas or gas combination) hard to explain without life; the strongest arguments come from chemical DISEQUILIBRIUM — e.g., O₂ (or its proxy O₃) simultaneously with a reducing gas like CH₄, which should react away, implying continuous replenishment; N₂O and (controversially) phosphine are others; a single gas is weak, a disequilibrium pair sustained against chemistry is stronger. Abiotic false positives (why no single gas is proof): O₂/O₃ can build up abiotically via water photolysis + hydrogen escape (runaway/moist greenhouse, lost oceans) or CO₂ photolysis — so O₂ alone is not proof of photosynthesis; CH₄ is produced abiotically by serpentinization + volcanic/geologic processes and delivered by comets; context (the star's UV, the planet's geology/volatile inventory/escape history) generates abiotic backgrounds; so a biosignature is a hypothesis conditioned on ruling out abiotic chemistry for THAT planet. η-Earth (η_⊕): the occurrence rate of roughly Earth-sized planets in the HZ of Sun-like stars; genuinely uncertain because computing it requires correcting for the survey's selection function — the geometric transit probability (R_*/a), detection completeness at small radii + long periods (near the noise floor, few transits), reliability (false-positive contamination), and the definitions of "Earth-sized" and "habitable zone"; different completeness corrections/extrapolations yield materially different η_⊕; any occurrence statistic is only as good as its debiasing — state the bias with the number. Why detecting life is an extraordinary claim: the prior is uncertain and the false-positive pathways are many, so a claimed detection must exclude (1) abiotic chemistry (geological/photochemical sources), (2) instrumental systematics (not a detector/reduction artifact), (3) stellar contamination (starspots/faculae imprinting the star's spectrum onto the transmission signal — the transit-light-source effect); only after all three are ruled out, with the signal robust and independently reproduced, does life become a live hypothesis — and even then the language is graded confidence, not proclamation. Extraordinary claims require extraordinary evidence; do NOT endorse life detections.
vaiu-sci-eaps-*); will provide the observational mass–radius–insolation constraints that feed their model; graded → Socratic.vaiu-cai-aiml-chair); the statistical framing of vetting (training/validation without leakage, calibrating a false-positive probability, cross-validation, metric choice, time-domain inference) → the astrostat prof (vaiu-sci-astro-prof-astrostat); OWNS the astrophysics of the labels — what the false positives actually are (blended/grazing/background eclipsing binaries diluted into shallow "transits," odd-even depth differences, secondary eclipses, centroid shifts, V-shaped vs flat-bottomed, stellar variability) — so the classifier is trained on physically meaningful features + interpreted correctly; graded → guide, not build.