Survey astronomy is an inverse problem: the catalogue is filtered by the instrument and by the scientist's cuts.
Survey astronomy is an inverse problem: the catalogue is filtered by the instrument and by the scientist's cuts.
Galactic astronomy observes projected light, velocities, and abundances from one moving vantage point. The inverse problem is underdetermined until geometry, selection, and a physical model are made explicit.
Definition: A model for the probability of the observed data as a function of parameters, \(p(D|\theta)\).
Why it matters here: It encodes the measurement uncertainties used to compare Galactic models with survey data.
Relationship: For independent Gaussian rotation points it is related to a chi-square residual sum.
Limitation or common confusion: A convenient likelihood can be precise yet wrong if errors are correlated or non-Gaussian.
Definition: A probability distribution for parameters before conditioning on the current data.
Why it matters here: It regularizes underconstrained distance and Galactic-structure inferences and makes assumptions explicit.
Relationship: Bayes' theorem combines it with the likelihood to form \(p(\theta|D)\).
Limitation or common confusion: A prior is not an additional observation, and an apparently weak prior can matter with sparse data.
Definition: The probability distribution of model parameters after combining likelihood and prior.
Why it matters here: It summarizes parameter estimates, uncertainty, and covariance for the adopted model.
Relationship: Posterior predictive checks connect inferred parameters back to observable catalogues.
Limitation or common confusion: A narrow posterior does not protect against a misspecified model or selection function.
Definition: The probability that an astronomical source is detected and retained as a function of its properties.
Why it matters here: It is part of the data-generating process and must enter population inference.
Relationship: The predicted underlying population is multiplied by selection before comparison with a catalogue.
Limitation or common confusion: Cuts applied by the analyst are part of the selection, not harmless bookkeeping.
Definition: A comparison between data simulated from the fitted posterior model and the actual observations.
Why it matters here: It tests whether the model reproduces features beyond the fitted parameter summary.
Relationship: Residual distributions, trends, and replicated catalogues reveal model-data mismatch.
Limitation or common confusion: Passing selected checks does not prove the model unique or physically true.
The central relation for this lecture is:
Define every symbol before substitution. State the coordinate system, units, and approximation. A formula is evidence only when its assumptions match the tracer and regime.
The evidence display for this lecture is the labeled model/data relationship introduced above. Read the axes and annotations before accepting the inferred trend.
Data context: the accompanying values are standard published or survey-like values; see the course reference log for provenance level and spot-check targets.
A chi-square comparison needs a stated number of fitted parameters and a justified error model.
Write the units beside each intermediate quantity. The numerical result is generated from the shared course constants and recomputed independently during review.
Ask whether changing the distance prior, selection function, or mass decomposition could move the conclusion. This is where a plausible diagram can become a misleading argument.
Prompt: Construct a simple likelihood for rotation-curve points with 8 km s^-1 errors and compare two models.
First estimate the direction of the answer. Then calculate. Finally name one assumption that could reverse the interpretation.
OpenStax Astronomy 2e, Chapter 25, sections 25.1–25.4, and Chapter 26 for comparative context. Exact section and embedded-figure offsets were checked against the local extracted text before release.
Additional sources: Gaia Collaboration (2018, 2021) releases; Bland-Hawthorn & Gerhard (2016), ARA&A 54, 529; McMillan (2017), MNRAS 465, 76; Eilers et al. (2019), ApJ 871, 120; Bland-Hawthorn et al. (2019), Science 365, 859.
References are standard published values not independently re-fetched in this production session unless explicitly marked level 1 in reference-log.md.