photometry-pipelines
Measuring brightness from images — aperture and PSF photometry, calibration to standard systems, and light curves.
- License 许可
- MIT
- 最近更新
- 2026-11-07
- Stars
- ★ 0
- Forks
- 1
01 前言 Front Matter
- name
- photometry-pipelines
- category
- scientific
- description
- Measuring brightness from images — aperture and PSF photometry, calibration to standard systems, and light curves.
03 SKILL.md 正文
Overview
Photometry converts pixel counts into physical fluxes and magnitudes — the basis of light curves, color–magnitude diagrams, and transient detection. This skill covers aperture and PSF-fitting photometry, sky background handling, calibration to standard systems (zero points, color terms, extinction), and building reliable light curves.
When to use
- Measuring magnitudes of stars, galaxies, or transients in images
- Building light curves: variable stars, exoplanet transits, supernovae, AGN
- Calibrating instrumental magnitudes to a standard system (SDSS, Pan-STARRS, Gaia)
- Doing difference imaging for transient detection
- Deciding between aperture and PSF photometry for a field
Core concepts
-
Magnitude system: m = −2.5 log₁₀(F) + ZP — logarithmic, relative; the zero point (ZP) ties instrumental counts to physical flux. Colors (differences of magnitudes) cancel the ZP.
-
Aperture photometry: sum flux in a circular aperture, subtract sky from an annulus — simple, robust for isolated sources; aperture corrections account for flux outside the aperture (curve of growth).
-
PSF photometry: fit the point-spread function (DAOPHOT/DoPhot/photutils) — mandatory in crowded fields; yields better precision and deblending at the cost of complexity.
-
Sky background: the dominant error source for faint sources — measure locally (annulus or background mesh), watch for gradients and contamination by neighbors.
-
Calibration ladder: instrumental mag → zero point (standard stars in the field) → color terms (filter mismatch) → atmospheric extinction (airmass correction) → standard system. Each rung needs doing.
-
Differential photometry: for light curves, measure target relative to stable comparison stars in the same frame — cancels clouds, airmass, and most systematics without absolute calibration.
-
Difference imaging (DIA): subtracting a reference template isolates variable sources in crowded fields — the standard for microlensing, supernova, and dense-cluster transient searches; convolution matching (Alard–Lupton) is the core algorithm.
-
Crowded-field PSF fitting: DAOPHOT/DoPhot iteratively fit and subtract stars — aperture photometry is simply wrong in globular clusters and the Galactic bulge; use the right tool for the field.
-
Color terms: filter+detector response never exactly matches the standard system — the linear (sometimes quadratic) color term in the calibration equation absorbs the mismatch; skipping it leaves 0.05–0.2 mag systematics.
Practical workflow
1. Prepare and detect
from photutils.detection import DAOStarFinder
from photutils.background import Background2D, MedianBackground
bkg = Background2D(data, box_size=64, filter_size=3,
bkg_estimator=MedianBackground())
finder = DAOStarFinder(fwhm=4.0, threshold=5.0 * bkg.background_rms)
sources = finder(data - bkg.background)
- Work on calibrated (bias/dark/flat-corrected) images; estimate and subtract background with a mesh, checking for over-subtraction around bright sources.
- Detect at ≥5σ; set FWHM from measured stars, not guesses.
- Choose aperture radius ~1.5–2× FWHM for the measurement aperture, sky annulus starting ~3× FWHM — then derive aperture corrections from bright isolated stars.
2. Measure
- Sparse fields: aperture photometry with local sky; apply curve-of-growth aperture corrections per frame.
- Crowded fields: build an empirical PSF from isolated stars; fit with PSF photometry; iterate (subtract neighbors, refit).
- Record per-source uncertainties: photon noise + sky noise + read noise — and check that the scatter of constant stars matches the error model (it should, within ~10–20%).
3. Calibrate
- Match to a reference catalog (Pan-STARRS, SDSS, Gaia synthetic photometry, APASS) with proper cross-matching radius and quality cuts.
- Fit ZP + color term: m_std = m_inst + ZP + c × color; fit per frame or per night; check residuals vs magnitude, color, and position (spatial trends mean flat-field or illumination problems).
- Correct atmospheric extinction via airmass (m₀ = m − kX) when comparing across elevations; better yet, calibrate each frame independently.
4. Build light curves
- Use differential photometry against an ensemble of comparison stars (weighted mean); reject variable comparisons by their own scatter.
- Detrend systematics (airmass, seeing, position on chip) — but verify the "trend" isn't the astrophysical signal (compare detrended vs raw).
- For transits/periodicity: normalize out-of-transit baseline, then run BLS/Lomb-Scargle; always phase-fold and inspect by eye — algorithms find periods in noise.
5. Run a transient-detection pipeline
- Build a deep reference template from the best-seeing images; difference each new image with PSF-matched convolution.
- Detect on the difference image (≥5σ), then vet: reject subtraction artifacts (dipoles from misregistration), asteroids (motion between exposures), and variable stars (catalog cross-match).
- Force-photometer the candidate position on all epochs — including pre-discovery non-detections, which constrain the explosion/brightening time.
Common pitfalls
- Aperture photometry in crowded fields: neighbor flux contaminates both aperture and sky annulus — switch to PSF fitting.
- Global sky subtraction: a single background value across a frame with gradients or nebulosity biases faint sources — go local.
- Catalog mismatch: calibrating against a catalog in a different filter system without color terms — systematic errors of 0.05–0.2 mag.
- Underestimated errors: photon-noise-only error bars ignore flat-fielding, sky estimation, and systematics — validate against constant-star scatter.
- Detrending away the signal: aggressive systematics removal can erase shallow transits or slow variables — detrend conservatively and compare.
- Saturation blindness: saturated stars have flat-topped profiles and garbage magnitudes — flag and exclude above the linearity limit.
- Aperture corrections from the wrong stars: deriving corrections from saturated or crowded stars corrupts every magnitude — use bright, isolated, unsaturated stars only.
- Ignoring correlated noise: resampled/stacked images have correlated pixels — naive sky-noise estimates understate uncertainties; measure noise empirically from blank regions.
04 同仓库更多 Skills
web-scrape
Turn web pages into LLM-ready markdown via the Firecrawl API (renders JS, anti-bot handling). Use when browser_open/fetch fails on JS-heavy pages, or when you need clean markdown/structured extraction from a site. Requires FIRECRAWL_API_KEY.
design-taste-frontend
tasteskill v2, the anti-slop frontend skill. Art direction for websites, distilled from a study of 247 curated gallery sites (One Page Love, Awwwards Sites of the Day). Deals researched directions (idea, composition, type, palette, photographs, section blueprints) from a library, enforces a strict no-slop house style with a real-browser inspector, and runs a look-and-fix loop on screenshots. Use this whenever you build, design, redesign, restyle, polish or review a landing page, marketing site, portfolio, product page, one-pager, hero or any public-facing web page, in any stack (HTML/CSS, React, Next.js, Vue, Svelte, Astro, Tailwind) — also when the user says a page looks generic, templated, "AI-made", sloppy, or asks to make it look designed, premium, award-level or less like every other site. Not for dashboards or app screens.
test-driven-development
Use when implementing any feature or bugfix, before writing implementation code
systematic-debugging
Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes
requesting-code-review
Use when completing tasks, implementing major features, or before merging to verify work meets requirements
brainstorming
You MUST use this before any creative work - creating features, building components, adding functionality, modifying behavior, or planning anything new, in software or out of it (a talk, a business, a renovation).