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

38_143685_box_Incident_Summaries_101-172

Official PDF: Open the verified Release 01 PDF. The Open Sky release file copy is the 178 page official PDF represented by the SHA 256 value in the provenance section. Derived page renders: The images below are page renders from that PDF for inspection. They are not separate sta…

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

38_143685_box_Incident_Summaries_101-172

Evidence media

  • Official PDF: Open the verified Release 01 PDF. The Open Sky release-file copy is the 178-page official PDF represented by the SHA-256 value in the provenance section.
  • Derived page renders: The images below are page renders from that PDF for inspection. They are not separate standalone photographs or new evidence items.

Derived Release 01 PDF page 003: Incident 101 checklist

Page 3 render: Incident #101 checklist for the Norcatur, Kansas fireball/explosion report. The page is a text checklist; it references a vapor-trail photograph but does not itself contain the photograph.

Derived Release 01 PDF page 076: Incident 124 radar checklist

Page 76 render: Incident #124 checklist for a North Atlantic radar-sighting entry.

Derived Release 01 PDF page 077: Incident 124 radar continuation

Page 77 render: typed continuation text for Incident #124. It describes an air-search radar contact; it is not a radar-scope image or plotted track.

Derived Release 01 PDF page 114: Chiles-Whitted checklist

Page 114 render: Incident #144 checklist for the Chiles-Whitted Eastern Air Lines report south of Montgomery, Alabama.

Derived Release 01 PDF page 118: Chiles typed statement

Page 118 render: typed Chiles statement describing the Eastern Air Lines Trip 576 sighting. It is a source statement, not an object photograph.

Derived Release 01 PDF page 131: Indianapolis/Fall Creek checklist

Page 131 render: Incident #151 checklist for the Indianapolis/Fall Creek report, noting low altitude, slow speed, and attached drawing references.

Derived Release 01 PDF page 134: Fall Creek sketch map

Page 134 render: hand-drawn Fall Creek area sketch map with roads, woods, bridge, and the marked path of object.

Derived Release 01 PDF page 137: object-profile sketch

Page 137 render: hand-drawn object-profile sketch with measurements and an intelligence-officer signature block. This is a source sketch, not a photograph.

Derived Release 01 PDF page 170: Arnhem/The Hague checklist

Page 170 render: Incident #168 checklist for an Arnhem/The Hague report comparing the reported sound and speed to a V-2. The checklist says sketches or photographs were none.

Investigation reading

This draft treats the file as a packet-level source, not as one sighting. The released PDF is a 178-page scan of incident-summary material. The PDF metadata reports 178 pages and a file size of 242,733,260 bytes. The OCR text contains 178 page sections, with text present for every page. The verified SHA-256 is:

f4b5cf6a93fe47c09123ce53706e430c814229ed688a49a96244a90b4bcf3e05

The reading used the complete OCR text, the PDF page/image inventory, and targeted searches across the whole document for incident numbers, dates, locations, observers, radar, photographs, sketches, diagrams, aircraft, weather balloons, meteors, rockets, V-2 references, clouds, smoke, reflection language, and named anchors such as Norcatur, North Atlantic, Chiles, Whitted, San Acacia, Kirtland, Indianapolis, Fall Creek, Turkey, and Moscow. Low-text pages were checked as possible visual attachment pages rather than treated as blank; several of those pages are sketches or maps.

The practical reading is that this is a 1948 incident-summary bundle. It contains checklists, witness statements, newspaper-derived accounts, correspondence, attached drawings, sketch maps, and administrative control pages. The packet should be split into incident-level child records before anyone uses it for case-level interpretation.

What the file appears to contain

The official manifest describes the PDF as a group of incident summaries using a "Check-List - Unidentified Flying Objects" format, with some witness lists, statements, and narrative descriptions. The file supports that description. It is not a single UFO case and it is not a clean photographic or radar exhibit set.

PagesIncident anchors visible in the OCRReading note
1-2Cover/control materialDeclassification, classification, and register/control pages for a flying-saucer subject file.
3-16101Norcatur/Norton, Kansas, 18 February 1948. The checklist describes a fireball/explosion, bluish-white smoke, a mushroom-shaped cloud, apparent construction as "Meteor," recovered meteorite material, and a referenced vapor-trail photograph. Follow-on correspondence includes speculative astronomy and rocket/moon arguments; that correspondence is source context, not an archive finding.
17-19102-103Related B-29 crew observations around Green River/Limon on 18 February 1948, describing a large multi-colored ball of fire and dense smoke. These pages cross-reference the Norcatur material and include redacted or not-stated fields.
20-34104-111March-April 1948 entries from Smyrna, Belmont, Bakersfield, Baltimore/Hamilton, and an airborne P-47 observation. The P-47 page describes a silver half-span or flying-wing-type object and rapid acceleration; the rendered page is a text form, not a photograph.
35-48112 and sublettersAshley/Delaware, Ohio witness cluster. The entries describe oblong, cylindrical, spherical, or blob-like forms and repeatedly show uncertainty about distance and scale.
49-60113-116Includes a parachute/canister-style exhibit sketch, a motion diagram of objects accelerating, and prosaic follow-up language about balloons, aircraft, and meteor possibilities. The rendered exhibit pages are drawings/diagrams, not photographs.
61-77118-124Includes European and U.S. entries and the North Atlantic radar-only report. Incident #124 says an air-search radar contact was tracked from about 6,500 yards to a longer range at about 30 mph; the continuation page is text only, with no radar-scope image or track plot. The rendered page appears to favor 18,000 yards where OCR may read 15,000 yards, so the exact range should be rechecked from the page image.
78-97125-133Short entries for bird, sphere, disc, ball, and questionnaire-style reports. Incident #132 carries an IR-coded heading, but the page is a checklist/record heading rather than an infrared image.
98-113136a-143Later June-July 1948 entries including fireball/orange-ball reports, Chapel Hill/Columbus material, aircraft-comparison language, and visibility limits.
114-120144 and sublettersThe Chiles-Whitted Eastern Air Lines report near Montgomery, Alabama, 24 July 1948. The packet includes checklist pages and typed statements describing a cigar-shaped or jet-like object, rear flame, and blue glow. The reviewed pages are text pages; the checklist references sketches or photographs, but no standalone object photograph is visible here.
121-134145-151Includes San Acacia/Kirtland-linked observer information and an Indianapolis/Fall Creek report of a broad, short, propeller-like object. Incident #151 has visible drawing and sketch-map pages showing a 6-8 ft object with protruding cups and a mapped path near Fall Creek.
135-149152-157Includes a cymbal-like object with a profile sketch, redacted entries, starlike and balloon-like descriptions, and a Columbus, Ohio object identified in the text as a small county-fair-style hot-air balloon.
150-178158-172Late packet entries include streak, cylinder, star-like, sphere, V-2/rocket, Arnhem/The Hague, Turkey, Moscow-area, and final control/stamp material. Several descriptions explicitly compare sounds or speeds to V-2 or aircraft; those are source descriptions to be checked at the incident level, not conclusions.

Representative source details that should stay separate from the mass of forms:

  • Norcatur / Incident #101: the checklist uses "Meteor" in the apparent-construction field and references a vapor-trail photograph. The follow-on page describes recovered meteorite fragments, including a much larger recovered stone. The rendered checklist page itself has no photograph.
  • Radar / Incident #124: this is the main sensor-specific entry in the packet. The continuation text describes air-search radar contact, visual search of the area, no surface-radar indication, and an S5 target strength fading at the longer range. The released page is typed prose, not raw radar imagery.
  • Chiles-Whitted / Incident #144: the file contains the familiar Eastern Air Lines report chain near Montgomery, including typed Chiles statement material. The pages preserve witness testimony and attachment references; they do not, by themselves, supply a photo plate or resolved object image.
  • San Acacia/Kirtland / Incident #146: the checklist describes seven objects seen near San Acacia Dam by Kirtland-connected observers. The shared Kirtland anchor is useful for searching, but it is not enough by itself to assert a match to later Kirtland records.
  • Indianapolis/Fall Creek / Incident #151: this is one of the clearest visible attachment examples in the packet. The form describes a low, slow, shiny, broad-propeller-like object; the following pages show a measured sketch and a local sketch map.
  • Prosaic leads appear throughout: meteor, meteorite, weather balloon, small hot-air balloon, sun reflection, aircraft traffic, clouds/weather, V-2 or rocket comparison, and witness distance-estimate issues recur across the packet. These checks need to remain attached to individual child incidents rather than being flattened into one explanation for the whole bundle.

Source custody and provenance

The WAR.GOV media URL is preserved as the canonical citation. Direct official-media probes can return 403 Forbidden; source custody for this draft therefore rests on the official URL, Release 01 row metadata, the verified release-file copy, and the matching SHA-256.

Graph context

The graph has exact provenance records for both the released PDF asset and the official Release 01 row record. For this asset, the graph context currently indexes:

  • 1 released Document asset record for the PDF.
  • 1 related official Release 01 row Document record.
  • 127 text chunks connected to the asset record, including OCR and manifest/context chunks.
  • 1,923 extracted source-text claim records.
  • 482 entity-mention records.
  • 66 sensor/platform event records.
  • 10 candidate crosslink records.

The graph material is useful for navigation, not adjudication. The 66 sensor/platform records are heavily influenced by ordinary source words such as aircraft, radar, and IR appearing inside forms and witness statements. Only a small part of that set is a true sensor-specific source passage, especially the Incident #124 radar text. Candidate crosslinks are also cautious: the Kirtland/San Acacia text around Incident #146 creates possible links to other Kirtland records in FBI Vault and NARA material, but those links remain review leads only.

The graph also shows why this packet should not be summarized as a single case. One PDF node contains dozens of incident numbers, subletters, and source threads. The next useful graph step is to split the bundle into reviewed child records for each incident or cluster while preserving this PDF as parent custody evidence.

Leads to check

  • Split the packet into incident-level child pages before promoting any case-level interpretation.
  • Verify attachment trails for referenced photographs and sketches, especially the Norcatur vapor-trail photograph and the Chiles-Whitted checklist's note that sketches or photographs existed.
  • For Incident #124, look for the separate radar-sighting file or any raw radar plot/track material; this PDF gives text about the radar report but not the instrument record.
  • Recheck the Incident #124 range wording against the page image. OCR text can read the longer range as 15,000 yards, while the rendered page appears to favor 18,000 yards.
  • Reconcile date anomalies and sublettered records, including 112a/112b, 136a, 140a, 144a/144b, and other out-of-sequence or missing incident numbers.
  • Preserve prosaic checks at the incident level: meteorite records, balloon releases, local aircraft activity, weather, clouds, sun/moon geometry, rocket/V-2 references, and witness-distance estimates vary across the bundle.
  • Review the Kirtland/San Acacia candidate crosslinks manually before connecting this page to NARA or FBI records as anything stronger than a shared-anchor lead.
  • Compare OCR against rendered page images for low-text visual pages. The Fall Creek map and profile drawings are visually meaningful and can be under-described by OCR alone.

Lead check notes

  • Partial — incident-level split: The Open Sky wiki already has a curated page for Chiles–Whitted Encounter, but this packet still lacks child pages for many embedded incidents, including the North Atlantic radar entry, the Indianapolis/Fall Creek report, and the San Acacia/Kirtland observer material.
  • Partial — attachment trail: OCR and page renders confirm references to a Norcatur vapor-trail photograph, Chiles-Whitted sketches or photographs, and visible Indianapolis/Fall Creek drawing material. No standalone Norcatur vapor-trail photo, Chiles-Whitted photo plate, or radar plot was identified in the reviewed renders.
  • Checked — Incident #124 radar lead: The Release 01 OCR corpus places the exact radar language in this PDF. The page says a copy was also filed in a radar-sighting file, but the current Release 01 file does not surface a separate raw radar plot by that anchor.
  • Partial — date and subletter anomalies: The OCR confirms subletter anchors including 112a, 112b, 136a, 140a, 144a, and 144b. Those anchors need source-level review before being treated as chronological or filing evidence.
  • Partial — prosaic checks: The packet itself contains incident-specific references to meteor or meteorite, balloon, aircraft, V-2/rocket, cloud/weather, sun/moon, and reflection leads. Those checks should remain attached to individual child incidents.
  • Partial — Kirtland/San Acacia crosslinks: The graph shows 10 candidate links anchored on Kirtland to FBI Vault and NARA records, all still marked for human review and not as findings.

Limits

This is a graph investigation draft awaiting human review. It does not assert that any incident in the packet is anomalous, solved, unsolved, or related to another case.

The packet is scan-heavy and sometimes dark or low contrast. OCR coverage is complete by page count, but OCR can still misread dates, initials, redactions, handwritten annotations, distances, or labels. Rendered page checks confirm visible drawings and maps on selected low-text/high-signal pages; they do not establish a complete photographic inventory.

The official media URL may be access-controlled for automated retrieval. The source custody for this page therefore depends on the official URL, the Release 01 row, the verified release-file copy, and the SHA-256 match.

Deep investigation — graph + web reconnaissance

Source reread

  • The verified Release 01 source is still best treated as a packet, not as one case: 178 scanned/copy-disabled PDF pages, 242,733,260 bytes, SHA-256 f4b5cf6a93fe47c09123ce53706e430c814229ed688a49a96244a90b4bcf3e05, and full frontier-OCR coverage for all 178 pages.
  • A fresh OCR keyword pass confirms why the file needs child-incident handling before analysis: radar appears 4 times, meteor 38, meteorite 6, balloon 17, rocket 33, aircraft 53, photograph/photo 88/91, and sketch 102. These are source and custody leads distributed across many incident numbers, not a single explanation for the packet.
  • Representative rendered-page rereads tightened three high-signal points. Page 77 confirms the Incident #124 continuation says air-search radar first contacted a target at 6,500 yards and tracked it to 18,000 yards at about 30 mph; the same page says the area was visually searched, no object was seen, no surface-radar target appeared, and target strength faded at 18,000 yards. Pages 134 and 137 are Fall Creek sketch/map/profile drawings, not photographs or raw sensor imagery. Page 170 is Incident #168, Arnhem/The Hague: a wingless/two-deck aircraft description, V-2-like sound/speed comparison, and Sketches or photographs marked None.
  • The Norcatur pages remain an important internal control: the checklist itself uses Meteor, and the continuation describes recovered meteorite material, while later correspondence adds speculative rocket/moon arguments. Those threads should remain separated as source text, meteorite evidence, and speculation.

Graph connections

  • Read-only graph checks found the expected official pair for this asset: the Release 01 row-26 Document record and the exact PDF asset Document record for the official URL/hash. The asset has 127 connected TextChunk records and is linked to the official Release 01 source; no graph writes were made.
  • Semantic graph material remains machine-extracted and not_a_finding: 1,923 Claim nodes, 482 EntityMention nodes, and 66 SensorEvent nodes are tied to the asset. Claim counts are dominated by observation/witness-testimony, motion, object descriptors, dates, platform text, and prosaic leads.
  • The 66 sensor-event nodes should be read as extraction leads, not confirmed instrumentation. The graph inventory is 61 unknown modality records, 4 RADAR, and 1 IR; sampled unknown records are mostly ordinary source words such as aircraft. The only sensor-specific passage re-confirmed in the source reread is the Incident #124 radar text, which still lacks a raw radar plot or separate instrument file in this PDF.
  • The 10 graph CANDIDATE_CROSSLINK relationships are audit-only Kirtland-anchor leads, not identity or causality claims. They point to FBI Vault UFO Part 06/07/08 pages and NARA Kirtland records from 1949-1959, while the source passage here is the 17 July 1948 San Acacia/Kirtland observer entry. That gap makes exact page-level counterpart review mandatory before promoting any crosslink.
  • Existing wiki context already has a curated Chiles–Whitted Encounter page. This PDF is useful source custody for Incident #144 pages 114-120, but it should not overwrite that event page's broader Blue Book/Battelle/NICAP context without source-by-source comparison.

External provenance and archive context

  • Direct checks of the WAR.GOV official PDF, WAR.GOV UFO landing page, and WAR.GOV CSV URL returned 403 Forbidden during this check. That is an access-control/custody note, not a contradiction of the verified release copy or the preserved canonical URL.
  • Internet Archive CDX returned a 200 application/pdf capture for the exact official PDF URL at 20260508133753. The archived length reported by CDX differs slightly from the verified release-file byte count, so it is useful as availability/provenance evidence but not a replacement for the verified local release copy and SHA-256.
  • The National Archives page Project BLUE BOOK - Unidentified Flying Objects is reachable and confirms the broader official custody lane for retired Air Force UFO records. Candidate NARA catalog IDs surfaced by the graph were reachable as catalog shell pages, but API-style fetches returned the catalog app shell rather than useful record JSON; treat those as follow-up URLs until exact record text/page images are pinned.
  • The FBI Vault UFO collection remains a relevant overlap lane for 1940s-1950s reporting, but direct fetches were Cloudflare-challenged during this check. The graph's FBI Vault page candidates therefore stay as secondary provenance leads until exact part/page comparison is completed.

Prosaic checks and unresolved work

  • Packet-level astronomy, weather, balloon, launch, satellite, or aircraft correlation would be misleading because the PDF bundles many dates and locations. Those checks should be run only after child pages are split by incident number, date, place, and witness/source chain.
  • Priority child records are Norcatur/meteorite material, the North Atlantic radar-only Incident #124, Chiles-Whitted Incident #144, San Acacia/Kirtland Incident #146, Indianapolis/Fall Creek Incident #151, and Arnhem/The Hague Incident #168. Each needs its own source-custody trail, attachment search, and prosaic check set.
  • The next evidence-custody question is not whether the packet is interesting; it is where the named attachments live: the Norcatur vapor-trail photograph, Chiles-Whitted sketches or photographs, the separate radar-sighting file for Incident #124, and the Fall Creek sketch/map originals.
  • Until that work is done, this page should stay at source-packet level. It preserves official custody, OCR/search leads, graph extraction coverage, representative visual checks, and unresolved crosslink questions without resolving any embedded report.

Audit note

This section used the current wiki page, verified release-file metadata, frontier OCR, representative page renders, read-only Neo4j queries, and official/archive web checks. It creates no graph writes and no finding, hypothesis, or resolution decision.

Sources