Red lights in Tarn-et-Garonne, France

On October 6, 2010 at 8:17 PM, two people observed from outside a building at the Golfech nuclear power plant a phenomenon flying over the plant. It is described as a black triangle with flashing white lights at the corners and a flashing red light in the center. The phenomenon flew at low altitude in a dark and cloudy sky. No noise was heard. A video was made by one of the witnesses (T2) using his mobile phone before the object gradually disappeared in the direction of Saint Loup. An initial investigation allowed GEIPAN to conclude on November 22, 2012 that it was probably an observation of the passage of an aircraft. A new investigation was conducted in October 2013 due to the possibility of accessing the Golfech nuclear power plant site. Several hypotheses (EDF maintenance device, airplane, drone, hoax, PAN D) were considered (see the investigation report). This complex observation case posed a methodological problem for GEIPAN, due to the following elements which significantly weaken the file: - The two witnesses interpret the phenomenon very differently. - T2 gives a statement unfavorable to the witness in his report to the gendarmes. - The video, considered as proof by T1, turns out to be a video of streetlights. Apart from the fact that it is obviously not proof, this video may have influenced the testimony after the fact, if the witness unconsciously brought his account closer to the video at the expense of the memory of his own sighting of the phenomenon. - T2 turns out to be easily influenced if one compares his statement to the gendarmerie and the interview given to C Albiero. Under these conditions, we do not know what value to give to his statements, while also not being able to ignore his testimony. Because of this, GEIPAN assigned a value of 0.6 to the consistency of this case, that is, a lower value than that assigned to an ordinary single witness (0.7). On the other hand, the airplane hypothesis, initially proposed by GEIPAN, suffers from several inconsistencies (see the investigation report); it would only be acceptable on condition of considering both strong distortions of the testimonies, and the existence of a low-altitude airplane not detected by radar or by the plant's surveillance. GEIPAN considers it certain that there was a passage of 3 flashing light points arranged in a triangle around a flashing red light point. The black triangle cited by the witnesses is probably the result of the frequent illusion called "contour" or "shape" illusion: the real object could just as well be a set of 4 independent light sources, or a craft much larger than the triangle described by the witnesses (the airplane hypothesis or other). The altitude is indeterminable as in all cases of observing light points at night: the witnesses here estimate very different altitudes, which is completely normal; it is very likely that the true altitude of the luminous phenomenon is still different. The strangeness of the phenomenon was judged as medium to high (0.55) since the closest hypothesis (airplane) suffers from several elements not conforming to the observation. We find ourselves facing a case of medium consistency and strangeness, on the borders of the 3 categories B, C, and D: a slight difference in scoring on strangeness or consistency would allow shifting into neighboring classes. Because of this sensitivity, GEIPAN asked for the opinion of the GEIPAN expert college, which met on April 12, 2016 in Paris. The opinions were divided as follows: 4 experts voted for a classification in D1, considering that the luminous phenomenon that flew over the plant was unexplained. 8 experts voted for a classification in C, considering that the elements of the file contained too many inconsistencies to be able to give a consolidated opinion. Consequently, GEIPAN classifies this case in category C due to insufficiently reliable information.