Argentina.Bariloche · Flight AR 674

On 31 July 1995, at night, between 20:00 and 21:00 local time, two aircraft from Buenos Aires were converging a few minutes apart on San Carlos de Bariloche's Teniente Luis Candelaria International Airport: the Boeing 727 of Aerolíneas Argentinas flight AR 674, with Captain Jorge Polanco, First Officer Carlos Dortona, Flight Engineer Jorge Allende and a company pilot riding in the cockpit, and a Gendarmería Nacional Piper PA-31T Cheyenne, registration GN-705, with Commanders Juan Domingo Gaitán and Rubén Cipuzak and a mechanic. The tower told them the city had no power and the airport was running on its generator. On the outbound leg of the approach procedure, the 727 crew saw a light 'at eleven o'clock' and reported it to the tower, which answered: 'OK, received, that traffic should now be forty miles from Bariloche'. Already on final, about ten miles out, the 727 transmitted: 'Yes, we have an object on our right flying in formation with us, and above we can see the other aircraft. The object has two bright flashing lights... I don't know, we think it's a flying saucer'. To the airport chief, Polanco later stated that he saw 'green lights at its ends and a strong orange light in the centre', and over the years he has described a sighting of about seventeen minutes, with the object by the wingtip. The episode became known as the 'Polanco Case'.

The night had two separate blackouts, and they should be kept apart. The first hit the whole city. The head of the Bariloche district of Energía Río Negro, engineer Etchegoyen, told researchers Heriberto Janosch and Virgilio di Pelino on 2 August 1995 that it was caused by a switching error at the transformer station that receives 132,000 volts from Alicurá, which overloaded the two transformers and tripped them; the Bariloche Electricity Cooperative also attributed it to human error, and the local press and the Air Force date it to 20:17 (Janosch placed it between 20:05 and 20:15). That ordinary cause is on record. The second blackout is the one that coincided with the approach: at 20:47, with the 727 on final, the airport's generator went out, and with it the runway lights, and the aircraft had to go around and climb to hold. According to Janosch, the airport chief, Major Jorge Oviedo, found no explanation for that failure because the unit had started normally; decades later, operations officer Daniel García told the Air Force that the generator could not power the runway, taxiway and apron lights all at once, which opens a non-extraordinary cause, though an unproven one. The tower recorder stopped for about four minutes because of the outage, and tower operator Alfredo Blanco attributes the anomalies he saw on his instruments to voltage fluctuations. There is also a statement no explanation fully closes: meteorologist Nicolás Araya declared that at 19:50 he was blinded by a strong yellow light moving at great height behind the tower; the Air Force holds that the time is wrong and proposes an overvoltage flash from an airport floodlight, something it admits it cannot reproduce.

The weather matters too. A storm dropped snow on the city for about fourteen hours, with accumulations of 40 to 60 centimetres according to the local press, and according to the local press the snowfall by itself caused no outage. The aviation report gave two octas of cloud, with thin layers of stratiform cloud and the freezing level at ground level; the pilot riding in the 727 cockpit recalls rather four octas and says they flew the outbound leg in cloud. Those layers of ice crystals are the screen on which, according to the Air Force, the searchlight beam was reflected.

During the approach, when the tower asked GN-705 whether it could see what the 727 was reporting, the answer was 'we still see nothing'; Cipuzak later told Lianza that at that moment 'something dazzled us', and according to the 2018 report a couple of miles further on they confirmed seeing a light to the right of the 727, although the 2025 report states that the Gendarmería crew never saw the light accompanying the 727. The second phase came after the go-around. GN-705, over the airport at about 11,000 feet, saw a light sweep rapidly across Lake Nahuel Huapi below them, heading west, which in 'a strange evolution' shot up at a right angle and stopped in a thin cloud layer; Gaitán would say it 'climbed abruptly, at 90° toward the cloud layer, and stopped there instantly'. The 727, in its return turn at 10,000 feet, confirmed by radio that it had it on its left: the simultaneous sighting from both aircraft that the commission itself regards as the firmest. The light then passed in front of and below the 727, heading south, and vanished. The 727 landed at 21:09 and GN-705 at 21:18. On the ground it was seen by the Cabral couple at Dina Huapi, the tower staff and the meteorologist. Bariloche had no air-traffic control radar at the time, and no photograph of the object has been published: the evidence is testimony and radio traffic.

The Argentine Air Force published its answer in two instalments signed by Commodore (ret.) Rubén Lianza. In the CEFAE's 2018 annual report it took up the hypothesis Janosch had raised in 1995, which the Air Regions commander, Brigadier Horacio Genolet, had already said at the time should not be ruled out: the beam of a skytracer searchlight that Juan Carlos Rivero was testing that night, on its own generator, at the Roket nightclub on the waterfront. A nightclub employee, Sergio Segade, wrote in 2004 that 'even I saw it reflected in the sky that night and realised it was the searchlight', and Roberto Benavente, the 747 captain riding in the 727 cockpit, had the impression, according to Janosch, that they were 'searchlight lights'. The radio transcripts also show that GN-705, through a navigation error, was about 11 miles from the airport when the tower believed it was at 22, and that it flew almost in formation on the 727's right and some 6,500 feet higher. Gaitán, by contrast, told Janosch on television that they were '18 miles away' behind the 727, which the 2018 report itself calls 'a clear contradiction' with Cipuzak's drawing and testimony. In the CIAE's 2025 expansion the commission acknowledged that the searchlight could not reach the area of the eastern phase (its eastward sweep was limited to a bearing of about 74° and its range to about 10 km) so it attributed that phase to two successive stimuli, the Piper's own lights and the searchlight's coloured gels over Dina Huapi, and classed it 'Type B', explained with solid grounds but without absolute proof. It classed the western phase, over the lake, as 'Type A': Rivero signed a statement before a lawyer describing a low sweep westward followed by a 90° climb, and his trajectory matches the drawing Cipuzak handed over in 1996. The weak point is the clock: a timetable circulated by civilian researchers places the end of the searchlight test at 20:31, before the 20:47 go-around, and the commission rejects it by reconstructing, from what Rivero remembered thirty years later, a start around 20:27 and an end around 20:52. Moreover, the commission's arithmetic adds the 25 minutes Rivero assigned to the whole job, moving the equipment included, starting from the end of that move; counted from the 20:17 outage, the test would have ended around 20:42, before the go-around and the lake phase. The commission also stresses that nine of its ten witnesses described 'a light' and only the captain a structure; in that tally, the view of the flight engineer, whom it could not locate, is taken from what the other two pilots reported, and in 1995 Janosch had quoted him describing a very strong green and orange light that came very close to the aircraft at a 'speed very hard to imagine'.

Its country's air force followed it up institutionally, with radio transcripts the Ministry of Defence made public in 2020 at a civilian group's request, and with testimony from two professional crews. Its international fame owes much to the 1999 French COMETA report, which, citing SEPRA, presented it as a case in which the city and airport blackouts were 'directly related to the presence of the object' and in which the other aircraft 'observed the same phenomenon'; that report resurfaced in the Pentagon's release of files in May 2026. Both claims clash with the record: the city blackout has a logged technical cause, and when the 727 reported the object on its right, GN-705 answered that it could not yet see anything.

Calibration must discount the element popular culture treats as decisive, without discarding what remains. The general outage has a documented ordinary cause; the airport generator failure does not have one proven, though it has a plausible one. What remains is a sighting of lights by two crews, with no radar or photograph, for which the Argentine Air Force has published a detailed mundane explanation: a misplaced Gendarmería Piper that accompanied the 727 and the beam of a searchlight whose operator signed a statement. That explanation is strong in form for the lake phase, where the manoeuvre described by Gendarmería and that of the searchlight match in shape and place, although its timing rests on an arithmetic that, redone from the power cut, has the test ending before the go-around; and it is weaker for the approach phase, which the commission itself left as 'Type B' and whose timeline depends on reconstructed recollections.