Data source. Everything from
the NASA Exoplanet Archive.
The ephemeris is taken whole from a single paper, from the
ps table — the period and the transit time always from the same source. This matters: the composite table
pscomppars gathers parameters from different papers, and for dozens of planets it joins a period from one study to an epoch from another. The combination is not self-consistent and produces errors of hours. Among all the papers publishing a complete ephemeris for each planet, the one chosen is the one giving the smallest uncertainty for today. Everything else — position, magnitude, radii — comes from the composite table, where the gathering does no harm.
What was checked. Among the planets within reach of a 40 cm telescope that have two or more precise ephemerides from independent papers,
88.8% agree on the transit time to within a quarter of an hour, and the median difference between papers is 0.87 minutes. In 21 planets the disagreement is larger, and some of those are known multi-planet systems such as WASP-47 and TOI-1130 where the pull between the planets really does move the transits — there
a linear ephemeris does not describe reality, no matter which paper it came from.
The timing uncertainty. Unlike the eclipsing binaries, here NASA publishes both a period error and an epoch error, so the number is measured rather than assumed: the square root of the sum of the squares of the epoch error and the period error times the number of elapsed cycles. It is recomputed for every date and grows the further you get from the ephemeris. The few planets with no published error are marked explicitly.
The transit depth is shown as a percentage and in millimagnitudes. The conversion is logarithmic: a one per cent drop in intensity is about 11 millimagnitudes. This is the quantity that decides whether the transit is detectable with your equipment — for comparison, a typical eclipsing binary eclipse is a hundred times deeper.
The tool does not account for cloud, a local horizon, or the Moon.
How the windows are chosen. Each window runs from the transit start minus the padding to its end plus the padding. The tool checks that the
whole window — not just the transit — falls inside the site's dark window and stays above the altitude threshold from start to finish. A window where only the transit itself qualifies is not included, because the scheduler would not be able to carry it out. The number of sites on each row is a measure of the chance the request gets scheduled.