THE SOURCE
DOW-UAP-PR096, PR097 and PR099, three infrared video clips from the U.S. Department of War’s PURSUE Release 02, published at war.gov/ufo on 22 May 2026. All three assessed by AARO to U.S. military sensors operating in the U.S. Central Command area. PR096 dated by its uploader to 3 July 2018, PR097 and PR099 to the same date in September 2019, four hours apart.
What this briefing is
Release 02 Briefing 14 read 2020 in the Central Command area as a sustained wave. These three clips put a small but real prequel under that wave. PR096 is from mid-2018, PR097 and PR099 from a single date in late September 2019. They are not part of the 2020 set; they precede it. Read together they say that the documented run of unidentified-object encounters in this theatre starts earlier than the 2020 footage alone shows. The general grounding for infrared footage is in Release 01 Briefing 3.
TL;DR
PR096 is 79 seconds. Its uploader titles it “HH11 03 July 2018 UAPs”. AARO’s description has two areas of contrast entering the lower right of the screen and being tracked by the sensor, with one of them ceasing to be visible at higher magnification. AARO assesses 2018, Central Command.
PR097 is 4 minutes 59 seconds, titled “Hi-Res: [CALLSIGN] Observes UAP on 25SEP19 at 2135Z”. The sensor cycles modes and zoom while tracking an area of contrast, and at higher magnification, between 1:38 and 2:10, the area “appears as several areas grouped together in the center of the field-of-view”. AARO assesses 2019, Central Command.
PR099 is 4 minutes 51 seconds, titled “Hi-Res: [CALLSIGN] Observes UAP on 25SEP19 at 1715Z”. A brief area-of-contrast transit at around 45 seconds, then mostly empty content. The callsign is the same as PR097 and the title’s date is the same, four hours earlier (1715Z versus 2135Z). AARO’s metadata field, oddly, dates the clip 2023 rather than 2019; another small label-versus-record mismatch of the kind Release 02 Briefing 15 set out for PR057.
These are three pre-2020 Central Command clips, two of them by the same observer four hours apart on a single day, that extend the documented timeline backwards. Labelled source tier 2. Their analytical value is mostly the timeline, plus one specific lesson about what a single “area of contrast” can resolve into at higher magnification.
A prequel to the 2020 wave
Release 02 Briefing 14 made the case that the 2020 Central Command clips, roughly eighteen of them, form a sustained operational pattern across one year. These three clips do not extend that wave, but they do show that the pattern did not start on 1 January 2020. The same theatre, the same kind of sensor, the same uploader-language of unidentified objects tracked from U.S. military aircraft, is present in mid-2018 and again on a single date in September 2019.
That is a modest finding. It does not turn the 2018 to 2019 footage into a wave in its own right, three clips across more than a year do not match the 2020 density. But it removes the temptation to read the 2020 cluster as a sudden onset. The encounters were already being recorded earlier. The 2020 wave is the densest patch of a longer, sparser run.
Two encounters on one day, same observer
PR097 and PR099 share two details that are worth pulling out. They use the same uploader callsign string, and they place that observer at two times on 25 September 2019, 1715Z and 2135Z, four hours apart. Read as labels alone, this is one observer recording two unidentified-object encounters on the same day.
The detail to keep proportional is the same one Release 02 Briefing 11 raised about the four April 2021 clips: a labelled “second encounter” by a “same observer” is consistent with one observer who really did encounter something twice, and equally with the same one event captured or filed twice, or with two unrelated transits in a busy theatre. The metadata-versus-title mismatch on PR099, where the title’s 25SEP19 date and AARO’s 2023 incident-date field disagree, is a fresh reminder that the uploader-supplied specifics are not always corroborated. The clean read: two clips that the uploader places at two different times on the same day; one observer link that depends on a callsign string; the rest takes care.
What a single object can resolve into
PR097 carries one specific teaching detail worth carrying forward. Between 1:38 and 2:10, at a higher magnification level, the single tracked area of contrast “appears as several areas grouped together” in the centre of the frame.
That is a familiar shape of observation. A bright blob at low magnification, zoomed in, can resolve into a tight cluster of smaller bright points. Two readings are open. The object really is a cluster, several things close together that the lower-magnification view smeared into one. Or the object really is one thing, and the higher-magnification view is breaking it up by sensor noise, focus artefacts or atmospheric effects rather than because there are multiple. The footage on its own cannot pick between them. The right discipline is to record that “single object” and “cluster” are both readings the same clip can support depending on zoom, and not to pick the more dramatic one by default.
What the file says
They establish that U.S. military sensors in the Central Command area captured and tracked unidentified objects in mid-2018 and again on a single date in September 2019, with PR097 and PR099 sharing the same observer label four hours apart. They establish that the operational pattern described in Release 02 Briefing 14’s 2020 wave did not begin in 2020; recordings of similar encounters precede it by at least a year and a half. They establish that PR097 contains the specific magnification-resolves-to-cluster behaviour, a useful teaching example of how zoom changes what a clip seems to show.
What the file does not say
It does not establish what any of the objects were. Three clips with no visible telemetry resolve no identities.
It does not establish that the two 25 September 2019 encounters are connected beyond a shared callsign string in the uploader titles. Same-observer-twice is plausible from the labels and unverified by the AARO descriptions.
It does not reconcile PR099’s title-date with its metadata date, which disagree. The same label-versus-record mismatch flagged in Release 02 Briefing 15 applies here too.
And it does not rise above tier 2. The pre-2020 set is a small, real extension of the timeline. The pattern is older than the wave alone showed; its identity is still unresolved.
What to watch
Two items stay open: the documented run begins in mid-2018 only because that is the earliest clip here, so any earlier footage a future tranche surfaces would move that line back; and PR099 carries unreconciled dates, 2019 in the title and 2023 in the metadata field, which a corrected catalogue could close.
Read the file. Decide for yourself.
References and further reading
- DOW-UAP-PR096, PR097 and PR099, PURSUE Release 02, U.S. Department of War, hosted at war.gov/ufo
- Presidential Unsealing and Reporting System for UAP Encounters, U.S. Department of War, war.gov/ufo
- AARO UAP Records, All-domain Anomaly Resolution Office, aaro.mil/UAP-Records
- Signals from the Periphery, Release 02 Briefing 14, on the 2020 Central Command wave; Release 02 Briefing 11, on what multi-clip clusters do and do not add; Release 02 Briefing 15, on label-versus-record mismatches; Release 01 Briefing 3, on reading infrared footage