Analyses temporelles

Home > Une étude analytique comparative et observationnelle des bases de données nord-américaines sur les phénomènes aériens non-identifiés

Le but tactique qui m'a incité à faire cette sélection quantitative fut de vérifer si des tendances significatives pouvaient être identifiées dans l'espace comme dans le temps. Pour accomplir cette tache des paramètres ont été opportunément choisis. Le choix le plus naturel d'un point de vue scientifique fut d'essayer de voir si une tendance quelconque (de n'importe quel type, pas nécessairement linéaire ou exponentielle seulement, mais aussi périodique) était détectée dans un échantillon de données contenant un nombre de cas à la population suffisamment statistique. Ce test commença avec les cas du Connecticut, et en fait le moment comme la durée des observations de PAN rapportées ont été tracés par rapport à la date date julienne, étant confiant que le système de calendrier julien, normalement utilisé en astronomie pour les études de précision de phénomènes célestes variables s1Hiltner, W. A. (Ed.), & G. P. Kuiper & B. M. Middlehurst (Series Eds.). (1962) Stars and Stellar Systems, Compendium of Astronomy and Astrophysics. University of Chicago Press s2Julian Date Converter., pourrait révéler quelque chose après qu'une analyse ultérieure ait été faite. Le résultat final fut que l'analyse de série chronologique s3"Time Series", Wikipedia ne montra aucune périodicité possible, aucune tendance significative de quelque type que ce soit, à l'exception du fait que le nombre d'observations enregistrées augmente clairement avec les années, à la fois en termes de nombre de cas à une date et une heure données et en terme d'une plage horaire élargie (voir figure 1).

Figure 1. Variation of the Time (up, 525 data points) and of the Duration (down, 466 data points) of UAP sightings with Julian Date, in Connecticut, USA, from January 1956 to June 2009. Figure 1

Quelle est la signification de tout ceci ? Ce sera discuté dans une section ultérieure, lorsqu'une telle augmentation du nombre de cas de PAN avec le temps se montrera elle-même de manière plus claire et marquée. Comme il sera discuté plus tard, ce n'est pas dû à une véritable augmentation des cas de PAN avec le temps. L'intention était juste de commencer immédiatement avec des analyses temporelles plus rafinées que des histogrammes normaux, en utilisant la même approche que celle que les astronomes utilisent normalement lorsqu'ils veulent mener des études de variabilité sur certaines classes d'étoiles particulières. Dans ce cas le système de calendrier julien est couramment utilisé comme outil de précision temporelle. Si alors une périodicité peut être identifiée alors une "phase temporelle" peut être utilisée comme variable indépendante, avec une valence hautement prédictive. Mais l'expérience réalisée montra que cela n'arrivait malheureusement pas lorsque la même procédure était appliquée aux bases de données sur les PAN. Comme on le verra plus tard, les observations de PAN signalées dans le Connecticut suivent une tendance temporelle très semblable à celle de New York et de l'Ontario : il n'y avait donc pas de raison convaincante de perdre un temps supplémentaire à appliquer ce type d'analyse temporelle aux 2 autres régions.

Therefore, the globally disappointing results obtained with this first study, where the sample is quite well statistically represented, invited me not to repeat the same test for the Ontario and New York cases. After this choice was done, the analysis was from then on focussed on the choice of more general and classical parameters such as the number of UAP cases as a function of time range, month and year. If we consider the first one of these three statistical checks that were carried out, where the number is counted during the last 60 years, what we obtain is shown in Fig. 2, which shows that cases reported in Ontario, New York and Connecticut are quite well correlated together.

Figure 2. Up. Number of UAP Sightings plotted vs. Time Range for three compared areas: 1) Connecticut state (USA): 543 data points; 2) New York state (USA): 2057 data points; 3) Ontario province (Canada): 969 data points. Down. Correlation between the three cases.

This similarity in the time behaviour of the three cases is not surprising at all and probably tells almost nothing on the intrinsic nature of the investigated phenomenon. The reported behaviour most probably shows a very marked “selection effect” due to the fact that witnesses of UAP sightings are much more numerous in a time range in which they are freer (mostly from work time) to look at the sky and when they are still awake. The higher frequency of cases reported at certain hours at night is probably due to the perceptive effect [Ref. 102] of the witness in general: it is here expected that “nocturnal lights” (of the structured and unstructured kind) in general tend to attract the attention of people much more than daylight anomalous aerial phenomena. But maybe, as it is strongly suspected, the most important component affecting this statistics is due to the fact that witnesses very often tend to misinterpret well known astronomic, atmospheric and aerial phenomena, such as stars or planets that give the impression to “run in the sky” when really the clouds are moving, luminous planets (such as Jupiter, Venus, Saturn and Mars) taken alone or in occasional conjunction, meteors and fireballs, occasional ball lightning or natural phenomena of different kind [Refs. 5, 13, 19, 21, 48, 106], airplanes and/or helicopters seen through particular perspectives compared to the observer, very luminous orbital satellites and space stations, bird flocks illuminated by city lights, single birds (such as owls) occasionally illuminated by street lights, fireworks, military flares, mirages [Ref. 57], moving laser beams hitting the clouds, experimental military aircrafts, Chinese lanterns, cars moving on the top of hills during foggy days, simple cottage lights, and probably several other causes that might create the illusion of seeing an UAP.

A *real frequency* of the UAP phenomenon, intended as a really anomalous phenomenon, seems to be completely buried inside these perceptual and highly deceptive factors. Therefore the extraction of the signal from the noise is extremely difficult because any possible study to evaluate the noise involves many concomitant effects, which might also be different from state to state.

What is a bit surprising is that if we make the same hourly plots just monthly and not extended to all the sightings that were reported during 60 years, the picture changes quite consistently. In fact the previously recorded correlation almost vanishes, and the three areas show a quite different hourly behaviour, despite the fact that here we can anyway see just a “general common increase” during dark hours. This can be seen quite precisely in Figs. 3A, 3B, 3C and 3D, where the Ontario, New York and Connecticut cases are considered together. How to explain this behaviour? Certainly one probable cause should be apparently given by the difference of daylight hours for areas that taken all together range from the parallel 41° (NY) to the parallel 54° (ON): this clearly shifts the time at which sightings are reported, with opposite conditions during extreme seasons such as winter and summer. Nevertheless this effect doesn’t seem to be decisive due to the fact that the most northern areas (northern Ontario) are much less populated than the southern ones: for instance, it is sufficient to compare Long Island, NY area with the most northern border of the Ontario province in Canada. As it will be seen later this drastic difference in population number has a very marked effect on the number of reports of UAP sightings. Conclusively, the difference of daylight is a reasonable factor that must be taken into account per se, but in the case of this specific confining group of states it must be accurately weighed state by state. However as a general impression in this specific case this factor seems not to be important in determining the often marked difference that is seen when the statistics on time range is done month by month.

All this said, it cannot be therefore excluded that the bulk of a possibly real UAP anomaly is hidden inside these last four charts, namely in the residual that comes out from the pondered difference of the number of UAP sightings at a given time slot and at specific months. This is the real reason of the importance of comparing together UAP sightings reported in more than only one state, without limiting the statistics to only one. There is no doubt that the difference of the trend that is shown by the three cases contains in itself something that should be better evaluated. As it was discussed previously, fishing the signal from the noise here is a quite difficult operation. The scope of this study, as it can be seen now, is not to solve this problem once and for all, but to highlight this possible perspective for further studies when it will be inevitably necessary to embark into a multi-faceted analysis of the many accurately weighed factors that contribute in creating the “noise” inside these charts. This further study might maybe shed some light into the evidence of a real UAP phenomenon directly from UAP databases that are opportunely screened and where the individual cases are more carefully and accurately selected and chosen.

Figure 3A

Figure 3B

Figure 3C

Figure 3D

What about now "UAP statistics" on a monthly and yearly base? Concerning the statistics by month (see Fig. 4) we clearly see that summer months are quite markedly favoured in all of the three states. This is logically expected due to the fact that summer months – in particular July and August – are the ones in which people pass their time more often outside due both to the good climate and to a less constraining link with work duties (and a consequent lesser necessity to go to bed earlier in these months). Nevertheless it is possible to notice some quite basic difference among the three considered areas.

Figure 4. Up. Number of UAP Sightings plotted vs. Month for three compared areas: 1) Connecticut state (USA): 543 data points; 2) New York state (USA): 2057 data points; 3) Ontario province (Canada): 969 data points. Down. Correlation between the three cases.

The general trend shows only a quite vague, but not very strict, correlation. For instance if we look at the high peak reached in October by the New York state area and we compare it with the same month in the Connecticut and Ontario areas, we notice a pretty huge difference. The differences with which the curve grows or decreases in the three cases can be noticed also in some other months of the year. These differences, in particular the October behaviour for the New York case, might show the existence of a real “UAP flap” (or in case more than one, at different times) that is just “hidden” inside these curves. In few words some events that have nothing to do with prosaic perceptual factors might have occurred. It must be clarified that such a flap must not necessarily ascribed to extraterrestrial [Ref. 17] and/or interdimensional visitation. It can be caused by “everything”, most probably by occasional flybys of experimental aircrafts [Ref. 27] or to other specific factors that are peculiar of the locations where these events have been occurring. The trend by year (see Fig. 5) shows in all of the three areas a quite strict correlation. This is a very important result and, as it will be noticed soon, it is most probably not due to a real increase of UAP events.

Figure 5. Up. Number of UAP Sightings plotted vs. Year for three compared areas: 1) Connecticut state (USA): 543 data points; 2) New York state (USA): 2057 data points; 3) Ontario province (Canada): 969 data points. Down. Correlation between the three cases.

Which other factor might account for this almost exponentially increasing trend with years? The answer is quite clear: the reason of this is mostly due to the exponential increase with time of the technology of communications. A similar conclusion seems to be suggested by previous extensive studies, including aviation-related UAP reports, dating back to 1930 [Refs. 114, 115, 116].

Figure 6. Up. Number of UAP Sightings plotted vs. Year for the three compared areas. Down. Evolution of diffusion of cell phones in North America. (2008 and 2009 data were not yet available for cell phones at the time of the writing of the present paper.)

Statistics on the technological evolution and diffusion of Internet and cell phones is shown everywhere on internet. In the case of the signalling of a UAP incident, cell phones, whose worldwide diffusion [Ref. 7] follows approximately the same trend as the one of Internet, should be considered more important because a witness can advise (sometimes in real time) an UAP centre from everywhere, including the car from which a given sighting has been just reported. Fig. 6 shows that a quite interesting correlation exists between the diffusion of cell phones in the world and the increase of UAP sightings with years. But this correlation is very sharp only for two thirds of the entire yearly curves. In fact the residuals might hold something relevant to UAPs.

What occurs in the last years seems to show that something that is probably inherent to the “phenomenon” itself is suddenly overlapped and this happens almost exactly in the same way for the New York, Connecticut and Ontario areas. Let’s comment more in detail what happens here.

In conclusion it is highly reasonable to suspect that mostly the UAP number is correlated with the increasing diffusion of cell phones, but also that: A) in some specific years temporal flaps occurred transiently; B) in the last 6 years a sharp change of the constant trend occurred, characterized first by a decrease and then by a sharp increase. These ones might be intrinsic effects due to the UAP phenomenon: in fact this trend correlates NY, ON, and CT areas together quite well.

Another important feature should be noticed if we look at Figs. 5 and 6. Apart from the same trend that the three cases show quite sharply, we notice that the amplitude of the three curves is markedly different. Do UAP pilots prefer a state than another in order to build their “underground bases”? Maybe, but not as expected. The amplitude of the shown curves depends mostly on the number of inhabitants of the New York (19.000.000) and Connecticut (3.500.000) states and of the Ontario (13.000.000) province. In fact, as it will be seen more clearly in a subsequent section concerning spatial analyses, the number of UAP sightings is quite strictly dependent on the number of inhabitants of a given area. And it is also logical that it happens this way. But is all this totally explanatory and exclusive? To give the answer we should compare together the ratio given by the number of inhabitants and the number of reported UAP sightings. Results give the following values:

New York state Ratio: 19.000.000 / 2.057 = 9.237
Connecticut state Ratio: 3.500.000 / 543 = 6.446
Ontario province Ratio: 13.000.000 / 969 = 13.416

Clearly here a smaller ratio is an indication of a bigger impact of UAP sightings (intended as reports and not necessarily an alien visitation). In fact these ratios show that the state that is more affected by the phenomenon is not New York (biggest amplitude of the curve) but Connecticut (smallest amplitude), while the Ontario state (intermediate amplitude) is the one where UAPs have less impact (in spite of the additional reports compared to the other two areas). This clearly shows that, apart from their trend in time and the peculiarities therein (which are what really counts in this study), the numerical amplitude of the histograms and the curves represented in the charts of Figs. 2 up to 6 are just illusory, namely: they are not numerically intrinsic to the signal that we are searching for.

Home > Une étude analytique comparative et observationnelle des bases de données nord-américaines sur les phénomènes aériens non-identifiés