Within the public record of the 13 March 1997 Phoenix Lights there are two temporally and behaviourally distinct phases that have frequently been merged in later popular accounts. The later, stationary string of lights seen around 10:00 p.m. is the better-documented of the two and aligns closely with a confirmed military flare exercise. The earlier travelling formation, observed roughly between 8:00 and 8:30 p.m., remains more open. Thousands of people across Arizona reported unusual lights that night; contemporary police and Department of Public Safety calls, newspaper reports, and multiple independent witness accounts establish the scale of the event. The residual analytical task is to separate what can be securely matched to known activity from what remains less fully constrained by the public data.
The 10:00 p.m. stationary lights
Multiple videos show a row of bright lights that appear to hang or drift slowly before disappearing sequentially behind the Estrella Mountains southwest of Phoenix. Time, location, brightness, parachute-descent behaviour, and sequential disappearance converge on high-intensity illumination flares (LUU-2 type) dropped by A-10 aircraft of the Maryland Air National Guard operating under Operation Snowbird from Davis-Monthan Air Force Base. In July 1997 a Maryland ANG spokesman confirmed the exercise; a participating pilot, Lieutenant Colonel Ed Jones, later corroborated that he flew one of the aircraft that released the flares. Video triangulation studies further support the geographic match to the Barry M. Goldwater Range. This phase can be treated as resolved to high confidence.
The earlier travelling formation
Witnesses described a V-shaped or triangular arrangement of lights moving southeast along a recognised aviation corridor. Some observers with telescopes resolved individual aircraft. Changing spacing between lights was reported by a subset of witnesses. No clear photograph or video of a continuous solid body exists; many solid-body perceptions appear consistent with the geometric arrangement of lights and the absence of intervening stars under glare conditions. Aircraft associated with the same broader training period and corridor account for the directional movement, the long route, the telescope observations, and the absence of a solid-body image. The residual uncertainty stems mainly from the lack of a complete public flight reconstruction for every light in the early formation and from the volume of prompt solid-body descriptions.
Later testimony, including that of Governor Fife Symington in 2007, must be weighed against the intervening decade of media imagery and discussion. Social and memory consolidation between the two phases is well-documented and contributes to the single-craft narrative that developed after the event.
Comparison of the principal explanations
Military illumination flares (10:00 p.m. phase) Confirmed unit, operation, munition type, time window, geographic match to the Goldwater Range, parachute descent behaviour, sequential disappearance behind the Estrella ridgeline, and multiple confirming statements including a participating pilot. Some residents familiar with flares initially felt the lights looked different; this is a subjective residual. High confidence explanation for the later stationary lights. Key sources: Maryland ANG statements (Capt. Drew Sullins, July 1997); Lt. Col. Ed Jones confirmation (2007); video triangulation studies.
Aircraft in formation (early travelling phase) Lights followed a recognised north-to-south aviation corridor. Telescope observation of individual aircraft. Formation aircraft can appear as one enormous object when lights are unresolved. Steady lights may be mistaken for fixed lights on a craft. Changing spacing reported by some witnesses. Timing consistent with Operation Snowbird activity. Complete public flight reconstruction remains incomplete. Strong / leading prosaic explanation for the early formation. Key sources: Contemporary telescope reports; corridor and landing timing data; Operation Snowbird context.
Single continuous giant craft accounting for both phases Multiple solid-body descriptions (including some recorded relatively promptly) and later high-profile testimony. No clear solid-body photograph or radar track. Size estimates are unreliable. Later testimony is delayed and follows years of media imagery. The two phases differ markedly in behaviour, timing, and location. Not uniquely demonstrated; the single-craft interpretation consolidates two separable events.
The public record establishes a large-scale multi-witness event on 13 March 1997 and provides a high-confidence prosaic explanation for the later stationary lights. The earlier travelling formation is more open but is consistent with aircraft operating in a recognised corridor under the same training programme. Progress on the residual questions depends primarily on fuller reconstruction of the early flight activity rather than on further reinterpretation of the already-public video and witness material.