Watch a distant star brighten for no visible reason - no host star, no resolved image, just a brief lift in brightness. That absence is the whole signature of a rogue planet: a starless, planet-mass object detected purely by gravitational microlensing.
Rogue Planets 3D Explorer
Scrub a gravitational-microlensing event with no host star anywhere in it, and see why that absence - not a resolved image - is how astronomers find free-floating planets drifting alone in the galaxy.
Every confirmed exoplanet you have probably heard of orbits a star. A rogue (or "free-floating") planet does not - it was either flung out of its birth system or never fully formed one, and now drifts through the galaxy with no host at all. Because it emits no light of its own and has nothing to orbit, the only way to catch one is to watch a distant background star and wait for the rogue planet's gravity to briefly, invisibly, bend and brighten that star's light as it passes in front of it.
- A monitored background star with a growing brightness halo driven by the real single-lens (Paczynski) magnification formula
- Three lens-mass presets - a comparison star, a Jupiter-mass planet, and an Earth-mass planet - each showing the real published event duration for that mass class
- A "Show inferred path" toggle that reveals a teaching-aid marker for the otherwise completely invisible lens - by default nothing marks it, matching what a real telescope actually sees
- A survey-estimate toggle comparing the widely reported 2011 population estimate against the 2017 survey that revised it down
- A live light-curve canvas tracing brightness vs. event progress, with Play/Pause and a scrub slider
- Runs fully in the browser with the vendored three.js engine - no account, no upload; your lens-mass and view choices are remembered on your device
| Quantity | Value |
|---|---|
| Shortest microlensing event ever recorded | OGLE-2016-BLG-1928, 41.5 minutes (Mroz et al. 2020) |
| That event's implied lens mass | about 0.3-2 Earth masses; no companion star within 8 AU |
| Typical stellar-mass microlensing event | about 20-40 days |
| 2011 population estimate | ~1.8 Jupiter-mass objects per star, ~75% unbound (Sumi et al. 2011) |
| 2017 revision | 95% upper limit of ~0.25 Jupiter-mass free-floating planets per star (Mroz et al. 2017) |
| JWST direct-imaging evidence (2023) | ~540 planet-mass objects in the Orion Nebula, including 42 JuMBO pairs/triples |
| Ground surveys | OGLE, MOA, KMTNet |
Why the lens itself is never seen
In every other microlensing page on this site, the foreground lens is a visible star and the planet only adds a small secondary wiggle to that star's own long brightening. A rogue planet flips that entirely: there is no star doing the lensing, so there is nothing bright to see in the foreground at all. The single brief brightening of the background star - and nothing else - is the entire detection. That is also why the event is so short: the smaller and lighter the lens, the faster it crosses the line of sight, and a planet-mass lens crosses in hours or less instead of the weeks a star-mass lens takes.
The 2011 population estimate suggested rogue Jupiter-mass planets could outnumber ordinary stars roughly two to one - a striking headline figure at the time. A much larger follow-up survey in 2017 did not confirm that excess, instead placing a far lower upper limit on how many exist. Both figures are real, published results from named surveys; the explorer presents the revision honestly rather than picking whichever number sounds more dramatic. Separately, in 2023 the James Webb Space Telescope found hundreds of young, still-warm planet-mass objects directly imaged in the Orion Nebula - a completely different detection method that does not depend on microlensing at all, and one of the few ways free-floating planets can be seen rather than only inferred.
For a planet detected while still orbiting its own star via the same microlensing technique, see the Gravitational Microlensing Exoplanet 3D Explorer - that page's secondary spike sits on top of a multi-week host-star event; this page's entire event is the short signature of having no host star at all.
Everything renders on your device with WebGL. The 3D engine loads once (about 0.7 MB) and is cached, and nothing about your visit is sent to a server.
This is an educational approximation, not a lensing simulation: the brightness curve uses the standard single-lens magnification formula, and the visual width shown for each mass preset is illustrative rather than a computed physical timescale - the real duration for each mass class is the cited figure in the facts panel and the table above.
Frequently Asked Questions
What is a rogue (free-floating) planet?
A planet-mass object that does not orbit any star - either flung out of its birth system or never fully attached to one. It drifts through the galaxy on its own, emitting no light of its own to detect.
How do astronomers find a planet with no host star and no light of its own?
Almost entirely through gravitational microlensing: when the rogue planet's gravity happens to pass directly in front of a distant background star, it briefly bends and focuses that star's light, causing a short, one-time brightening. No resolved image ever forms - only the brightness change reveals it.
Why is a rogue planet's microlensing event so much shorter than a normal one?
Event duration scales with the mass of the lens - a lighter lens crosses the line of sight faster. A star-mass lens produces a multi-week event; a Jupiter-mass rogue planet's event lasts hours to about a day; the lightest confirmed candidate, OGLE-2016-BLG-1928, lasted only 41.5 minutes.
What is OGLE-2016-BLG-1928?
The shortest gravitational-microlensing event ever recorded, lasting 41.5 minutes (Mroz et al. 2020). Its properties point to a lens of about 0.3 to 2 Earth masses with no companion star detected within a projected separation of 8 AU - the shortest, and one of the most closely-studied, rogue Earth-mass planet candidates found so far.
How many rogue planets are there really?
Estimates have changed as surveys grew. A 2011 study suggested roughly 1.8 Jupiter-mass objects per main-sequence star, with about three-quarters likely unbound - close to twice as many rogue Jupiter-mass planets as stars. A much larger 2017 survey did not confirm that excess, instead setting a 95% upper limit of about 0.25 Jupiter-mass free-floating planets per star. This explorer shows both figures rather than picking one.
Has the James Webb Space Telescope found rogue planets too?
Yes, by an entirely different method. In 2023, JWST directly imaged about 540 young, still-warm planet-mass objects in the Orion Nebula's Trapezium region, including 40 paired and 2 tripled "Jupiter-mass binary objects" (0.6-14 Jupiter masses). Direct imaging works there because the objects are young and still glowing with formation heat - it does not rely on microlensing at all.
How is this different from the Gravitational Microlensing Exoplanet explorer on this site?
That page shows a planet still orbiting a visible lens star - the planet only adds a short secondary spike on top of the star's own multi-week brightening event. This page has no host star at all: the planet-mass object itself is the entire lens, and the whole event (not a secondary spike) is short, which is the defining signature of a genuinely free-floating planet.
Is this a real physics simulation?
No. It is an educational approximation, not a lensing simulation. The brightness curve uses the standard single-lens (Paczynski) magnification formula; the visual width shown for each mass preset is illustrative, not a computed physical timescale - the real event duration for each mass class is the cited published figure shown in the facts panel.