The Roosevelt-associated aircrew record extends beyond the two familiar infrared clips. Redacted VFA-11 hazard forms describe contacts during East Coast training in April 2014, before the January 2015 recordings subsequently released as Gimbal and GoFast. These records provide specific dates and distinctions between sensor tracking and visual acquisition. They do not supply a complete named roster or establish that the earlier contacts and later recordings share an identity. This collective review covers the supporting documentary history; Ryan Graves's separately attributable testimony remains in his own dossier.
Supporting naval aircrew and squadron reporting record across East Coast training and the Roosevelt-era recordings
Three April reports and the January recordings
A Naval Safety Center hazard report identifies an event on 23 April 2014 involving VFA-11 F/A-18F aircrew in Warning Area W-72. Its narrative describes two radar contacts with displayed speed of zero Mach, at approximately 15,000 and 12,000 feet, and says that infrared observations persuaded the crew that the contacts were not false radar files. It also reports two additional objects crossing the infrared field of view between the aircraft and the second radar contact. Their described high speed is a crew interpretation in the report; no time-resolved trajectory accompanies the released form.
The 24 April report describes two aircraft interrupting basic fighter manoeuvring to investigate a stable unidentified radar track at about 11,000 feet. The lead aircraft approached roughly 200 feet below the indicated contact while the second remained about a nautical mile away. Both obtained a CATM-9X sensor track, according to the narrative, but neither achieved visual identification. This is a reported radar and additional-sensor encounter with an explicit visual limit, rather than a close appearance description.
The released four-page summary for 27 April records a near-midair encounter with a balloon-like object in W-72. It classifies the risk as critical and records recommendations to brief relevant units, alongside the commander's description of repeated occurrences. It does not preserve a detailed cockpit narrative, a cube-and-sphere description or an exact clearance. It cannot therefore be identified as the flight later described by Graves on the basis of this summary alone.
The Department of Defense officially released three historical Navy videos on 27 April 2020, dating two recordings to January 2015 and describing their subjects as unidentified. These later clips belong to the public history of Gimbal and GoFast. NASA's independent study team used GoFast as a geometric example in September 2023; AARO published its more detailed case analysis in February 2025. The latter addresses the movement of the target in that clip, rather than the full sequence of squadron reports.
Evidence review
The April forms document what the squadron recorded as a training hazard. Their print timestamps are 29 May 2019, while the event fields specify April 2014. The copies inspected here were published by a FOIA recipient and redact personal identities. They preserve original report wording, but do not establish the exact filing date or permit the aircrew to be assigned to particular named public witnesses. Neither the report count nor the number of sensor channels supplies a verified count of independent observers.
The crew's assessment that infrared observation confirmed a radar contact is evidence about its identification process. The released forms contain neither the underlying radar history nor synchronised sensor recordings. Their additional sensor observations give the reports more detail than an isolated unexplained display return, while leaving target association and displayed speed available for testing only through the narrative supplied. The explicit failure to obtain visual identification on 24 April limits comparison with reports containing an appearance description.
The forms' safety categories describe the perceived consequences of unknown traffic in training airspace. Balloon-like appearance in the 27 April summary is a description, not a recovered object's identification. Conversely, a potentially ordinary object would still present a collision concern. The documents support a history of reported hazards and briefing recommendations without establishing a common operator, unusual propulsion or one recurring physical design.
All five pages of the original redacted Naval Safety Center form read in text and rendered images. FOIA-recipient-hosted reproduction, not a direct Navy archive download. WESS 1398374727732; local report 09-14. The event date and 29 May 2019 print timestamp are distinct.
All five original rendered pages read, with narrative and visual-acquisition wording checked. WESS 1398374605670; local report 10-14. Released names are redacted; no underlying sensor recording was obtained. The source explicitly distinguishes additional sensor tracking from failure to acquire a visual identification.
All four original rendered pages read. WESS 1399061379010; local report 11-14. This released summary uses balloon-like appearance and critical-risk wording, but supplies no detailed crew account identifying it with the cube-and-sphere story.
Recipient's release page acquired and its original PDF links followed. Source provenance is a private reproduction of released Navy forms; no direct Navy-hosted copy or official-file hash comparison was completed in this review.
Complete short original statement of 27 April 2020 read. It dates two videos to January 2015 and describes the subjects as unidentified at release. Direct NAVAIR video downloads were inaccessible during this review; no claim of complete video or audio inspection is made.
Final report's printed page 28, PDF page 30, read and inspected as an original rendered page. The example describes parallax and a roughly 40 mph estimate, explicitly neglecting aircraft wind effects. The report as a whole was not fully inspected for this dossier.
Original 26-page publication acquired. Narrative and methodology text read; selected consequential pages 1–7, 17–18, 22 and 24–26 inspected as rendered originals. The report distinguishes ground speed from wind-compensated speed and records missing original metadata, unknown aircraft heading and unavailable aircrew accounts. No independent recomputation of its model was performed.
Separately attributable person dossier. His account of the cube-and-sphere near-pass is a report received from another pilot, and is not assigned to the redacted 27 April summary without an identifying source.
What the hazard records and GoFast reconstruction establish
The available documentation breaks the collective into events rather than a uniform block of corroboration. Two April reports concern contacts pursued through aircraft sensors; a third provides only a near-midair summary. The January recordings are later events. Shared unit and operating context explain why they entered one public narrative, but do not supply the missing trajectory or identity link between them. Redacted participants cannot be counted as additional named witnesses independently of Graves without further source identification.
GoFast demonstrates why apparent movement across an infrared background is not a direct measure of target speed. A moving aircraft changes the viewing angle to a target between it and the sea, making the background sweep past even when the target moves much more slowly. NASA's 2023 example used the displayed range, camera angle and aircraft altitude to place the target at about 13,000 feet, well above the water. It estimated roughly 40 mph over a 22-second interval, while explicitly acknowledging that it neglected wind effects on the aircraft. That estimate is conditional geometry, rather than a measured track showing a fast object skimming the ocean.
AARO's February 2025 analysis adds a reconstruction of the aircraft's turning path and historical winds at the aircraft and estimated target altitudes. Its principal calculation uses a thirteen-second segment with nearly constant displayed altitude, bank and airspeed. Because aircraft compass heading is unknown, it examines possible directions relative to the wind. The resulting ground-speed range is approximately 72–161 mph; after removing the modelled wind contribution, the report gives about 5–92 mph relative to the surrounding air. The latter is an air-relative modelled range; its upper end includes substantial motion relative to the air, so the model does not imply passive drift in every scenario.
The reconstruction's limits are part of its result. AARO says that the original file and accompanying metadata were no longer available, that exact aircraft location and heading were unknown and that it was unable to obtain the recording aircrew's accounts. It extracted values from the compressed public video and estimated a continuous path from rounded display readings. Historical weather information for the general area was introduced because atmospheric observations were not supplied with the report or video. Its appendix cautions that quantitative speeds should be used qualitatively because of uncertainty in the inputs and flight assumptions. These are modelled bounds, not a recovered navigation track or an independent on-site wind measurement.
Within that reconstruction AARO assesses with high confidence that GoFast did not demonstrate anomalous performance, while leaving the object's identity undetermined. The two conclusions address different propositions. The geometric work substantially weakens an extraordinary-speed interpretation of this recording without identifying its target as a particular balloon, bird or aircraft. Its scope also stops at GoFast: the analysis does not settle Gimbal's target identity, establish that an apparent infrared shape is a resolved airframe or classify every earlier squadron contact.
The reporting history therefore has a firmer documentary basis than a general claim about extraordinary craft. Original hazard forms preserve the squadron's concern and some distinctions between observation channels; the released clip permits a later technical test of one speed impression. The missing named encounter statements and underlying mission data matter because they would connect a particular crew's perception to a particular measured event. The available reports describe separate events with different viewing access and investigative outcomes, without a demonstrated common trajectory or identity.
Key dates
23–27 April 2014
Three VFA-11 event records describe different training-area contacts and hazards.
January 2015
The two later Navy recordings are dated to this month in the official release statement.
29 May 2019
The inspected hazard-report copies bear this print date; their event dates remain April 2014.
27 April 2020
DoD publishes the clips and confirms Navy provenance, without identifying their targets.
September 2023
NASA's final study report presents a GoFast parallax example with an explicit wind-related limitation.
February 2025
AARO publishes its GoFast case card and methodology; an editor's note dated 20 February records unsuccessful efforts to obtain aircrew accounts.
All five pages of the original redacted Naval Safety Center form read in text and rendered images. FOIA-recipient-hosted reproduction, not a direct Navy archive download. WESS 1398374727732; local report 09-14. The event date and 29 May 2019 print timestamp are distinct.
All five original rendered pages read, with narrative and visual-acquisition wording checked. WESS 1398374605670; local report 10-14. Released names are redacted; no underlying sensor recording was obtained. The source explicitly distinguishes additional sensor tracking from failure to acquire a visual identification.
All four original rendered pages read. WESS 1399061379010; local report 11-14. This released summary uses balloon-like appearance and critical-risk wording, but supplies no detailed crew account identifying it with the cube-and-sphere story.
Recipient's release page acquired and its original PDF links followed. Source provenance is a private reproduction of released Navy forms; no direct Navy-hosted copy or official-file hash comparison was completed in this review.
Complete short original statement of 27 April 2020 read. It dates two videos to January 2015 and describes the subjects as unidentified at release. Direct NAVAIR video downloads were inaccessible during this review; no claim of complete video or audio inspection is made.
Final report's printed page 28, PDF page 30, read and inspected as an original rendered page. The example describes parallax and a roughly 40 mph estimate, explicitly neglecting aircraft wind effects. The report as a whole was not fully inspected for this dossier.
Original 26-page publication acquired. Narrative and methodology text read; selected consequential pages 1–7, 17–18, 22 and 24–26 inspected as rendered originals. The report distinguishes ground speed from wind-compensated speed and records missing original metadata, unknown aircraft heading and unavailable aircrew accounts. No independent recomputation of its model was performed.
Separately attributable person dossier. His account of the cube-and-sphere near-pass is a report received from another pilot, and is not assigned to the redacted 27 April summary without an identifying source.