The Black Triangle of the Seventh Tranche: Two Papers That Don’t Match, 1,800 mph Nobody Measured and a Second Night That Is Missing
The FBI and AARO describe the same black triangle an hour apart, three seconds versus more than thirty, and they disagree on whether it ever stopped moving. The 1,800 mph comes from misreading a distance, “no thermal signature” is the opposite of what the papers say, and the second night the official entry announces is not in the PDF.
On October 23, 2026 the seventh PURSUE tranche published two papers on a single June 2026 sighting: the statement of two special federal agents transcribed by the FBI (FBI-UAP-D043) and the memorandum of the AARO interview with them (DOW-UAP-D176). They describe a matte black triangle that rises vertically over a ridge, without noise, and is lost from sight. They disagree on the time, the duration, the size and the hovering. Nobody measured the speed.
Executive summary of the case file. Everything essential before you dig deeper.
Fact sheet
OSINT - Traceability
“Thermally elevated aerial object.” In plain language: something that shows up hotter on a thermal camera than what surrounds it. That is how a special agent of the U.S. Government wrote it in a statement that the FBI transferred to its FD-302 form and that the Department of War published on Friday, August 7, in the fifth PURSUE tranche, as FBI-UAP-D033.
He saw it around nine at night, “in the same general western sector and along a similar heading as the object observed the previous night.” His paper does not tell that previous night. Since Friday, October 23, I think I know what it was: the seventh tranche brings a black triangle, an official entry that speaks of two days and a PDF with room for only one.
Back to my own business, friends: papers, clocks and marker pen. I separate what the documents say from what I say, and I warn you every time I speak in Astro language.
The verifiable first: one triangle, two papers
The first is FBI-UAP-D043, another FD-302, the form on which the FBI records interviews. This one contains no questions: it copies “verbatim” a written statement signed by two special agents, with date, place and names blacked out. It is two pages of text and nine of reenactment: the same photo of low scrub-covered hills under a blue sky, with a black silhouette pasted in a different spot on each sheet. There is not a single photograph of the object.
The repeated photo was reported on Saturday by IBTimes UK, based on a report from NHI Anomalous; I have checked it sheet by sheet. A clerk’s nuance: the portal entry calls “interview” what the paper presents as a written statement.
The second is DOW-UAP-D176, a memorandum for the record from AARO, the Pentagon office that investigates these cases. Three pages. It records the interview with two agents (SOURCE 1 and SOURCE 2), who by what they say and by the portal’s pairing are the ones who sign the statement, and the account of a third person, SOURCE 3, who from another observation post detected the object first, with long-range thermal equipment. Before telling anything, the memorandum covers itself:
“This report reflects the reporter’s interpretation at the time of the event.” AARO writes it itself, in the first line of its summary.
DOW-UAP-D176, section “Overview”
What the two accounts say, in short: at dusk, the agents run across to the other side of their hill, see a matte black triangle rise vertically over a ridge, without noise, and lose sight of it as it descends toward the southeast (according to the FBI) or into a canyon (according to AARO). The next morning, AARO says, the canyon was searched: no debris, no marks on the ground, no damaged vegetation.
Eight differences in one sky
Set side by side, sentence by sentence, the two papers do not tell the same story. Some differences are matters of nuance. Others are not.
Wingspan: 9 to 15 feet (2.7 to 4.6 m).
The ridge, about 3.2 miles (5.1 km) to the west.
Rises about 200 feet (61 m) above the ridge.
Hovering: does not appear.
Total duration: “approximately three seconds.”
Leaves “approaching our position” and descends toward the southeast; they lose it at about 1.5 miles (2.4 km).
They compare it to a “delta wing drone” and call it “unmanned aerial system.”
Wingspan: 13 to 15 feet (4 to 4.6 m).
Distance bounded “between approximately 1.4 and 3.2 miles.”
Rises 200 to 300 feet (61 to 91 m).
Hovering for about 20 seconds.
Phases: about 10 s of ascent, 20 stopped and 2 or 3 of departure.
Turns, tilting, and descends hugging the terrain inside a canyon.
Calls it “unidentified craft.”
The duration had already been pointed out by IBTimes UK, also from NHI Anomalous: three seconds versus more than thirty. The one that stops me is the fifth. The hovering, that going from rising to being still in the air and from being still to flying like a fighter with no acceleration ramp, is what turns the account into an anomaly. And it appears in the AARO interview, not in the statement the two agents signed.
I don’t know which of the two papers is more faithful, or which was written first: the FD-302 has its date blacked out and the memorandum shows none. I know that the heart of the anomaly beats in only one of them, and that this has to be said before arguing about propulsion.
The heat: the opposite of what has been published
The Turkish agency Anadolu wrote the same Friday that the object’s thermal signature “could not be detected by thermal equipment.” Avi Loeb, who chairs the scientific advisory board on UAP (the one from the PR135 birds), summed it up on his blog with a “showed no heat source.” The two papers say something else, and they say it clearly.
Hotter than ambient and uniform is not “no thermal signature.” It is a thermal signature without hot spots, which is not the same: it says no exhaust is visible, not that nothing is visible. The object was, in fact, found by its heat. CBS reported it correctly on Friday afternoon, with the exact quote.
The 1,800 miles per hour nobody measured
The number that has gone around the world is another one. The New York Post published on Friday at 8:02, New York time, that black triangles were “speeding up to 1,800 miles per hour,” and explained it this way: “The objects were observed moving at a rate of 1.5 miles in three seconds, which would be 1,800 mph, observers noted in an FBI report on the matter.” The figure is also in UNILAD, which cites the Post, and in TRT World.
The FBI paper does not say that. It says they lost sight of the object “at an estimated range of about 1.5 miles from our position,” that is, when it was about 1.5 miles from them. That is a distance to the observer, not a path traveled. And the three seconds are “the total duration of the sighting,” the whole sighting, ascent included. Dividing one by the other gives 1,800, yes. And it means nothing.
There is another calculation that should have stopped someone. 1,800 miles per hour is about 2.4 times the speed of sound at sea level, which the Air Force rounds to “about 750 miles per hour.” Whatever flies faster than sound keeps dropping sonic booms, and both papers speak of silence. One or the other.
0measured
“No instrumented speed measurement was obtained” (AARO)
3s
the whole sighting, ascent included
32-33s
10 ascending, 20 still, 2 or 3 departing
0
the previous six included video
Nobody measured the speed, as AARO says in the first box. The agents estimated it from how far the object advanced along the ridge before they lost sight of it, and compared it to that of a fighter.
The Post adds that there were triangles “on at least three separate occasions over the past year” and presents the AARO memorandum as “another triangle sighting.” It is not: the portal pairs the D176 with the D043, and the entries for both speak of two special federal agents at a national security site in 2026. I have not been able to reconstruct where the three occasions come from.
The Post also states that some officials ruled out that it was U.S. technology, without saying who or where. That is not in the D176, nor in the D043, nor in their entries. This paper has made me open the PDFs before: with the helicopter orbs it illustrated the story with reenactments from another case.
The time: 19:23 or 20:23
Now the difference I have not seen reported anywhere. The FBI writes “approximately 1923 hours.” AARO writes “2023L,” twice, with an L the memorandum does not explain. Exactly one hour apart, which is just what separates daylight saving time from standard time, or two neighboring time zones. My reading: someone converted an hour wrongly. The paper does not say who.
Which one fits the sky? AARO describes “late-dusk illumination”: the Sun already below the horizon and a remnant of sunset light outlining the western ridges. My own calculation, with NOAA’s solar position formulas: the Sun’s altitude at 19:23 and 20:23, local time, in 29 cities of the western U.S., on June 1, 15 and 30, 2026. That makes 87 combinations.
- At 19:23 the Sun was still above the horizon in 82 of the 87. It had set, or was setting, only in southeastern Arizona: Tucson and Springerville at the beginning of the month, Douglas all month.
- At 20:23 it was below in 52 of the 87 (one of them, Reno in mid-month, with the Sun right on the horizon line): southern and central California, Nevada, Arizona, New Mexico, El Paso and, at the beginning of the month, Denver and Cheyenne. In the north (Seattle, Boise, Salt Lake City, Helena) it had not yet set.
- The calculation does not include refraction, which makes the Sun visible a little longer. That only worsens the fit of 19:23.
- Nor does it include terrain, which works the other way: a high range to the west hides the Sun before the textbook sunset, and that could indeed save 19:23 in some valley. The agents, according to AARO, were on top of a hill.
Translated: with a clear horizon, the memorandum’s sky is the 20:23 one in the southern half of the West, and the 19:23 one only in a corner of Arizona. I have not tried to locate the site and I will not: according to the entry, it is a sensitive national security place. What I am saying is more modest. One of the two clocks is wrong and, with what there is, I lean toward the FBI’s.
The second night that is missing
The official entry for the memorandum, the text that accompanies it on the portal and in its data sheet, says it records observations “over two days in 2026” and that it “further records a secondary observation on the following evening.” It also mentions “associated analytic support material.” I have read the three pages of the PDF from top to bottom. There is no second night. There is no analysis material. It ends at “REPORT PREPARED BY” and a redacted name.
Either the entry describes a longer document than the one published, or it describes another one.
The August 7 paper
In the D033, an agent “positioned in the same general observation area referenced in my prior statement” sees, around 21:00 (the portal entry specifies “local time”), a hot object in the same sector, “flying southbound, consistent with the general direction of travel observed during the prior night’s sighting.” He places it about two ridges farther west “than the location where the” [black box] “object was observed.” The adjective for that object, if it was one, is blacked out.
He could not make out its shape because of the distance. He also filmed an Army Black Hawk flying through the area, on purpose, to have a size reference, and estimated the object was “approximately half the size of the Black Hawk,” “an estimate only.” He could not say whether it was fixed-wing, rotor “or other unmanned aerial system.” The coincidences with the triangle, one by one:
- The same heading in both FBI statements: “Unidentified Aerial Object Observation Report evening of,” followed by the redacted date.
- The same expression: the D043 opens with “a detection for a thermally elevated aerial object”; the D033 carries that title.
- Western U.S. and 2026, according to the portal, in both cases.
- A previous night in one; a second night announced, and absent, in the other.
- The heading: toward the southeast in the D043, toward the south in the D033, “consistent with” the previous night.
My reading, and I mark it as such: the D033 is the triangle’s second night, and it came out eleven weeks before the first; the PDF metadata, in the timeline below, tell the order. No paper says it is the same case. If the complete memorandum appears and it is not the D033, I will correct it here, with a date.
The August tranche brings more FBI papers on the West in 2026, but none serves as a previous night: the D032 is from four in the afternoon, and the D037 and the D040 tell of red lights seen by two people whom the entry does not call agents.
There is one detail that doesn’t fit, and I put it on the table before someone else does. Half a Black Hawk is about 25 feet if you take the fuselage and about 32 with the rotors, 7.6 to 9.9 meters. The first night’s triangle had a wingspan of 9 to 15 feet, 2.7 to 4.6 meters. The D033’s would be at least 1.7 times larger. Two ridges away and by eye, that rules nothing out. It confirms nothing either.
The marker-pen law
The 2024 defense act is enacted; its section 1843 sets the grounds for deferring publication of UAP papers.
Source ↗The first night
Two special federal agents see a black triangle over a ridge. The FBI notes 19:23; AARO, 2023L. Day and place redacted.
Source ↗The second night
According to the D176 entry, another observation. The D033 tells of a hot object at 21:00, two ridges farther west. That it is the same series is my own reading.
Source ↗1843 shows up
A Navy report from the fourth tranche (D089) already cites section 1843 on its release sheet.
Source ↗The D033 PDF is generated
Creation date in the file metadata.
Fifth tranche
The D033 comes out, with the previous night untold.
Source ↗The D043 and the D176 are generated
Creation dates in the metadata of the two files.
Seventh tranche
28 PDFs, no videos. The D176 and the D043 come out paired.
Source ↗The 1,800 mph
The New York Post publishes the figure at 8:02, New York time, and others repeat it.
Source ↗This reading
The three papers, the portal CSV and the press cross-checked.
The marker pen of section 1843
The marker pen, which is my thing. The redactions in the D176 cite “1843(a)(2),” “1843(a)(3)” or “1843(a)(1)(C)”: section 1843 of the defense act for fiscal year 2024 (Public Law 118-31), the one that sets the grounds for deferring publication of a paper on UAP. I already went through that law with the National Archives’ empty entries. The National Archives ask agencies that, if a paper comes out redacted, they cite those grounds or their declassification guidance.
- 1843(a)(1)(B): would reveal an intelligence source or method in use and not officially disclosed.
- 1843(a)(1)(C): any other defense, intelligence or foreign policy matter whose release would harm national security “demonstrably and substantially.”
- 1843(a)(2): releasing it would violate the Privacy Act of 1974.
- 1843(a)(3): it would be an unwarranted invasion of privacy, so serious that it outweighs the public interest.
- In the D176 there are 22 codes: 10 of (a)(2), covering the witnesses; 6 of (a)(1)(C), among them those on the date; 4 of (a)(3), covering the interviewers and whoever wrote it; 2 of (a)(1)(B).
It is not entirely new. A Navy report published on July 10, the D089, already had a final sheet with “FY2024 NDAA § 1843(a)(1)(C)” and “§ 1843(a)(3).” What is new is that now there is nothing else.
In the other papers from earlier tranches that I have opened (the D010 from May, the D101 from August, and the D106, the D108, the D109, the D111 and the D112 from September), the labels cite FOIA exemptions, the freedom of information law. In the seventh, the ten Department of War documents with redactions cite 1843, and none cites FOIA. It is a sample, not a census. And the FBI goes its own way: in the D043 it blacks out without saying why.
The same section, in its subsection (c), obliges the head of the agency that defers to notify congressional leadership and its oversight committees within 15 days, with the reason. Those notices are not on the portal, and as of October 25 I have found none published. The law does not require publishing them either. If anyone has them, my folder of FOIA requests has room.
A dead man shielded by the Privacy Act
The Privacy Act does not protect the dead: “Deceased individuals do not have any Privacy Act rights, nor do executors or next-of-kin,” sums up the Department of Justice guide. And I don’t know what intelligence source or method in use is protected by redacting the name of a 1947 press chief in a text that came out in a book nineteen years ago and that The Register was already quoting in July 2007, in the same words.
With the same codes, the officers it names are redacted, among them Blanchard, Marcel and Ramey, dead in 1966, 1986 and 1963. This time the marker is not even black: they are white boxes with a border.
On Friday, on NewsNation, the statement was presented as part of what was released that day, according to Loeb’s transcript. It is. New it is not. Where it came from and who wrote it is told by its Wikipedia entry, and what really happened in 1947 was sorted out by Rogelio in his Roswell case file.
The marker used to cite the freedom of information law. Now it cites the UAP one. That is something. What’s left is for someone to read it whole, starting with the part about the living.
What it could be (no Martians required)
The first hypothesis was put forward by the agents themselves in their statement: “most resembling a delta wing drone,” and at the end they call it “the unmanned aerial system.” A fixed wing that takes off vertically is already on the market. And drones over sensitive installations are not exotic: I told here about the NRC’s 67 reports on nuclear plants. The objection, according to AARO, comes from the witnesses: from that hill, even a small drone would be heard at 1.4 miles.
But 1.4 miles is the near limit. AARO bounds the distance between 1.4 and 3.2 miles, and size, speed and silence depend on that figure. My own calculation: an object of 13 to 15 feet takes up 2.6 to 7 arcminutes of sky, between one twelfth and somewhat less than a quarter of the diameter of the full Moon. A 30-centimeter rotor at 1.4 miles, less than half an arcminute.
The agents estimated the wingspan by comparing the silhouette with the trees on the ridge, which is a good idea if the object was on the ridge. If it was closer, it was smaller. Second hypothesis: something of our own. The Post places the scene over “top secret sites where the US government tests new technology,” which the papers do not say.
Third: the dusk, the rush, the switch to the other side of the hill and a pair of binoculars that, according to AARO, passed from hand to hand. Three seconds, the length of the sighting according to the FBI, is enough to see a shape, not to measure a maneuver.
Triangles with a history
Black triangles have a history, and not all of it good: Rogelio told how the most famous photo of the Belgian wave turned out to be a piece of polystyrene foam. In PURSUE there were already triangles in the West, two from the FBI in Colorado Springs, from 2023, also published on August 7.
According to the entry, AARO sees a resemblance to other cases in the region, without saying which. Close at hand is the “Western US Event” of October 2023, which the portal lists among AARO’s most compelling: special federal agents, two days, a sensitive site in the West, orbs and two kite shapes.
What would clear it up is the full version of the D176, with its second night and its analysis material; the thermal recording, if there was one; and the Black Hawk video that the D033 agent filmed as a reference, rather than a hangar. None of that is in this tranche. The marker pen, meanwhile, has changed law, and on the Roswell paper even color.
OSINT – Source traceability
V3 and not V1/V2: the documents are official and verifiable, but what they describe is testimonial, with no published sensor data, and the two papers contradict each other. Own reading, marked in the body: that the one-hour difference comes from a faulty conversion; that the wrong clock is the FBI's (own solar calculation with a clear horizon, deliberately without locating the site); that the D033 is the D176's second night; the change in redaction codes (sample of eight earlier documents, not a census; one from July, the D089, already cited 1843). No attempt was made to geolocate the place.