sidramiconi.comTransit Lab

Real Kepler data · Kepler-6 b

How big is the planet crossing this star?

NASA Kepler measurementsYour planet's dip

  1. 1Watch the planet cross and the dip draw itself.

  2. 2Guess its size until your line fits NASA's dots.

    11.2× Earth
    Closeness —
  3. 3Check your guess against the published fit.

How it works, the exact method and the data

What the game draws

The star and the planet are drawn to scale, from the published values: the star is 1.391 times the Sun's radius, and the planet crosses 0.141 of a star radius from the centre on an orbit 7.503 star radii wide (inclination 88.93°). Earth, Neptune and Jupiter are drawn at the same scale for comparison.

Your line is a model of the dip: a dark disc of the size you pick crossing a star that is brighter in the middle than at its edge (limb darkening, with Kepler-6's fitted coefficients u₁ = 0.480 and u₂ = 0.138), averaged over Kepler's 29.4-minute exposures. The dots are NASA's real measurements, folded on the published orbit.

Closeness compares your size with the size whose model dip best fits these dots by least squares (0.0932 of the star's radius). Every 1% away from it costs 3 points. That best fit uses one quarter of data and the published orbit, so it differs slightly from NASA's published fit (0.0942), which was fitted to more data with the orbit free.

Measure the dip: the exact method

When a planet crosses its star, it blocks light. For a uniformly bright star, the share of light blocked equals the share of the star's disc the planet covers. So the depth of the dip is roughly (planet radius ÷ star radius)². Take the square root of the depth, multiply by the star's radius, and you have the planet's radius.

Measure the dip takes the median brightness outside the blue window as the baseline, and the median of the central hour as the bottom. The comparison uses the planet-to-star size ratio, the quantity the depth measures directly; the radius in Earths then depends on the star's radius (1.391 × Sun in the archive's table). The archive lists the planet's radius (14.62 × Earth) and its size ratio separately, and they don't quite agree: 0.09424 × 1.391 Suns is 14.30 × Earth. So compare the size ratios directly.

Brightness of Kepler-6, folded on the planet's orbitKepler · NASA MAST
The whole quarter, unfolded

Drag the gold line down to the bottom of the dip, or press Measure the dip.

Planet radius
—× Earth
Dip depth
—%
Planet / star size
—
In Jupiters
—× Jupiter
Archive-listed radius
14.6× Earth

Planet / star size: yours vs the published fit

You—
Published0.0942

Measure the dip to compare.

A star is brighter in the middle than at its edge (limb darkening), so a planet crossing the middle blocks more light than its size alone would. This approximate correction uses Kepler-6's fitted limb-darkening coefficients and published impact parameter, treating the planet as small compared with the star.

What a simple depth estimate leaves out

  • Limb darkening. The star's centre is brighter than its edge. Kepler-6 b crosses close to the centre (impact parameter 0.14), so the dip is deeper than the planet's size alone predicts and the plain estimate comes out too big. The correction uses the quadratic coefficients fitted to Kepler-6's light curve (u₁ = 0.480, u₂ = 0.138).
  • Geometry. For the same planet size, a planet that only grazes the star's edge generally makes a shallower, V-shaped dip; the published fit models the full shape, including the orbit's tilt.
  • Blending. Light from a nearby star in the same pixels would make the dip look shallower. The Kepler pipeline's PDCSAP flux already corrects for the crowding it knows about.
  • Folding. The points are folded with the published period and transit time, so this measures the depth, not the orbit.

This is an educational measurement, not a replacement for a full transit fit. The published value comes from such a fit.

Where the data comes from

Kepler data from NASA MAST. Planet, star and orbit values from the NASA Exoplanet Archive (Esteves et al. 2015). Limb-darkening coefficients from Müller et al. 2013, as tabulated by Espinoza & Jordán 2015 (Table 1). Built by Sidra Miconi with Claude Opus 5.5 writing the code. @SidraMiconi