A curved streak in a space photograph can look like an object making an extraordinary turn. Hubble offers a more useful possibility: the camera itself is moving. In the Hubble Asteroid Hunter project, researchers used that changing viewpoint to estimate distances to small asteroids. The curves were not merely something to explain away. With observation times, spacecraft positions and a tested model, they became measurements. For UAP research, the lesson is to reconstruct the observing geometry before turning an image shape into a claim about propulsion.
What the photograph actually records
The official Abell 370 parallel-field image is a useful starting point. Against distant galaxies, nearby asteroids leave curved and sometimes S-shaped streaks. The caption explains that the picture combines multiple Hubble exposures. Its 20 asteroid sightings represent seven distinct objects, not 20 separate visitors.
The telescope orbits Earth while the asteroids follow their own orbits around the Sun. Hubble’s changing viewpoint shifts the apparent position of a relatively nearby object against the distant background: parallax. Earth’s motion and the asteroid’s motion also contribute. The resulting line on the image is therefore not a direct overhead map of the asteroid’s physical flight path.
This distinction does not make the asteroid unreal. Unlike a particle hit on a detector, the trail records light from an external object. What needs correction is the interpretation of its apparent motion. An authentic image and an authentic moving object can coexist with a misleading description of the trajectory.
From interesting streaks to a research sample
ESA/Hubble’s May 2022 project account describes a search through more than 37,000 composite images from observations made between April 2002 and March 2021. Volunteers identified trails that helped train a machine-learning algorithm. The combined effort produced a dataset of 1,701 trails in 1,316 images.
The catalogue comparison left 1,031 trails without a match to known asteroids. Here, “unknown” means unmatched in the relevant database using the observation’s position and time. It does not mean that researchers had established an object beyond ordinary astronomy. A faint asteroid can be recognizable as a likely natural object without having a previously catalogued identity.
The April 2024 follow-up release described how the project used the curves to investigate small main-belt asteroids. The 2022 discovery account and the 2024 analysis concern overlapping work, not two independent discoveries of the same number of objects. Keeping that chronology straight prevents a repeated headline from becoming a doubled discovery.
How a curve becomes a distance estimate
The primary source is García-Martín and colleagues’ 2024 paper, Hubble Asteroid Hunter III: Physical properties of newly found asteroids. Its method uses the trail’s sky coordinates, the exposure’s start and end times, and the telescope’s trajectory. Researchers calculate predicted trails at different object distances and compare them with the observed trail.
That is more demanding than pointing to a bend and saying “parallax.” A candidate distance must reproduce the measured shape while accounting for the observer’s movement and the object’s apparent intrinsic motion. The researchers checked their implementation with simulated observations and with trails from known asteroids whose distances could be compared with JPL Horizons predictions.
After quality assessment and filtering, the analysed sample contained 632 objects: 454 unknown objects and 178 known objects. Those are the study’s retained analysis counts, not a replacement description of every trail initially found. A detection catalogue and a usable measurement sample answer different questions.
The limits are part of the result
The paper explains that some trails are too straight to yield a well-constrained distance. A visually clean streak can still be poor evidence for this particular calculation. Other exclusions involved faint signals, nearby sources affecting the extracted coordinates, or uncertainty about the trail’s endpoints. More pixels do not automatically mean more useful geometry.
Even a successful distance estimate does not provide a complete orbit. The authors note that the velocity component along the asteroid–Earth direction is missing, leaving ranges of possible orbital parameters. Their size estimates also depend on assumed reflectivity, or albedo. Brightness plus distance is informative, but a darker object and a brighter object need not have the same size to produce the same observed signal.
These qualifications are not decorative disclaimers. They specify what the method measures, what it assumes and what remains unavailable. A credible explanation should expose those boundaries rather than conceal them behind an impressive image or a confident institutional headline.
What transfers to UAP analysis
This is not a newly reported UFO encounter, and the Hubble results do not identify an unrelated aircraft-camera or smartphone sighting. Those observations have different platforms, exposure times and environments. The transferable principle is narrower: apparent motion combines the object’s behaviour with the observer’s motion and the way the image was assembled.
A useful investigation therefore asks for the original sequence, timing, camera direction, platform movement and any independent distance information. It should compare an ordinary-motion model with the actual observations, not simply announce that parallax is possible. Conversely, a claim of an impossible turn should not treat a two-dimensional trace as a complete three-dimensional trajectory.
Managing expectations
- Distinguish trails from objects. Repeated exposures can record the same body more than once.
- Define “unknown.” No catalogue match is not evidence of extraterrestrial technology.
- Ask what constrains the explanation. A model needs timestamps, geometry and a comparison with measurements.
- Retain the limits. Distance, size and a complete orbit require different information; one successful estimate does not deliver all three.
Hubble’s curved trails show that an initially strange-looking feature can become useful evidence without becoming an extraordinary vehicle. The productive question is not whether the line looks surprising. It is whether the observing conditions let us measure what produced it.
Sources and scope
- Official image record: ESA/Hubble — Abell 370 Parallel Field with Asteroids. Composite exposures, repeated sightings and parallax.
- Official project account: ESA/Hubble — Global Citizen Science Project Finds Over 1700 Asteroid Trails in Hubble Images (May 6, 2022).
- Official follow-up: ESA/Hubble — Hubble goes hunting for small main-belt asteroids (April 18, 2024).
- Primary peer-reviewed research: García-Martín et al. — Hubble Asteroid Hunter III, Astronomy & Astrophysics 683, A122 (2024). Methods, validation, selection criteria and uncertainty.
Source check: November 10, 2026. Four source pages/documents reviewed. The institutional releases summarize related research; they are not independent replications. UAP comparisons are editorial applications, not findings of this asteroid study.
UAP / UFO Research: People, Films & Sightings
Source-grounded research, clear evidence labels and conclusions proportionate to the record.
Back to UAP Topic Back to Blog