← Back to Files & Wiki
Wiki page · asset · graph_investigation_draft

DOW-UAP-PR26, Unresolved UAP Report, United Arab Emirates, October 2023

Open Sky release file copy: open the verified 43 second MP4. This is the Release 01 video copy for DVIDS ID 1006063; release metadata gives a file size of 27,493,716 bytes and SHA 256 46fdeec860cdaa6b27b6fcc51d0b7e045347d494832cdc65f2038e702eaa58aa. Official source page: DVIDS v…

Release 01#war-gov#pursue#release-01#official-source#evidence#video#dvids#infrared

DOW-UAP-PR26, Unresolved UAP Report, United Arab Emirates, October 2023

Evidence media

  • Open Sky release-file copy: open the verified 43-second MP4. This is the Release 01 video copy for DVIDS ID 1006063; release metadata gives a file size of 27,493,716 bytes and SHA-256 46fdeec860cdaa6b27b6fcc51d0b7e045347d494832cdc65f2038e702eaa58aa.
  • Official source page: DVIDS video 1006063.
  • Companion context: DOW-UAP-D23 is the paired mission-report PDF, and DOW-UAP-PR27 is the longer companion video. Keep the three records separate until a timestamped source crosswalk assigns specific video intervals to specific D23 form lines.

Investigation reading

This released item is the short DVIDS video paired in Release 01 with DOW-UAP-D23, a United Arab Emirates / Persian Gulf MISREP packet. The verified MP4 is a 43.17-second, 1920 × 1080 H.264 video at 30 fps with an AAC audio stream. The release description says the footage came from an infrared sensor aboard a U.S. military platform in 2023 and describes an “area of contrast,” not a resolved object identification.

The media pass treated the video as its own evidence item rather than simply repeating the manifest prose. A whole-clip contact sheet was reviewed, with additional frame checks around the publisher-described intervals near 17–18 seconds, 29 seconds, and the end of the clip. The underlying scene is a monochrome/infrared-looking sky or cloud field with broad gray gradients, cloud-like texture, faint banding, and compression/sensor noise. Several display elements are fixed overlays rather than scene content: cyan reticle/corner brackets, a cyan N indicator, and multiple black rectangular masks or redactions around the frame.

A tiny bright, unresolved contrast feature is visible in sampled frames, generally left of the central reticle and sometimes slightly below or above the reticle line depending on the sampled time. Around 17 seconds it appears as a very small pale smudge just left and a little below the reticle center. Around 29 seconds it is farther left of center and somewhat above the reticle, clear of the black mask blocks. In the final sampled frame after the reported non-reacquisition period, no clearly reacquired target is visible; the frame mainly shows gray cloud/IR background plus fixed display overlays. This draft does not identify the feature or infer size, range, speed, or object type from the video alone.

What the file appears to contain

  • A short grayscale/IR-style sensor clip published through DVIDS as video 1006063 and represented in Release 01 as DOW-UAP-PR26.
  • A small point-like area of contrast against a low-detail sky/cloud background. It remains unresolved in the reviewed frames and is not visibly shaped enough to classify.
  • Fixed display symbology and redaction/mask blocks. Those black rectangles and cyan marks should be separated from the original scene and not described as objects in the sky.
  • Motion that is difficult to separate from sensor/platform movement. The release text says the sensor pans around 17–18 seconds, stops tracking at about 29 seconds, and does not reacquire the area of contrast from 30–43 seconds. The sampled frames are consistent with changing line of sight across a cloudy/gray background, but they are not enough to reconstruct kinematics.
  • A direct relationship to the D23 mission-report packet. D23 contains two UAP form lines at 0241Z and 0322Z; this video page does not decide which line, if either, this clip should be assigned to without a timestamped source crosswalk.

Source custody and provenance

  • Official/source URL: DVIDS video 1006063.
  • Open Sky release-file copy: war-gov-dow-uap-pr26-unresolved-uap-report-united-arab-emirates-october-2023-1006063.
  • SHA-256: 46fdeec860cdaa6b27b6fcc51d0b7e045347d494832cdc65f2038e702eaa58aa.
  • File size: 27,493,716 bytes.
  • Video metadata checked from the verified Open Sky release-file copy: H.264 video, 1920 × 1080, 30 fps, 43.17 seconds, with AAC stereo audio. The MP4 metadata creation time is May 7, 2026.
  • Release metadata ties the video to DVIDS ID 1006063, CSV row context 42 / 84, Department of War agency metadata, and incident-location language for the Persian Gulf / United Arab Emirates. The associated D23 PDF should remain the documentary companion record rather than being collapsed into this video page.

Graph context

The graph has an exact VideoEvidence record for DVIDS 1006063 and release-record context linking this item to DOW-UAP-D23. The semantic layer currently preserves 19 source-text claim records, 11 entity mentions, 0 sensor-event records, and 0 table rows for this video record. Those counts come mostly from the official manifest/DVIDS description: agency names, D23/PR26 identifiers, the “one UAP” / “area of contrast” wording, the 43-second infrared-footage description, and the time-coded video description.

Related graph records include the D23 mission-report PDF and the companion PR27 video. The D23 report has its own page because it is a nine-page form packet with two UAP line entries, FMV/IMINT/SIGINT context, redacted coordinates, and thermal-signature language. This PR26 page should be used as the media-side read-through: a short video with an unresolved contrast feature and limited public telemetry.

Leads to check

  • Reconcile the PR26 clip with D23’s two UAP form lines at 0241Z and 0322Z. The public release pairs them, but this draft does not assign PR26 to a specific form line.
  • Compare PR26 against PR27, the longer 4-minute-57-second UAE infrared clip, before treating them as separate observations or repeated views of the same report thread.
  • Look for any official source packet that provides sensor mode, IR polarity, platform/range, field of view, azimuth/elevation, target-track state, or unredacted coordinates. Without those, apparent motion cannot be converted into object motion.
  • Review whether the graph row numbering needs cleanup: video context lists rows 42 / 84, while a related release-record node carries a current-row anchor for PR26. Treat this as provenance bookkeeping, not an evidence conflict.
  • Run prosaic checks before escalation: cloud background, sensor panning, platform motion, compression artifacts, display overlays, and image-processing effects can all affect a tiny point-like feature in this clip.

Lead check notes

  • Partial — D23 line assignment: the paired D23 mission-report page preserves two UAP form entries at 0241Z and 0322Z, while the PR26 source description gives a 43-second infrared video with an area of contrast. The current linked release material does not assign PR26 to one specific D23 form line.
  • Partial — PR26/PR27 comparison: PR27 is a separate DVIDS video (1006067) with a 4:57.33 runtime and SHA-256 3c510fe34677a0784c656105b9653ce3039d83de46f929f1ee6d6fb820ce3b8c. The two videos should remain separate media records until a source-backed frame/timestamp comparison shows whether they are different views, different form lines, or repeated views of one report thread.
  • Blocked — Missing sensor/platform data: the PR26 release page and D23 report do not provide sensor mode or polarity, field of view, range, azimuth/elevation, track state, platform motion, or unredacted coordinates. Apparent motion in the clip cannot be converted into object motion without that source material.
  • Partial — Row/provenance cleanup: the video evidence record carries related row context 42 / 84, while the current release-record anchor for PR26 appears as row 81. Treat this as release bookkeeping until the manifest history is reconciled, not as an evidence conflict.
  • Needs external source — Prosaic checks: the media review separated fixed overlays/masks and sensor panning from scene content, but ordinary explanations such as cloud texture, platform/sensor motion, compression or image-processing artifacts, distant aircraft, balloon, bird, debris, or other small-point sources need unredacted time/location/geometry or raw sensor/track data before they can be tested.

Limits

This is a graph investigation draft, not a finding. The public video shows a very small unresolved contrast feature in a masked infrared-style display, but it does not provide enough public metadata to establish range, altitude, size, speed, identity, or origin. The black rectangles and cyan symbology are display/redaction elements, not scene objects. The contact-sheet and frame checks support cautious visual description only; they do not resolve the event. The paired D23 PDF supplies mission-report context but does not embed the video frames or raw sensor data.

Deep investigation — graph + web reconnaissance

Source reread and media checks

The deep reread keeps PR26 anchored as the short DVIDS media record for video 1006063, not as a resolved object identification. The verified release-file copy is 27,493,716 bytes with SHA-256 46fdeec860cdaa6b27b6fcc51d0b7e045347d494832cdc65f2038e702eaa58aa. Media decoding confirms a 43.17-second H.264 MP4 at 1920 × 1080 and 30 fps with an AAC stereo stream; the decoded audio stream measured as silent during this check, so there is no public narration or transcript to use.

Representative frames at 0, 5, 10, 16, 17, 18, 29, 30, and 42 seconds show a grayscale/infrared-style display with fixed cyan reticle/corner marks, a cyan N, and black redaction/mask blocks. Those overlays are display elements, not scene objects. The background is low-contrast and cloud- or haze-like, with mottling, banding, and sensor/compression noise. A tiny bright speck or smudge is visible in several sampled frames left of the reticle or elsewhere in the display, but it is not resolved enough to classify. The 29-second frame still shows a small bright feature; the 30-second and 42-second samples do not show a clear reacquisition, matching the release description's claim that the area of contrast leaves the field of view and is not reacquired from 30 to 43 seconds.

Graph connections and provenance cleanup

Read-only graph checks found the exact VideoEvidence node for DVIDS 1006063, with the expected title, canonical DVIDS URL, CloudFront MP4 URL, current official-CSV status, and full_download_sha256 matching the verified MP4. The semantic layer for this video currently contains two manifest-description text chunks, 19 machine-extracted claim nodes, 11 entity mentions, 0 sensor-event nodes, and 0 candidate crosslinks. The claims mostly restate release metadata and publisher description: DOW-UAP-PR26, DOW-UAP-D23, Department of War / DVIDS / AARO, United Arab Emirates / Arabian Gulf, 43 seconds, the observer statement that one UAP was reported, and the time-coded sensor-pan / non-reacquisition description. They remain machine_extracted_needs_human_review and not_a_finding unless checked against the source media and manifest text.

The graph also preserves the companion-document context. PR26 is tied to the D23 mission-report packet, whose public pages contain two UAP form lines at 0241Z and 0322Z, and PR27 is a separate DVIDS video record for the longer companion clip. A graph hygiene issue remains: one current PR26 release-record node is matched by title and DVIDS ID but still carries a stale file-pointer field from an older mission-report record, and a direct RELATED_TO neighbor can surface a PR29 row through row 84 bookkeeping drift. Those are provenance-cleanup leads, not evidence that PR26, PR29, or D14 are the same event. Stable identity for this page is the DVIDS ID, title, MP4 URL, byte count, and SHA-256.

External official-source checks

The live DVIDS page for video 1006063 was reachable and returned the PR26 title, video ID 1006063, VIRIN 231001-D-D0360-8883, category B-Roll, length 00:00:43, location AE, Date Taken: 10.01.2023, and Date Posted: 05.07.2026 23:31. The direct CloudFront MP4 endpoint also answered a range request for DOD_111688816.mp4.

The same live DVIDS page currently carries a description mismatch: its metadata description describes a 2022 still image, a red line encircling an area of interest, and an accompanying DoW-UAP-D12 mission report, rather than the expected PR26 / D23 43-second infrared-video description preserved in the Release 01 manifest and graph chunks. Treat that as an official-source metadata anomaly to reconcile, not as a reason to merge PR26 with D12 or to assign this clip to a D23 form line. Direct WAR.GOV landing, CSV, and D23 PDF probes returned 403 Forbidden during this check, and Internet Archive availability probes returned rate-limit errors, so the custody anchor remains the verified release-file copy plus the previously ingested official manifest records.

Prosaic checks and limits

The available public evidence supports cautious media description, not kinematic reconstruction. The source describes sensor panning, tracking loss, and non-reacquisition, and the video itself shows fixed overlays, masks, low-detail background texture, and a tiny unresolved contrast feature. Without the missing sensor mode/polarity, field of view, range, azimuth/elevation, platform track, raw target track, unredacted D23 coordinates, exact video-to-form-line crosswalk, and weather/traffic/satellite context, apparent screen motion cannot be converted into object speed or behavior.

Graph labels for weather, astronomy, and launch context exist, but exact-date probes for the public date candidates (2023-10-01, DVIDS; 2023-10-24, D23 body; 2023-10-31, release metadata) did not return a modeled context row tied to this record. That is a graph-coverage limit, not an exclusion of ordinary explanations. Priority prosaic lanes remain sensor/platform motion, cloud or sea-background contrast, compression/gain/display artifacts, small distant aircraft or airborne clutter, balloon/bird/debris possibilities, and the DVIDS description/manifest bookkeeping mismatch.

Follow-up leads

  • Reconcile the live DVIDS PR26 description mismatch against a later DVIDS page snapshot, official WAR.GOV CSV revision, or preserved release manifest.
  • Build a timestamped source crosswalk before assigning PR26 to the D23 0241Z line, the D23 0322Z line, or a separate media/reporting thread.
  • Compare PR26 and PR27 only at the source-media level; graph adjacency and shared D23 context are not enough to decide whether they show different form lines or repeated views of one report thread.
  • Keep stale row/file-pointer relationships in the graph as maintenance issues until reconciled; do not use them as corroboration.

Audit note

This section is a source-grounded audit note, not a finding. It distinguishes the verified MP4, publisher/manifest claims, machine-extracted graph claims, live official-source metadata, and unresolved prosaic checks. No hypothesis or resolution decision is created from this video page.

Sources