A Tumbling Object Filmed over Jinzhou, Liaoning
The object is unquestionably there, at thirty two standard deviations above the sky. What it is not is rigid.
What did witnesses see at Jinzhou?
A witness in Jinzhou, Liaoning province filmed a small dark object moving against a pale evening sky on 16 October 2026. The clip runs just under twenty seconds and the object is in it for almost all of them.
The structured filing puts the sighting at 5:31 PM CST on 16 October 2026 over Jinzhou, with the observer facing west. One object is reported. Its closest distance is given as very close, under fifty feet, its apparent size as a basketball held at arm's length, and its movement as hovering, floating around, straight and steady, circular, unpredictable, erratic, random and smooth, which is most of the options on the form at once. The witness wrote that it flew extremely smoothly and was completely unaffected by its surroundings, that the clip is part of a longer recording, and that the sighting was still in progress when the report was filed.
The footage shows an urban skyline of apartment towers under a broken altocumulus sky still holding daylight, with lit windows beginning to show. The object is a small dark shape above the rooftops. The camera is handheld and follows it, and the frame swings across most of the available sky during the clip.
One field in the filing deserves to be read before anything else. Under the five observables, the question asking whether the object had wings is answered Yes. That is the reporter's own observation, recorded in their own form, and it sits uneasily beside a report of an object under intelligent control.
What is the official explanation?
No Chinese agency has commented on this sighting, and no coverage of it was found in Chinese, English or any other language. There is no official narrative to weigh.
There is also no usable local weather record. The Iowa Environmental Mesonet holds a station for Jinzhou, ZYJZ, but its archive for that site ends in August 2014. The nearest active reporting stations are Shenyang Taoxian, about 190 kilometres north east, which logged wind from 010 degrees at four knots at 09:30 UTC with no cloud layers reported, and Dalian Zhoushuizi, about 250 kilometres south, which logged wind from 100 degrees at twelve knots with scattered cloud at 1,500 feet. Neither describes the air over Jinzhou. The honest position is that the surface wind at the site is unrecorded, while the footage itself shows a textured altocumulus deck in motion overhead.
The filing carries its own automated note, and it is a limitation rather than a finding: the video has no creation timestamp and may have been re encoded, downloaded or re saved, and the reporter chose not to publish the device and camera settings or the capture date and time the file originally contained. The archive copy on the reporting site's media subdomain could not be retrieved because the parent page returns a Cloudflare challenge that cannot be cleared without an interactive browser. The oldest traceable appearance of this footage is therefore the Reddit post of 19 October 2026 at 22:18 UTC.
What did the witnesses think it was?
The thread reached a verdict quickly and unanimously, which is itself unusual. The first commenter asked whether it could be a plastic bag. A second agreed it was a plastic bag or balloon. A third wrote that it is obviously moving with the wind. A fourth described it as a light piece of plastic swept up and carried around. Only one reply pushed back, and it did so with a rhetorical question rather than an argument.
A thread consensus is not evidence. It is a reason to go and measure, which is what the archive did.
Is the A Tumbling Object Filmed over Jinzhou, Liaoning real? The two-pass assessment
Pass one, how could this be mundane. The test that settles a case like this is the silhouette stability test, and it needs nothing but the clip. A rigid body holds its outline. Its apparent area changes slowly with range and its aspect ratio changes smoothly as it turns. A thin flexible sheet does neither: it crumples, it presents broadside and then edge on, and its silhouette collapses and reopens within a second.
The object was tracked by subtracting a Gaussian blurred background from each frame, thresholding the negative residual at four times the local sky noise, and following the detection from frame to frame with a continuity gate. It was recovered in 490 of the 554 frames between t equals 1.40 and 19.83 seconds.
The object is real and the measurement is not noise. It sits 38.8 grey levels below the local sky against a noise floor of 1.24 grey levels, which is about 32 standard deviations. Whatever it is, it is a physical thing in front of the sky.
The silhouette is not stable, and not by a small margin. Its area averages 195.8 pixels with a standard deviation of 247.0, a coefficient of variation of 126.2 per cent, ranging from 20 to 806 pixels, a factor of 40. Its aspect ratio runs from 0.23 to 3.00. At t equals 14.17 seconds it is seven pixels wide and thirty tall. At t equals 17.57 seconds it is nine wide and three tall. The largest swing inside any single second is a factor of 30.2, from 21 pixels to 634, between t equals 14.33 and 15.93 seconds. The area filling its own bounding box averages 0.630, so the outline is ragged and concave rather than a disc or an oval.
The obvious objection is range: perhaps the object is simply moving toward and away from the camera. That objection is testable and it fails. If the apparent area were collapsing because the object were receding, its contrast against the sky would fall with it. It does not. The correlation between area and contrast across the whole track is 0.110, which is effectively none, and the contrast measured on the frames in the lowest decile of area is 34.9 grey levels against 45.3 on the frames in the highest decile. The object stays almost exactly as dark while its silhouette collapses by a factor of thirty. That is rotation, not range.
Look at the enlargements and the numbers stop being necessary. At 2.9 seconds it is a nine by six pixel fleck. At 4.4 seconds it is a thirty one by twenty seven pixel hooked mass with a concave bite out of one side and a brighter lobe beside a darker one. At 7.3 seconds it is a thin diagonal sliver. At 12.4 seconds it is a seven by four pixel speck. At 15.4 seconds it is a twenty nine by twenty four pixel irregular form again. A structured craft does not pass through those shapes in twelve seconds. A crumpled sheet of thin film tumbling in air does, and the near edge on moments are exactly what a flat sheet produces when it turns through the line of sight.
Pass two, if it is real. The pass two question here is narrow, because the measurement has already removed the reading that made the case interesting. There is no honest version of this in which the object is a rigid craft under control, because a rigid outline is the one thing the footage demonstrates it does not have. What remains is whether the thing is anomalous as a flexible object, and nothing suggests it is. The witness's own strongest argument, that it moved smoothly and was completely unaffected by its surroundings, is the argument against them: an object with no propulsion, no control surfaces and very low mass per unit area moves exactly like that, because it goes wherever the air goes. The Yes in the wings field is the reporter recording the flat, sheet like appendages the enlargements show.
Tiered Strongly Disputed, following the precedent this archive set at Campo Largo and Spring Hill, and deliberately not Discredited. An honest witness, genuine footage and a misidentification are not a hoax, and nothing here suggests any intent to deceive. The limit is stated plainly: the measurement demonstrates a thin, flexible, tumbling, wind carried body, and it does not identify that body. Plastic film is the commonest thing of that description in the air over a Chinese industrial city, and it is the thread's answer, but the archive is demonstrating a class of object and not naming a specific one. The checks that would close it completely, the original file's capture metadata and a local wind record, are both unavailable: the reporter withheld the metadata and the nearest working weather station is 190 kilometres away.
Sources
- www.reddit.com/r/UFOs/comments/1xaaxdz/unknown_object_hovered_and_moved_erratically_over/
- ufosighting.report/sighting/13160/unknown-object-hovered-and-moved-erratically-over-jinzhou-li
- mesonet.agron.iastate.edu/request/download.phtml?network=CN__ASOS
- mesonet.agron.iastate.edu/sites/networks.php?network=CN__ASOS
More cases from this region: UFO sightings in China
