What you're looking at
Everything shares one heliocentric ecliptic J2000 frame, in astronomical units. The Sun sits at the centre; the inner planets out to Jupiter carry their orbits from JPL's approximate Keplerian tables. Each meteor shower's parent body — thirteen comets and two asteroids — is drawn from osculating elements queried from JPL Horizons at this map's epoch and propagated to the instant you are viewing by solving Kepler's equation, the same two-body propagation those daily element files are published for. Its debris stream is a schematic ribbon of particles laid along that orbit: real streams are shed near perihelion and spread over centuries, and are far thinner and clumpier than a smooth tube, but the ribbon shows the essential truth — the dust follows the parent's orbit.
The physics that makes a shower is entirely in the geometry. Where a parent's orbit crosses the plane of Earth's orbit — its node — the stream and our planet can meet. When that node sits right at 1 AU, Earth drives through the densest dust every year and the shower is strong and dependable, like the Perseids (Swift-Tuttle's node: 1.014 AU) or the Leonids (Tempel-Tuttle: 0.981 AU). When planetary tugs have dragged the node inward or outward, Earth clips only the stream's edge, or misses it entirely — which is exactly why comet Biela's once-storming Andromedids have faded to nothing now that both its nodes have drifted past 1.3 AU. Select any shower to see its encounter geometry judged live.
Reading the map honestly
Parent positions are two-body propagations of a single osculating element set per object, taken from JPL Horizons at epoch 2026‑07‑19 TDB. Near this epoch they reproduce Horizons' own state vectors to better than a pixel; scrub years away and the two-body model slowly diverges from the real, perturbed motion — for the Jupiter-family comets most of all — so treat distant epochs as indicative of orbital phase, not a precision ephemeris. Comet 3D/Biela's elements are its last pre-disruption orbit (1852); it no longer exists as a single body, and its marker is shown only to locate the relic stream.
The debris ribbons are deliberately schematic. A real meteoroid stream is not a uniform tube: it is densest along the trails shed at recent perihelion passages, it fans out with particle size and age, and its nodal footprint precesses over millennia. The ribbon here is drawn evenly along each parent's current orbit to show which orbit the dust belongs to and where it threads Earth's path — not the fine structure that decides an outburst year. The node distances quoted in each card are computed from the parent's present elements; for the broad, ancient complexes (the Taurids from Encke, the Southern δ‑Aquariids from Machholz) the shower we actually see comes from the spread of the whole stream, not from the parent's instantaneous node, and the cards say so.
Peak dates and the countdown are generated at run time: the map finds the instant Earth reaches each shower's peak solar longitude using the same planetary ephemeris it draws, so the calendar is internally consistent and recomputes for whatever year you scrub to. Radiant coordinates, entry speeds and Zenithal Hourly Rates are the conventional values from the IAU Meteor Data Center and the IMO working list; a ZHR is an idealised clear-dark-zenith rate, and real observed counts are almost always lower and strongly Moon-dependent.