Kepler DR25 · observed to intrinsic

The catalogue is the visible tip of a selected population

A catalogue records what a telescope could see. To ask how common small, long-period planets may be, I had to reconstruct the selection process, account for unreliable candidates and carry those uncertainties into a population model.

50–500 days · 0.5–2 Earth radii · conditional on the selected GK-dwarf sample

Inspect the equations →Read the claim limits →Download the research products ↗

OBSERVED

catalogue candidates

what passed the observed KOI contract

MODEL-INFERRED

latent valid candidates

mean after reliability + FPP imputation

MODEL-INFERRED

78.3

effective stars

median exposure for the inferred shape

MODEL-INFERRED

0.692

planets per selected star

0.267–1.918 · 95% interval

Checking the newest verified posterior…

Selection lens

Four readings of the same survey

Step through the reasoning. Each transition changes the question being asked, so every number keeps its evidence label and unit.

Evidence changes meaning here

Count what the survey recorded

Eighty-nine KOIs pass the fixed stellar and planet-domain contract. That count describes the selected catalogue, not the intrinsic population.

Three-panel Kepler DR25 figure showing catalogue candidates, the total selection surface, and the inferred intrinsic period-radius density — Candidate colour is reliability. The middle panel is total selection probability, including geometry and target-wise pipeline plus vetting response. The final panel shows relative intrinsic density at the median fitted slopes.

What the posterior can say

0.246

Gamma Earth

per selected star per d ln P d ln R at the Earth pivot

0.122

Hsu fixed-box projection

237–500 days · 0.75–1.5 Earth radii

0.041

Earth ±20% projection

fixed period and radius · not a habitable-zone rate

Claim boundary

This is a fixed period-radius occurrence result. It is not a probability of habitability, biology or a star-dependent habitable-zone eta Earth.

What I learned

The missing planets are part of the result

I used to read a planet catalogue as if it were a map of what exists. This analysis taught me to read it as the output of an instrument, an observing schedule, a detection pipeline and a human-designed vetting process.

The posterior is broad because the data are genuinely weak near an Earth-like period and radius. Keeping that width visible is part of the finding. It tells me where a future survey can add the most knowledge.

Continue to my research dictionary and conclusion →