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

341_110448_Records_Relating_to_the_Collection_and_Dissemination_of_Intelligence_1948-1955-TS_CONT_No.2_2-5300-2-5399

Official/source PDF: WAR.GOV PDF Open Sky release file copy: hash verified PDF Source media status: this is a scanned seven page PDF, not a photograph, video, map, table, radar plot, or raw instrument record. The images below are derived page renders from the official PDF for re…

Release 01#war-gov#pursue#release-01#official-source#evidence#graph-investigation-draft#pdf

341_110448_Records_Relating_to_the_Collection_and_Dissemination_of_Intelligence_1948-1955-TS_CONT_No.2_2-5300-2-5399

Evidence media

  • Official/source PDF: WAR.GOV PDF
  • Open Sky release-file copy: hash-verified PDF
  • Source media status: this is a scanned seven-page PDF, not a photograph, video, map, table, radar plot, or raw instrument record. The images below are derived page renders from the official PDF for reading context.

Derived render of page 4: USAFE 10 unidentified-aircraft sighting report

Page 4 is a typed USAFE 10 / TT 1524 item dated 4 Nov 1948. A high-resolution render supports the reading 1402Z 5 Sep 48, 5155N/0355E, 307th Bomb Group, Operation Dagger, 30,000 feet, smoke/condensation trails, sudden acceleration and climb, and the source evaluation line B-2.

Derived render of page 5: USAFE 14 flying-saucer discussion

Page 5 is a typed USAFE 14 / TT 1524 item dated 4 Nov 1948. It is visibly text-bearing even though the stored OCR body for page 5 is blank. The page discusses recurring “flying saucer” reports, Neubiberg Air Base, and a Swedish Air Intelligence Service discussion.

Derived render of page 6: adjacent Soviet radar/TU-2 intelligence context

Page 6 is adjacent Soviet radar/TU-2 intelligence, not the Holland or Neubiberg sighting thread. Page rendering supports the Kholomia/Kholonia airfield context, a radar-set name that appears to read Redut but remains OCR-sensitive, 150 kilometers, 3000 meters, and IFF type S. CH. (C-4); the stored OCR conflicts with several of those fields, so the set-name spelling should stay under transcription review.

Investigation reading

This Release 01 item is a seven-page scanned Air Force intelligence packet. The WAR.GOV manifest describes it as a November 1948 Air Force intelligence report relating to unidentified flying objects and flying saucers, with release row 23, agency Department of War, manifest incident date 11/8/48, and manifest location Netherlands.

The file contains more than one thread. Pages 1-3 are custody and forwarding material for Extracts from TT #1524. Page 4 is a Holland-area unidentified-aircraft report from 5 September 1948. Page 5 is a separate flying-saucer discussion that quotes a Swedish Air Intelligence Service conversation and mentions a recent Neubiberg Air Base hovering report. Pages 6-7 are adjacent radar/aircraft intelligence items. These threads should not be flattened into a single incident.

This is a graph investigation draft only. It preserves what the source appears to say, what the page renders clarify, and what the current graph models. It does not decide whether any reported object was extraordinary.

What the file appears to contain

PageReading notes
1Cover page for Directorate of Intelligence, Headquarters U.S. Air Force, with declassification and top-secret security-information markings.
2AFOIR cover sheet dated 8 Nov 48 for Extracts from TT #1524, routed through D/I, HQ USAF via A-2, USAFE. It records distribution notes, including that USAFE 14 was distributed to an Armament Intelligence Branch recipient, rather than an incident narrative.
3USAFE 2 / TT 1524, dated 4 Nov 48, to Gen. Cabell. It says a complete set of reports from a special intelligence organization of European Command is being forwarded for inspection and final disposition, with caution that the reports were forwarded as received.
4USAFE 10 / TT 1524, dated 4 Nov 1948. It reports an unidentified aircraft sighting from the 307th Bomb Group during Operation Dagger. The high-resolution page render supports the sighting line as 1402Z 5 Sep 48, off the west coast of Holland at 5155N/0355E. The reporting aircraft were at 30,000 feet. The unidentified aircraft was first seen cruising at normal jet speed on heading 120 degrees, then leaving smoke and condensation trails, accelerating suddenly, and climbing. The report says observers generally agreed it was a single jet-propelled aircraft, probably using rocket assist, with more reserve power than normal 1947 jets. It was never within identification range, and the item gives the evaluation as B-2.
5USAFE 14 / TT 1524, dated 4 Nov 1948. The page says recurring “flying saucer” reports had continued to appear and “cannot be disregarded.” It says one object had recently been observed hovering over Neubiberg Air Base for about thirty minutes. It then describes a visit with Swedish Air Intelligence Service personnel who reportedly said reliable technically qualified people considered the phenomena to show “a high technical skill” not credited to any known culture on Earth, and were considering unknown or unidentified technology, possibly outside Earth. The page also recounts a Swedish lake case: a technical expert reportedly observed an object crash or land in a lake; Swedish intelligence reportedly sent a naval salvage team; divers reportedly found a previously uncharted crater or depression on the lake floor; and the results were not yet available. These are source statements requiring independent checks, not findings.
6A top-secret Soviet aviation/radar intelligence page. It describes a Soviet-built radar set near Kholomia/Kholonia Airfield, PVO assignment, a radar-set name that the render appears to read as Redut but that remains a transcription-review item, radar-training ranges, TU-2 aircraft equipment, IFF type S. CH. (C-4), airborne radar/tail-warning-device testing, airfield dimensions, fuel handling, and TU-2 crew/manufacturing notes. This looks like adjacent intelligence context rather than the Holland or Neubiberg report. The stored OCR conflicts with several high-signal fields, including the set name and the distance/climb readings; the page render supports 150 kilometers and 3000 meters, but those values should remain tied to the page transcription rather than treated as separate sensor evidence for the UAP threads.
7A short USAFE 16 / TT 1524 SECRET item. The rendered page favors the date 4 Nov 48. It says a CIC source reported that Russians had ordered CSR missions in foreign countries to purchase radar tubes in large quantities regardless of cost, and to purchase a complete radar set. It carries an Eval C-3 line.

No photographs, maps, diagrams, tables, object images, or instrument displays were identified in the rendered-page checks. The evidence value here is documentary: typed intelligence reporting, routing/custody context, and source-quoted claims.

Source custody and provenance

  • Official/source URL: WAR.GOV PDF
  • Open Sky release-file endpoint: hash-verified PDF
  • SHA-256: 45b2fdb6c919cd8ed9c4406f4d60c7add073669dff5a5497595d4ecf08d72073
  • File size verified from the release-file copy: 9,888,054 bytes
  • PDF page count: 7
  • WAR.GOV release row: 23
  • Agency: Department of War
  • Source/container kind: pdf
  • OCR status recorded for the asset: frontier_ocr_complete, with 7 page markers and 6 non-empty OCR pages.

The PDF is image-based and copy-disabled; ordinary selectable-text extraction produced only page breaks. The public reading above therefore uses the stored OCR where it exists and high-resolution rendered-page inspection for page 5 and for contested page 4, page 6, and page 7 fields. Direct official WAR.GOV HEAD and byte-range GET requests can return 403 Forbidden, so this page relies on the hash-verified release-file copy while citing the official URL.

Graph context

The graph has two exact official-primary document records for this item: a WAR.GOV release-row record for row 23, and a PDF asset record for the canonical media file. The PDF asset is linked to the source dataset war_pursue_uap_release_2026_05_08 and the semantic dataset war_pursue_release01_semantic_2026_05_12.

Current semantic graph counts for this asset are:

  • 35 extracted claim records
  • 23 extracted entity mentions
  • 13 sensor/platform event records
  • 0 table-row extractions
  • 0 candidate crosslinks returned for this item in the current context extract

The graph captures the page 4 Holland aircraft sighting, including the 307th Bomb Group, aircraft/speed/altitude language, smoke and condensation trails, sudden acceleration/climb, and the source evaluation line. It also captures page 6 and page 7 radar/aircraft terms, but those are adjacent intelligence references and should not be treated as raw radar evidence for the page 4 or page 5 sighting threads.

The graph under-models page 5 because the stored OCR has an empty page-5 body even though the rendered page contains the Neubiberg and Swedish Air Intelligence Service flying-saucer discussion. It also carries OCR-derived page 6 readings that the high-resolution render does not support. Those are extraction quality-control issues, not source conclusions.

Deep investigation — graph + web reconnaissance

Source reread and media audit

This deep pass rechecked the hash-verified seven-page release-file copy against the WAR.GOV manifest metadata and the page renders. The PDF remains a documentary packet, not sensor media: pages 1-3 establish custody/routing for Extracts from TT #1524; page 4 reports a 307th Bomb Group unidentified-aircraft observation off Holland; page 5 is a separate flying-saucer discussion involving Neubiberg Air Base and Swedish Air Intelligence Service claims; pages 6-7 are adjacent Soviet radar, TU-2, and radar-procurement intelligence.

Key transcription controls from this pass:

  • Page 4 render favors 1402Z 5 Sep 48 rather than the OCR-like 14022; the report gives 5155N/0355E, 30,000 feet, normal jet speed on heading 120 degrees, smoke/condensation trails, sudden acceleration/climb, and official evaluation B-2.
  • Page 5 is text-bearing despite the stored OCR gap. It is therefore a priority source-text repair target before the Neubiberg and Swedish lake material is promoted in graph form.
  • Page 6 is not a UAP narrative. It is Soviet radar/TU-2 intelligence; the render supports Kholomia/Kholonia ambiguity, likely Redut, 150 kilometers, 3000 meters, and S. CH. (C-4), while the stored OCR contains conflicting readings. Those conflicts should stay marked as transcription issues.

Graph reconnaissance

Read-only Neo4j checks found two exact official-primary document records for this item: the Release 01 CSV row record (row 23, manifest incident date 11/8/48, manifest location Netherlands) and the linked PDF asset record for the canonical WAR.GOV media URL. The PDF asset is modeled with the verified SHA-256 45b2fdb6c919cd8ed9c4406f4d60c7add073669dff5a5497595d4ecf08d72073, 9,888,054 bytes, seven pages, four frontier-OCR text chunks, and six non-empty OCR pages.

Graph relationships around the asset currently include 35 machine-extracted claims, 23 entity mentions, 13 sensor/platform event records, 7 text chunks across the row/asset/manifest/OCR records, and no candidate crosslinks returned for this slug. These semantic records remain machine_extracted_needs_human_review / not_a_finding; they are useful navigation aids, not conclusions. Several graph records correctly point to the Holland aircraft-language thread, but page 5 is under-modeled and page 6 contains OCR-sensitive radar/TU-2 terms that should not be blended into the UAP sighting threads.

External provenance and web checks

Official/archive-first reconnaissance found that the Internet Archive CDX has a 2026-05-08 200 application/pdf capture of the same WAR.GOV PDF path, and the archived PDF reports the same 9,888,054 byte content length as the local release-file copy. A 2026-05-12 archived WAR.GOV CSV snapshot also preserves row 23 with the same title, Department of War agency, manifest incident date 11/8/48, manifest location Netherlands, PDF link, and description blurb.

Searches for the archival scan identifiers and source language (DocId:34714985, NW 90307, USAFE 10, 307th Bomb Group, 5155N, Operation Dagger, and Kholomia/Redut/TU-2 combinations) did not surface an independent official NARA, Project Blue Book, USAFE, or Swedish archival corroboration during this pass. The only indexed non-official duplicate surfaced was a community GitHub conversion of the same Release 01 file, which is useful as a mirror lead but not as corroboration.

Prosaic checks and limits

The page 4 description itself points first toward aircraft/aviation explanations: observers described an aircraft-like object at normal jet speed, smoke and condensation trails, sudden acceleration/climb, and probable rocket assist, while also noting it was never close enough for positive identification. The Open Sky graph has no exact 1948 astronomy, weather, or launch-context records indexed for this date/location query, and the packet supplies no weather log, flight log, radar plot, photographs, or instrument data. Page 6 and page 7 radar references are real source facts inside the packet, but they are adjacent intelligence items and do not corroborate the Holland or Neubiberg sighting threads.

Follow-up leads

  • Replace the blank page-5 OCR with a clean page-aware transcription, then rerun/repair semantic extraction for the Neubiberg, Swedish Air Intelligence Service, lake-crater, and naval-salvage claims.
  • Seek official or archival corroboration for the page 4 Holland report in 307th Bomb Group, Operation Dagger, USAFE, European Command, Project Sign/Blue Book, or NARA holdings.
  • Treat the Swedish lake story as a lead bundle: identify the lake, the technical expert, Swedish Air Intelligence Service contact, naval salvage order, diver report, and any crater/depression documentation separately.
  • Keep the packet split into four source threads: forwarding/custody, Holland aircraft sighting, Neubiberg/Swedish flying-saucer discussion, and Soviet radar/TU-2/radar-procurement intelligence.

Audit note

This section was added from source reread, render inspection, read-only graph queries, and official/archive-first web reconnaissance. No graph writes, findings, hypotheses, or resolution decisions were created.

Leads to check

  • Produce a clean page-aware transcription/OCR update for page 5, then update semantic extraction so the Neubiberg and Swedish lake material is represented with citations.
  • Confirm the page 4 Holland sighting through independent 307th Bomb Group, Operation Dagger, USAFE, European Command, or Project Sign-era records before treating the report as corroborated.
  • Check Neubiberg Air Base reporting for the week before 4 Nov 1948 to determine whether the alleged thirty-minute hovering report exists in an operational log, intelligence report, or witness statement.
  • Identify the Swedish lake case described on page 5. The page points toward Swedish Air Intelligence Service personnel, a technical expert witness, naval salvage activity, divers, and an uncharted crater/depression; each element needs separate archival confirmation.
  • Correct or flag page 6 OCR fields in the graph/source text: stored OCR and page-render readings conflict on the radar-set name and several measurements; current render inspection supports Kholomia/Kholonia, likely Redut, 150 kilometers, 3000 meters, and S. CH. (C-4), but the set-name spelling should remain under transcription review.
  • Separate the packet’s distinct threads in future analysis: page 4 Holland aircraft sighting, page 5 flying-saucer discussion, page 6 Soviet radar/TU-2 intelligence, and page 7 radar-procurement note.
  • Continue prosaic checks before escalation: aircraft operations, jet/rocket-assist testing, contrails, exercises, weather, reporting-chain reliability, and possible confusion between adjacent intelligence items.

Lead check notes

  • Checked — source size/hash/page count: The release-file copy is 9,888,054 bytes, 7 pages, and matches SHA-256 45b2fdb6c919cd8ed9c4406f4d60c7add073669dff5a5497595d4ecf08d72073.
  • Checked — page 4 time, coordinates, and unit context: The stored OCR and page-4 render support 307th Bomb Group, Operation Dagger, 5155N/0355E, and 30,000 feet. The rendered page favors 1402Z 5 Sep 48 over the OCR-like 14022.
  • Checked — page 5 is text-bearing despite missing OCR: The page-5 render shows a full typed USAFE 14 / TT 1524 item about recurring flying-saucer reports, Neubiberg, Swedish Air Intelligence Service discussion, and a Swedish lake salvage lead. The stored OCR body for page 5 is blank.
  • Partial — independent corroboration: Current Open Sky coverage surfaces this packet and a related Neubiberg pointer, but no separate operational log, Swedish archival record, or independent 307th Bomb Group source has been linked to this page yet.
  • Checked — page 6 is adjacent intelligence context: The page render supports Soviet radar/TU-2 intelligence as a separate thread and shows no flying-saucer or unidentified-object narrative on that page. The radar-set name appears to read Redut, but because OCR/page-render readings conflict, the spelling remains a transcription-review item.
  • Checked — page 7 date: The rendered page favors 4 Nov 48, not 4 Nov 46.

Limits

  • This page is a graph investigation draft and still needs human review.
  • The source is a scan of typed intelligence documents, not direct sensor data, object imagery, a contemporaneous flight log, or an adjudicated case file.
  • Page 5 is visibly text-bearing but missing from stored OCR. The page-5 reading should be backed by a clean transcription before promotion.
  • Several date fields need careful handling: the manifest gives 11/8/48, the packet includes multiple 4 Nov 1948 item dates, page 4 reports a 5 Sep 48 sighting, and page 7 is a separate radar-procurement intelligence item.
  • The graph extraction is useful for navigation, but it includes machine-extraction artifacts such as title-range years and adjacent radar/aircraft references. Those are not adjudicated timeline or sensor facts.
  • The page does not establish whether any reported object was extraordinary. Prosaic checks and source-chain corroboration come before escalation.

Sources