Unidentified Aerial Phenomena. Characterization of Dark UAPs

Boris Zhilyaev; Sergey Pokhvala; Vladimir Petukhov

REVIEW 4 major objections 6 minor 9 references

Unidentified Aerial Phenomena. Characterization of Dark UAPs

T0 review · 4 major / 6 minor · reviewed 2026-08-07 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read Dark UAPs are real 3-to-6-kilometer objects flying at Mach 2.5 or faster, this paper argues.

desk verdict New UAP footage and some real photometry, but the kilometer-size/Mach-2.5 conclusion rests on an unvalidated contrast-to-distance model. read the letter →

arxiv 2503.05627 v1 pith:UB3EOPOW submitted 2025-02-13 physics.space-ph physics.pop-ph

Boris Zhilyaev , Vladimir Petukhov , Sergey Pokhvala This is my paper · ORCID

classification physics.space-phphysics.pop-ph

keywords unidentifiedaerialphenomenadarkUAPsalbedoparallaxthermalinfraredcolorimetrylow-albedoobjectstransatmosphericspeeds

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper claims that dark UAPs — objects seen as silhouette-like contrasts rather than glowing lights — are real physical bodies several kilometers across, moving at supersonic and, in one case, transatmospheric speeds, and reflecting almost no sunlight. The claim is built from three unrelated sightings: a synchronized two-camera daytime parallax measurement over Kyiv, a night video of objects crossing the Moon, and a night infrared drone video recorded in the Ukraine combat zone. In all three cases the authors infer sizes of roughly three to six kilometers, reflectivity below three percent, and speeds of Mach 2.5 or higher, and they take that agreement as evidence for a single type of object. If true, this would mean structured objects far larger and faster than any known aircraft or balloon can travel through the atmosphere and near-Earth space.

What carries the argument

The machinery is colorimetric flux comparison. For daytime objects the authors compare reflected sunlight from the object ($I_{obj}=F_{\rm sun}\,r^2/R^2\,\alpha$) with scattered daylight from the sky ($I_{sky}=F_{\rm sky}\,\Omega$) to derive the object size $r$ and albedo $\alpha$ once the distance $R$ is determined by parallax. For the Moon video the same comparison is made between the object and the lunar surface using the Moon's known albedo. For the night drone video the distance comes from a contrast map: the residual intensity along the line of sight is assumed to scale with geometric distance through a homogeneous 10-km atmosphere via $S = (10/\sin h)\,r$. Synchronized two-station parallax with millisecond timing is the anchor method that gives absolute distance and velocity.

What would settle it

Point the same DJI Mavic 3T at an object of known size and range, such as a large ground target or a tracked aircraft at measured distances, and compare the resulting residual-intensity curve with the paper's graph; a mismatch between true and predicted distances would falsify the contrast-to-distance relation and with it the 88-km distance, 8-km altitude, 6-km size, and Mach 2.5 speed claimed for the combat-zone object.

Watch

Extended reading notes

Core claim

The central discovery is that dark UAPs, visible only as low-contrast shadows, can be characterized quantitatively. Over Kyiv, a parallax of 0.0464 rad places the object at 2600 km distance and 1130 km altitude, moving at 78 km/s, with a size of about 3 km and an albedo of 0.037. For the Moon video, colorimetry comparing the object to the lunar surface gives an albedo near 0.025 and sizes between 3.0 and 6.8 km. For the drone video, a contrast map against a homogeneous 10-km atmosphere places the object 88 km away, 8 km high, about 6 km wide, and moving at 806 m/s. The paper holds that all three cases agree within the measurement errors and that the objects behave like completely black bodies.

Load-bearing premise

Everything about the warzone object — its altitude, size, and Mach 2.5 speed — rests on an uncalibrated assumption that residual pixel intensity on a thermal contrast map is linearly proportional to geometric distance through a uniform 10-km atmosphere; if that relation is wrong, that entire case collapses.

Editorial extensions

If this is right

If the three cases describe one population, night-time infrared imaging becomes the reliable detection channel for dark UAPs, since visible-light observation works only in daylight.
Kilometer-scale objects flying at Mach 2.5 at 8 km altitude would be detectable by existing air-defense radar and would represent a physical presence in controlled airspace.
The 78 km/s object at 1130 km altitude is far faster than any known satellite or meteor interpretation for that event, so its confirmation would challenge conventional near-Earth object models.
A network of two synchronized wide-field cameras on a 120-km baseline could, in principle, map such objects at ranges beyond 1000 km, as the Kyiv case demonstrates.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

The uncalibrated contrast-to-distance relation is directly testable: pointing the same thermal drone camera at aircraft of known range would validate or falsify the 88-km and 75-km distances that carry the combat-zone analysis.
The Moon-case photometry assumes that YouTube-compressed video pixels preserve relative intensities; recovering the original video stream would test whether the 0.025 albedo estimate is stable.
If dark UAPs are a real class, archival all-sky surveys and weather-satellite imagery could be searched for large, slow apparent transits with low visible contrast and strong thermal signatures, a search the paper does not perform.
The consistency of the three cases could also be checked by looking for the claimed 6-by-1.5-km shape in other drone or aircraft infrared videos recorded at low elevation angles.

Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

Three new UAP observations with a genuine attempt to extract physical numbers. The Kyiv two-camera parallax is the most interesting piece: synchronized shots give a distance of 2600 km, and the albedo estimate of 0.037 follows from that geometry. The Moon video colorimetry is also a real measurement attempt. That part is worth something.

The problem is that the paper's headline claim - 3-6 km objects at Mach 2.5 - hangs on Section 4.1. There the distance to the drone-IR object is obtained from S = 10/sin(h)·r, where r is the residual intensity on a contrast map. No calibration, no radiative-transfer forward model, no error budget for that mapping. The cited 6% error covers replacing an exponential atmosphere with a homogeneous slab, not the assumption that residual intensity is proportional to path length. In the 8-14 μm band, contrast between an object and the sky depends on object temperature, emissivity, atmospheric emission, absorption, aerosols, and the camera's response. A single scalar can't become a range without validation. And the drone video itself is an uncooled microbolometer in compressed MP4; there's no plate scale, so the 40-pixel size turning into 6 km is unsupported. The Moon case uses YouTube compression as calibrated photometry; no validation against a standard star or the Moon's known colors.

There are also logical gaps. The object is called 'dark' yet the Kyiv beta=2 means it's brighter than the sky; low albedo is not the same as a dark image. The parallax velocity of 78 km/s is extraordinary, but no error bars or astrometric reduction details are given. And no false-positive analysis appears anywhere.

On the plus side, the authors are engaging with real physics - Allen's data, standard photometric equations, color transformations. The paper is not incoherent, it's just overconfident about a fragile premise. The observations might be worth reanalysis with proper calibration, but the derived parameters as presented are not credible.

I would not cite this, and I would not send it to a serious referee in this form. The load-bearing inference is too weak. A better path is to ask the authors for raw data and a validated atmospheric model; then the parallax case alone might support a shorter, more cautious paper.

Referee Report

4 major / 6 minor

Summary. The paper reports three opportunistic observations of objects it calls 'dark UAPs': a two-station daytime observation over Kyiv, a YouTube video showing objects near the Moon, and a drone thermal video from a combat zone in Ukraine. From these it derives distances, sizes, albedos, and velocities, and concludes that dark UAPs are kilometer-scale objects (about 3 to 6 km), have albedo below about 3%, and move at speeds from about Mach 2.5 up to 78 km/s. The paper further claims that the same object type was observed in all three cases, with identical characteristics.

Significance. If the claims were correct, they would imply the existence of structured, kilometer-sized, extremely low-albedo objects at altitudes from 8 km to more than 1000 km, moving at transatmospheric speeds. The paper deserves credit for attempting quantitative estimates, for using a two-station parallax approach with millisecond synchronization, and for presenting explicit formulas for size and albedo. However, the central physical conversions are asserted rather than demonstrated: the drone case uses an uncalibrated contrast-to-range relation, the Moon case treats a compressed YouTube video as photometric data, and the parallax case carries no error budget. No raw data, calibration files, or analysis code are provided. Given the extraordinary nature of the claims, the evidence presented is far below the standard needed to support them.

major comments (4)

[Section 4.1] The distance model S = 10/sin(h) * r is asserted without calibration or a forward model. In the 8-14 micron thermal band, the residual intensity on a contrast map depends on the object's temperature and emissivity, atmospheric emission and absorption along the line of sight, aerosol scattering, and the detector's spectral response; it is not simply proportional to geometric path length through a homogeneous 10 km atmosphere. The cited 'error of no more than 6%' refers to the homogeneous-atmosphere approximation for a known optical depth, not to the validity of the r-to-distance proportionality. Because the derived 88 km and 75 km distances set the object's 8 km altitude, 6 km width, and roughly 800 m/s velocity, this unsupported mapping is load-bearing.
[Section 3] The Moon case treats pixel values from a compressed, re-encoded YouTube video as calibrated photometry. The albedo alpha_obj = 0.025 is obtained from RGB intensity ratios between the Endymion crater and an object, with no flat-fielding, linearity check, compression-artifact analysis, or uncertainty estimate. Likewise, the pixel scale derived from Endymion crater is applied to objects assumed to be at the Moon's distance without an independent astrometric reduction. The statement that 'a paper was recently published that attempts to prove the original video is a fake' is not accompanied by a citation or a rebuttal. This case cannot independently support the claimed albedo or size.
[Section 2] The parallax measurement for the Kyiv object is reported without an error budget. The distance of 2600 km, altitude of 1130 km, and linear velocity of 78 km/s follow from a single parallax angle of 0.0464 rad and an angular velocity, with no plate solution, no treatment of timing or refraction errors, and no cross-check from the second station's independent trajectory. The object's size estimate of 3 +/- 0.4 km is based on a 7 +/- 1 pixel extent against a 2 x 2 pixel PSF, which is only marginally resolved; the quoted uncertainty does not include PSF-subtraction or deconvolution errors. The albedo estimate from Eq. (5) inherits these unquantified errors and further assumes reflected sunlight.
[Section 5] The Discussion's central claim that the Moon and drone objects have 'identical characteristics' is internally inconsistent. The drone case yields about 6 km width and Mach 2.5 at about 8 km altitude above Earth, whereas the Moon case object is 6.8 x 3.5 km, travels at about 31 miles per second (roughly 50 km/s), and its 8.7 km altitude is above the lunar surface, not above Earth. These are not the same quantitative characteristics, and no physical argument is given for why objects near Earth's surface, in near-Moon space, and at 1130 km altitude should belong to one population. The cross-case claim therefore does not follow from the presented measurements.

minor comments (6)

[Section 2.1] Equation (1) uses r for the object's size, while Section 4.1 uses r for residual intensity; this notation collision makes the distance formula S = 10/sin(h) * r confusing.
[Section 3.1] The color transformations in Eqs. (6) and (7) are taken from an external arXiv paper without stating whether they are valid for the camera and video pipeline used by Carlotto, and the resulting (B-V)_J and (V-R)_J values have no uncertainties.
[General] The paper states that the authors analyzed the Carlotto video but provides no link to the original file, no calibration frames, no timestamp metadata, and no analysis code; raw data and code should be made available if these claims are to be checked.
[Section 2] The sentence 'Practice shows that fireflies are visible only in the daytime sky' appears to be a typo or an unclear translation; as written it is confusing.
[Figures 11 and 12] The axes of Figures 11 and 12 lack units, and 'residual intensity' is not defined in the text; it is unclear whether this quantity is a normalized pixel value, a contrast, or a radiance ratio.
[Section 4.2] The comparison of the object to Laputa in Gulliver's Travels is literary rather than scientific and should be removed from a research paper.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the derivation chain uses independent parallax, lunar-calibrated photometry, and a stated (if unvalidated) atmospheric contrast model; no equation reduces to its own input.

full rationale

Walking the paper's derivation chain, the quantitative claims do not reduce to their inputs. For the Kyiv case, the distance is obtained from a two-station parallax with millisecond synchronization, the angular size is pixel-resolved against the PSF, and Eq. (5) derives albedo from the measured contrast ratio, the sky brightness, and the geometric distance/size ratio, with the geometric ratio coming from the independent parallax and angular-size measurements. For the Moon case, the albedo is derived from a contrast ratio against the known lunar albedo (0.067) and sizes are calibrated with the Endymion crater, an external image benchmark; the color transformation is checked against published lunar (B−V) values. For the drone case, the distance is obtained from the explicitly stated homogeneous-atmosphere model S = 10/sin(h)·r, with the 6% error estimate quoted from Allen [1]; the coefficient is an assumed atmospheric scale height, not a parameter fitted to the target size or velocity. The statement that the drone object 'does not emit' is an interpretive premise used to justify the contrast-to-distance mapping, not a quantity derived from that mapping, so the later black-body language is an unsupported generalization rather than a circular derivation. Reference [9] is a self-citation to the authors' prior UAP classification, but it is not load-bearing: none of the sizes, velocities, altitudes, or albedos depend on values imported from [9]. The main validity threat is that the contrast-to-range relation and the compressed video photometry are uncalibrated and lack an error budget; that is an input-validation and correctness-risk issue, not circularity by construction.

Assumptions & free parameters 3 free parameters · 4 assumptions · 0 invented entities

The central claims rest on several unvalidated domain assumptions, especially the thermal contrast-to-distance law, plus a handful of chosen parameters such as the 10 km homogeneous atmosphere. No new physical entity is introduced.

free parameters (3)

Homogeneous atmosphere scale height H = 10 km
Used in S = 10/sin(h) * r to convert residual intensity to distance in Section 4.1; the 88 km/75 km distances, 8 km altitude, 6 km size, and Mach 2.5 velocity all scale with this chosen value.
Sky/object brightness contrast beta (Kiev case) = 2
Taken from the color map in Fig. 4; enters Eq. 5 and sets the albedo 0.037.
Moon/object brightness contrast beta (Moon case) = 2.65
Taken from RGB intensity estimates in Section 3; determines albedo 0.025.

assumptions (4)

domain assumption Objects reflect sunlight as diffuse, uniformly bright surfaces with a single albedo, and the sky background is given by Allen's tabulated values
Used in Equations (1)-(5) and in Section 3 to convert brightness ratios into albedo and size.
ad hoc to paper The YouTube video of the Moon (Carlotto) shows real physical objects transiting the lunar surface and its pixel values are usable as photometry
Section 3 asserts authenticity after 'calculations' but provides no calibration, no raw video, and no comparison with the fake-analysis paper it mentions.
ad hoc to paper Thermal infrared contrast in the 8-14 micrometer drone image is dominated by atmospheric scattering, and residual intensity decreases linearly with distance through a homogeneous 10 km atmosphere
Section 4.1 uses this to derive distances of 88 and 75 km; no independent calibration or radiative transfer calculation is given.
domain assumption The distance and velocity from parallax are correct despite lack of detailed astrometric reduction
Section 2 states a parallax of 0.0464 rad and one-millisecond synchronization, but no raw measurements or uncertainty are shown.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Unidentified Aerial Phenomena. Characterization of Dark UAPs." pith.science (2026). https://pith.science/paper/UB3EOPOW

@misc{pith2026250305627,
  author       = {Pith},
  title        = {Pith review of: Unidentified Aerial Phenomena. Characterization of Dark UAPs},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/UB3EOPOW}},
  note         = {Machine review of arXiv:2503.05627}
}

read the original abstract

We use high-tech observations of Unidentified Aerial Phenomena (UAP) class objects to evaluate their characteristics. We present data in three cases. (1) Multi-side daytime observations of UAPs over Kiev. (2) Night observations of a group of objects in the vicinity of the Moon. (3) UAP observations in the combat zone in Ukraine. Dark UAPs in the visible wavelength range are observed only during the day. At night they can only be seen in the infrared wavelength range. We note large sizes of UAPs, from three to six kilometers.They exhibit large velocities, from 2.5 Mach and much larger. They have low albedo, from three percent and below, that is, they actually exhibit features of a completely black body.

Figures

Figures reproduced from arXiv: 2503.05627 by the authors.

**Figure 3.** Image of the object. 0 10 20 30 40 50 0.8 1 1.2 1.4 1.6 1.8 2 #10 B G R [PITH_FULL_IMAGE:figures/full_fig_p002_3.png] view at source ↗

**Figure 5.** Frame from original video. 920 940 960 980 1000 1020 1040 300 310 320 330 340 350 360 370 380 0 50 100 150 200 250 [PITH_FULL_IMAGE:figures/full_fig_p004_5.png] view at source ↗

Figures from the paper (3 more)

**Figure 7.** Figure 7: The color diagram of the object in the RGB [PITH_FULL_IMAGE:figures/full_fig_p004_7.png]

**Figure 9.** Figure 9: A typical frame of the video. pixel pixel 10 20 30 40 50 60 70 10 20 30 40 50 60 70 80 420 440 460 480 500 520 [PITH_FULL_IMAGE:figures/full_fig_p005_9.png]

**Figure 11.** Figure 11: Contrast map at the beginning of the ob [PITH_FULL_IMAGE:figures/full_fig_p006_11.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

9 extracted references · 9 canonical work pages

[1]

W., 1963, Astrophysical Quantities, 2d ed., Lon don, Athlone Press

Allen C. W., 1963, Astrophysical Quantities, 2d ed., Lon don, Athlone Press

work page 1963
[2]

Carlotto, https://www.youtube.com/watch?v=ScBx 2EwSuDo 6 7

M.J. Carlotto, https://www.youtube.com/watch?v=ScBx 2EwSuDo 6 7

work page
[3]

Carlotto, 1995, Journal of Scientijc Exploration, Vol

M.J. Carlotto, 1995, Journal of Scientijc Exploration, Vol. 9, No. 1 , pp. 45-63

work page 1995
[4]

https://www.instagram.com/reel/C3vGfLDr-HB/?igsh= MWZweDV3c3BqczE5Ng

work page
[5]

https://alpha-photonics.com/en/produkte/dji-mavi c-3-thermal-en/

work page
[6]

Parka, S

W. Parka, S. Paka, H. Shimb, et al., 2015, arXiv:1501.047 78v3 [astro-ph.IM] 2 Sep 2015

work page 2015
[7]

Thejll P., Flynn C., Gleisner H., et al., 2014, A&A, 563, A 38

work page 2014
[8]

Zagury F., 2012, The Color of the Sky, Atmospheric and Cli mate Sciences, 2012, 2, 510-517

work page 2012

Show all 9 references

[9]

Evaluation of UAP properties, DOI:https://doi.org/10.48550/arXiv

Zhilyaev B.E., Petukhov V.N., Reshetnyk V.M., 2022, Unid entiﬁed aerial phenomena II. Evaluation of UAP properties, DOI:https://doi.org/10.48550/arXiv. 2211.17085 7

2022 doi arXiv

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.

[1] [1]

W., 1963, Astrophysical Quantities, 2d ed., Lon don, Athlone Press

Allen C. W., 1963, Astrophysical Quantities, 2d ed., Lon don, Athlone Press

work page 1963

[2] [2]

Carlotto, https://www.youtube.com/watch?v=ScBx 2EwSuDo 6 7

M.J. Carlotto, https://www.youtube.com/watch?v=ScBx 2EwSuDo 6 7

work page

[3] [3]

Carlotto, 1995, Journal of Scientijc Exploration, Vol

M.J. Carlotto, 1995, Journal of Scientijc Exploration, Vol. 9, No. 1 , pp. 45-63

work page 1995

[4] [4]

https://www.instagram.com/reel/C3vGfLDr-HB/?igsh= MWZweDV3c3BqczE5Ng

work page

[5] [5]

https://alpha-photonics.com/en/produkte/dji-mavi c-3-thermal-en/

work page

[6] [6]

Parka, S

W. Parka, S. Paka, H. Shimb, et al., 2015, arXiv:1501.047 78v3 [astro-ph.IM] 2 Sep 2015

work page 2015

[7] [7]

Thejll P., Flynn C., Gleisner H., et al., 2014, A&A, 563, A 38

work page 2014

[8] [8]

Zagury F., 2012, The Color of the Sky, Atmospheric and Cli mate Sciences, 2012, 2, 510-517

work page 2012