DOW-UAP-D017 is a 116-page Department of War case file released in PURSUE Release 02 on May 22, 2026. Originating from the Armed Forces Special Weapons Program (AFSWP) and the U.S. Air Force, it compiles investigative material spanning December 1948 through May 1950 concerning unidentified aerial phenomena observed in and around Sandia Base, New Mexico -- one of the most sensitive nuclear-weapons storage and development installations in the early Cold War.12
Provenance and Chain of Custody
The compilation was assembled by the 17th District Office of Special Investigations (OSI), Kirtland Air Force Base, New Mexico, and formally summarized on 25 May 1950. The original classification was Secret; the document was later declassified under NW 91526. Distribution at the time of production reached a broad cross-section of the national security apparatus: the Director of Special Investigations at HQ USAF, Air Force Materiel Command, the Armed Forces Special Weapons Project at Sandia Base, Holloman AFB, the Atomic Energy Commission Security Division at Los Alamos, FBI field offices in El Paso and Albuquerque, the Air Force Scientific Advisory Board, and the Fourth Army at Fort Sam Houston, among others. The breadth of that distribution is itself evidence of how seriously the phenomena were treated.
The document is structured as a multi-part dossier: security inspection correspondence from April 1949, dust-collection research from July through August 1949, scientific analytical reports by Dr. Lincoln LaPaz running through May 1950, and tabulated sighting records covering the full period. Some of the investigations contained here fed directly into Project Grudge, the Air Force's successor effort to collect and analyze UAP reports from military installations across the country.3
209 Sightings Near a Nuclear Weapons Base
The compiled tables record 209 sightings of phenomena described as green orbs, discs, and fireballs reported near Sandia Base, Los Alamos, Camp Hood in Texas, and associated locations. Witnesses were military personnel -- guards, security inspectors, officers, and scientists -- whose accounts were rated by investigators on a reliability scale (Very Reliable, Reliable, Unknown). The phenomena were consistently described as maneuvering, flying out of sight, disappearing, or in some cases exploding. In the great majority of events no sound was reported, which became one of the central anomalies in the scientific review.
The sighting tables use standardized columns: date, time, number of objects, observer reliability, location, apparent direction and altitude, course, color, duration, sound, shape, apparent size, speed, and a classification or explanation field. The dominant color across the event record is green or yellow-green, with orange, blue, white, and red variants appearing in smaller numbers. Durations cluster around one to five seconds. Geometric paths were predominantly horizontal or slightly descending.
The Green Fireball Phenomena: Scientific Analysis
The scientific core of the document rests on the work of Dr. Lincoln LaPaz, Director of the Institute of Meteoritics at the University of New Mexico, who served as the primary scientific consultant to the OSI investigation. LaPaz conducted triangulation studies on several events and formulated a systematic comparison of the observed phenomena against known meteoritic behavior, identifying at minimum ten distinguishing characteristics that set the green fireballs apart from conventional meteors.
The January 30, 1949 event was the most widely witnessed. Observed at 17:54 MST, its real path was calculated to range from approximately 12 miles altitude at appearance to 8 miles at disappearance, covering roughly 143 miles at a velocity of 12 to 14 miles per second, making it "among the brightest observed in the last quarter-century." Despite that magnitude, no detonations or acoustic phenomena were recorded by any of the several thousand witnesses across New Mexico and Texas -- an absence LaPaz considered decisive, since meteorites penetrating to such low altitudes invariably produce violent noise.
The December 12, 1948 "Starvation Peak" event was the subject of the most rigorous quantitative analysis. LaPaz personally observed the fireball and triangulated its path from simultaneous observations near Bernal, New Mexico, and from within the Los Alamos reservation. The computed path was nearly horizontal, approximately 25 miles in length, at an altitude of 8 to 10 miles, with a velocity of 8 to 12 miles per second. Color tests in laboratory conditions matched the emission wavelength to copper salts in a blowpipe flame at approximately 5,218 Angstrom Units. The December 20, 1948 event added an anomaly of a different kind: two red lights reportedly trailed the green fireball in fixed relative position -- the only event in the record where a sharp change of direction was also observed.
LaPaz enumerated the distinguishing characteristics with precision. Green fireballs followed nearly horizontal paths, whereas genuine meteors rarely move horizontally. They appeared at altitudes of 8 to 12 miles, far below the 40-plus miles typical for meteors. Their velocities were much lower than conventional meteorites but far higher than any known human aircraft or rocket. They appeared instantly at full brightness rather than building up as meteors do. They left no particle trail of sparks or dust. They produced no sound even at close range. Their colors were a vivid green not characteristic of ordinary meteoric material. They were geographically confined to a triangle bounded roughly by Los Alamos, Las Vegas (NM), and West Texas, despite nationwide observation capacity and observation conditions that were average or better elsewhere. Maximum sighting frequency occurred around 20:30 local time -- coinciding with neither the meteor peak at 03:00 nor the meteorite-fall peak at 16:00. And sightings showed an anomalous concentration on Friday through Monday, with a distinct decline in the final two months of the reporting period coinciding with a simultaneous rise in "flying saucer" reports in the region.
Dr. Edward Teller's Physics Assessment
On February 16, 1949, Commander Richard S. Mandelkorn of USAF Research and Development at Sandia Base convened a conference at Los Alamos. Dr. Edward Teller provided the key theoretical physics analysis. Working from the observed velocity of approximately 8 miles per second, Teller calculated that a material object of sufficient mass to produce the observed light output would generate a shock wave easily audible 10 kilometers from the source. No sound was recorded. His tentative conclusion: "They are not material objects passing through air. Should look to electronics and optics for explanation rather than in the field of hydrodynamics."
Conference participants noted it was "almost incredible that a large object such as a guided missile or unmanned vehicle could pass through the atmosphere at a height of eight miles at a velocity of seven to eight miles per second without producing loud noise which would have been audible to observers." Dr. N. E. Bradbury, Director of Los Alamos Laboratory, demurred on the electronics hypothesis, noting it would create "many more difficult problems" than it solved.
Dr. LaPaz separately raised the possibility of Soviet origin, noting that the primary sighting window of 17:00 to 23:00 MST corresponded to 07:00 to 13:00 in the Ural region of the USSR, and that missiles with the observed velocities could travel from the southern Urals to New Mexico in under 15 minutes. He stressed: "If guided missiles are NOT of U.S. origin, intensive systematic investigation should not be delayed." Dr. Joseph Kaplan of the Scientific Advisory Board confirmed: "Frankly, I don't know of any U.S. experiments that would result in the appearance of these unconventional objects, and neither does Von Karman."
Dust Collection and Chemical Analysis
A parallel physical investigation ran alongside the visual sightings program. Following a green fireball over Socorro, New Mexico on July 24, 1949, Dr. William Crozier and Ben Seely of the New Mexico School of Mines conducted systematic dust collection using impactment equipment processing air at approximately 34 liters per minute. The collection medium used a glycerin-gelatin mixture with rubeanic acid (dithiooxamide) additive to identify copper, nickel, and cobalt.
The results were unusual. First collections on July 25, approximately 13.5 hours after the fireball, yielded copper particles including at least one showing a positive copper reaction before ammonia treatment, indicating a soluble copper compound. An extended eight-day series revealed spherical cobalt-rich particles approximately 12 microns in diameter -- described as unique in Crozier's experience -- with descent calculations suggesting they would have had to fall from roughly 20,000 feet in 41.5 hours if of meteoritic origin. Critically, five-micron copper particles recovered at 127.5 hours post-fireball could not have descended from above 10,000 feet in that timeframe under normal atmospheric conditions, implying some form of downward forcing.
LaPaz's interpretation was cautious but pointed: "If future more detailed work shows that numerous copper particles found by Crozier and Seely ARE indeed floating down from green fireballs, then the fireballs are NOT conventional meteorites. Copper is one of the rarest elements found in meteorites. No known case in which even the finest particle of copper has been reported in dust collection supposedly of meteoritic origin."
A subsequent B-25 aircraft collection run on August 8, 1949, 41 hours after a Kirtland Field event and reaching 23,000 feet maximum altitude, found no copper, nickel, or cobalt particles attributable to anything other than surface-origin material.
Camp Hood, Texas: Coordinated Sightings Near a Nuclear Site
The Camp Hood section is among the most operationally detailed portions of the dossier. Beginning March 6, 1949, military guards at Site Baker ("Q Area") at Killeen Base, Texas -- a nuclear weapons storage site -- reported a series of unexplained lights at night over multiple consecutive days. Individual sighting reports include witness names, service numbers, azimuth and elevation readings, object color, duration, and the consistent absence of sound.
The March 17, 1949 incident stands apart. An alert guard of the 2nd Armored Division had been pre-positioned to test earlier observers' reliability and to fire calibration flares. Before that exercise could begin, eight separate lights appeared in a defined pattern. Coordinates were measured by dual sightings from tank-mounted sights, timed at the instant of each sighting. Three lights appeared north of the Q Area; five appeared south of it, forming what investigators described as a rough circular pattern around the target area. Captain Horace McCulloch, the Assistant G-2, personally observed three of the lights. An ordnance check confirmed that no pyrotechnic devices had been issued or used in the area for months. The FBI's San Antonio office, the Office of Naval Intelligence, Air Materiel Command intelligence, and the Commanding General of Sandia Base were all notified.
What The Record Supports
DOW-UAP-D017 establishes that a sustained series of unidentified aerial phenomena was observed by credible military and scientific witnesses near multiple U.S. nuclear-weapons installations over approximately 18 months. The OSI formally concluded that "a logical explanation was not proffered with respect to the origin of the green fireballs" and that "the continued occurrence of unexplained phenomena of this nature in the vicinity of sensitive installations is cause for concern." Dr. LaPaz's expert assessment was unambiguous: the December 12, 1948 fireball was "definitely non-meteoric," and most of the others probably were as well. Dr. Teller's physics analysis independently ruled out conventional material objects traveling through atmosphere at those altitudes and velocities. The dust analysis produced anomalous copper, cobalt, and nickel signatures for which no satisfactory meteoritic or surface-origin explanation was provided.
What the record does not establish is the identity, origin, or nature of what was observed. The Soviet guided-missile hypothesis was raised explicitly by LaPaz but was not confirmed. The electronics-and-optics hypothesis from Teller was noted but not developed. No physical debris was recovered. Systematic follow-on investigation using aircraft, spectrographic instruments, and high-altitude collection -- recommended by LaPaz in his May 1950 summary -- was not completed within the scope of this file. The phenomena documented here remain unresolved and unidentified.