Spatial analyses

Home > A Comparative Analytical and Observational Study of North American Databases on Unidentified Aerial Phenomena

So far we have discussed variables that mostly involve, in a way or another, the "time parameter". We have seen how UAP sightings are distributed with Julian date, time range, month and year and how their increase with passing years is extraordinarily correlated with the diffusion of new technological communication means. We have also pointed out how this long-term temporal behaviour is markedly anti-correlated with the geomagnetic field and we ventured some work hypotheses that might explain the reason of this. We have already introduced for a while the "space parameter" showing how the geographical distribution of UAP sightings has nothing to do with geophysical anomalies.

The following section will be entirely devoted to the analytical study of the spatial distribution of UAP sightings in the US states of New York and Connecticut and in the Canadian province of Ontario, using the same databases that were mentioned in the beginning of this paper. Before showing the results of these graphical-numerical studies it is important to make two important considerations (one specific and one general) concerning the accuracy and the reliability of this specific part of the study. These points are as follow:

  1. As it has been already said in a previous section, the data describing Ontario UAP cases are much more complete and accurate [Ref. 32] than the data concerning the two US states. In particular the Ontario Latitude/Longitude charts do not include only inhabited centres but intermediate locations and roads too. Therefore the spatial resolution that can be obtained in this way is sensibly higher than in the two US cases. As it will be seen this offers both an advantage and a disadvantage.
  2. The geographical position at which a UAP sighting is reported is not referred to the UAP itself but to the position of the witness seeing it. Of course the object that is seen in the sky (according to probability theory) most of the times is not just on the vertical of the observer but at several angular height degrees far from it. This means that this angular distance compared to the zenith may be the result of the intrinsic distance of the UAP from that point, especially for objects that are seen very close to the horizon. Apart from the angular dimension of objects that is reasonably easy to resolve in their shape when they are sufficiently close to the observer, in the case of sightings occurring at night the real distance of an object doesn't depend only on its angular height but also on its intrinsic luminosity: this means that a very luminous object can be seen from large distance, while a weakly luminous object can be seen only at short distance. Clearly the observer is not able to evaluate the difference between apparent and absolute luminosity. All this means that data concerning the geographic position at which an UAP sighting is reported are referred to the sighting position itself and not to the real position of the UAP. This is a source of positioning error, which is not so easy to evaluate (mostly for sightings where the NESW directions are not given), but which might reasonably range from 1 to 30 Km or more. This cause of error must be considered implicit within every data point that will be shown in the Lat/Long charts that have been plotted using UAP databases.

Fig. 11 shows the overall distribution of UAP sightings reported from 1949 to 2009, which includes the two US states and the Canadian province all together. As we can see data points mark quite well the shape of these states, except for the less inhabited areas (such as Northern Ontario). Just to reassure hard-cooked "UAP believers" this chart shows no "UAP invasion" and this will be demonstrated in the next sections. Fig. 12 shows UAP sightings reported in the entire Ontario area, while Fig. 13 shows a zoomed high-resolution snapshot of the places where UAP sightings have been reported more often inside this Canadian province.

Figure 11. Geographical distribution of UAP sightings in the US states of New York and Connecticut and in the Canadian province of Ontario reported in the last 60 years (1949-2009). This chart includes 1546 data points, several of which are time repeaters (cities, towns and their surroundings where several or many sightings were reported in the course of time).

Fig. 13 shows the high spatial resolution with which these data have been obtained. These include inhabited centres (73%) and intermediate locations / specific positions at precise roads (27%). Some particularly clustered areas can be easily identified, especially the one between parallels 43° and 44° and meridians 79° and 80°. The Ontario "UAP map" is certainly the most accurate among the three big areas that have been considered in this study. But all of this doesn’t show at all an intrinsic UAP geographic distribution but a pure selection effect due to the population number. Most populated areas are of course those where more UAP sightings have been reported. In fact big cities and towns such as Toronto, Kingston, Ottawa, Oshawa, Etobicoke, Hamilton, London, Peterborough, Scarborough, Windsor, Missisauga, Sudbury, Whitby record each one many sightings in the last 60 years, typically ranging from 5 to 80. Repeated sightings at these same towns and cities cannot be easily shown in this map (where only the locations are plotted). At intermediate locations and specific roads, as it is expected, only one sighting for each is typically reported. Hamlets (of which it is not easy to obtain in this specific case the population number) record only one sighting too.

Figure 12. Geographical distribution of UAP sightings in the Canadian province of Ontario occurred in the last 60 years (1949-2009). This chart includes 599 data points, several of which are time repeaters (cities, towns, their surroundings, hamlets, intermediate locations and roads where several or many sightings were reported in the course of time).

Despite the good details present in the Ontario map of Fig. 13, it is practically impossible to build up a pondered statistics on population vs. UAP sightings. Statistical studies might be limited to the big towns quoted above (of which population number is well known, including the one of smaller towns except for hamlets), but their too spread extension would make this survey uncertain and inaccurate. In order to make a good statistics where the real local frequency of UAP sightings can be derived independently from the population number, we need to concentrate the study on smaller towns or hamlets (whose population number is known for an acceptable number of cases) where more than one sighting was reported during the last 60 years. As it will be seen soon, this procedure can be done quite well only for the cases of New York and Connecticut geographic areas.

Figure 13. The part of the geographical distribution that was most “crowded” of UAP Sightings in the Canadian province of Ontario occurred in the last 60 years (1949-2009).

Let’s now see in detail the "UAP map" (see Figs. 14 and 15) for the cases of the US New York and Connecticut states. Here we have much less spatial resolution and precision, because intermediate locations (between towns) and specific positions at roads are not furnished.

Figure 14. Geographical distribution of UAP sightings in the US state of New York, reported in the last 60 years (1949-2009). This chart includes 794 data points, several of which are time repeaters (cities, towns, and hamlets where some or many sightings were reported in the course of time).

Therefore in these two cases we have to limit our analysis only on UAP sightings reported from cities, towns and hamlets. But here we have an important double advantage: a) the population number of many small towns and hamlets is available very often; b) many of these small centres register at least three UAP sightings in the course of 60 years. As we’ll see soon this will permit us to derive something relevant concerning the real distribution of UAP sightings on a geographical map.

Before passing to the geo-statistics that has been announced before, let’s first see how and how much the population number is able to affect the distribution of data points in these maps. This is clearly shown in Figs. 16, 17 and 18. The geographic distribution of UAP sightings follows very strictly the population number of the areas where they have been reported [Refs. 12, 47, 53]. This behaviour can be verified more quantitatively [Ref. 92] from the graphs of Fig. 19, where it is possible to ascertain a correlation (of various factors for the ON, NY and CT areas) between the number of reported UAP sightings and the population number. This shows once and for all that a "UAP map" taken alone has really no sense if we really want to try to identify some "recurrence areas" in a given territory of the world. All this has been just a test to verify how and if the scientific method and data processing are able to treat properly that which comes out from UAP databases. The answer is negative if we simply limit our analysis to putting data points on a Lat/Long chart. In such a way there is no hope to find anything that is really able to inform us on the true behaviour of the UAP phenomenon that is reported so often. Of course this concept is valid not only for the three North American areas analyzed here but for whatever area in the world. The answer can be positive if we "adjust the shot" by using a subsequent procedure, on condition that data are available in order to achieve a result describing the distribution of UAP sightings independently from "noise factors".

Figure 15. Geographical distribution of UAP sightings in the US state of Connecticut occurred in the last 60 years (1949-2009). This chart includes 154 data points, several of which are time repeaters (cities, towns, and hamlets where some or many sightings were reported in the course of time).

The only way to extract a scientifically reliable information here consists only in "weighing UAP sightings" reported from specific locations. Big towns are not suitable for this kind of analysis because the big extension of their surroundings renders the statistics very vague and inaccurate: in fact telling that, for instance, a UAP was seen in Toronto or New York City means that we are very often in the most total uncertainty (unless NESW directions are furnished by witnesses, but this happens only rarely) because that sighting was reported from an error box that might be of the order of 10.000 square kilometres or more. Moreover here we have also the complication that the location of a UAP sighting almost never coincides with the position of the UAP itself, which might also be quite far from the sighting point, being quite well visible only due to its possibly very high luminosity.

Figure 16. Comparison of the spatial distribution of UAP sightings in Ontario (right) and the population density of this province of Canada (left).

Figure 17. Comparison of the spatial distribution of UAP sightings in the US New York state (right) and the population density of this area (left).

Figure 18. Comparison of the spatial distribution of UAP sightings in the US Connecticut state (up) and the population density of this area (down). The "UAP map" in this specific case doesn’t show only the geographic distribution of sightings but also the amplitude of time repeaters (represented by circles of different size).

Are we in a condition to skip such a severe uncertainty problem? If the data that we need are available we can certainly avoid the problem and try to focus more on what is really important, namely the intrinsic spatial distribution of UAP sightings despite the fact that data regarding this are obtained not from measurement instrumentation but from simple witness databases.

Figure 19. Number of UAP Cases vs. Population Number for US New York state (left), Canadian Ontario province (centre) and US Connecticut state (right). (New York state: 547 data points, Ontario province: 143 data points, Connecticut state: 141 data points).

After verifying that the necessary data are available in sufficient number for the New York and Connecticut states, a search has been done for all the inhabited centres of which population number was available and which have been registering at least three UAP sightings in the last 60 years. Due to the selection effect caused by population, obviously some big towns record up to 212 UAP sightings. Of course this doesn’t mean at all that these cities are preferred by UAP visitation: this is only an effect due to highly inhabited areas.

In order to try to restrict the most probable locations at which a real UAP phenomenon had an incidence, the following criterion of choice has been decided:

  1. Due to the lack of sufficient data concerning the population number of small towns and hamlets and due to the lack of this kind of locations that have recorded more than only one UAP sighting in 60 years, the Ontario area has been excluded by this statistics. Therefore, only the New York and Connecticut areas have been considered.
  2. Only little towns and/or hamlets, which typically have a few thousands of inhabitants, have been considered, due to the fact that bigger towns are too much “spread out” and subject to a high level of inaccuracy concerning the localization of UAP sightings. Of these only the ones that recorded not less than 3 UAP sightings in the last 60 years have been included in the computation.
  3. The following ratio has been adopted: PL = NI / NU, where PL stays for “Probable Location”, NI stays for "Number of Inhabitants" and NU stays for "Number of UAPs". Only the inhabited centres having a PL ratio equal to PL ≤ 1000 have been considered for this statistics.

The selection furnishes the following small towns and hamlets:

Figure 20. Map of UAP sightings in the US states of New York and Connecticut, occurred in the last 60 years (1949-2009). Up. Without statistical selection (948 data points). Down. With statistical selection, where PL ratio is PL ≤ 1000 (32 data points).

Such a statistical screening was intended to identify only the inhabited centres that are characterized by a large excess of UAP sightings compared to the population number. The result of this procedure shows that a possibly real geographic frequency of UAP sightings that is not dependent on the number of inhabitants is limited to very little populated centres, typically ranging from 500 to 5000 inhabitants (see Fig. 20). This reduces to only 32 the inhabited centres that might have been really characterized by “UAP visitation” during the last 60 years, whatever the nature of the UAP phenomenon may be. This number is about a factor 30 smaller than the number of all the plotted NY and CT locations together. A few comparative examples are just shown here in order to show the order of magnitudes of the PL ratio: PL (Toronto) = 30100, PL (Brooklyn) = 23400, PL (Manhattan) = 7700, PL (Utica) = 4600, PL (Pine Bush) = 140. This means that only the last one can be chosen.

The most reasonable interpretations that can be done after this operation are as follows:

  1. UAP sightings tend to be reported much more often from little inhabited centres and at locations that are quite isolated from big towns and cities.
  2. c) As it can be seen from Fig. 20 the Hudson Valley (approximately between meridians 73° and 74.5°) is the most important “area of probability” for UAP sightings in the case of the New York state and the confining Connecticut state. This confirms thorough studies that have been carried out in the past both analyzing witness reports and doing monitoring operations on site [Refs. 6, 14, 15, 26, 58].

Figure 21. Left. Identification of the best probability area where UAP sightings have been reported: the Hudson Valley. Right. Exact geographic position of the nucleus of statistically selected cases.

Fig. 21 shows what might be the “broadened nucleus” of the said area of probability, here located on a World Wind map. This area includes data points (present, among all the other data points, inside the lower chart of Fig. 20) where statistically selected UAP locations are more clustered together: this coincides with an important part of the Hudson Valley.

Figure 22. Up. Frequency distribution of UAP shapes reported by witness from the US states of New York and Connecticut and from the Canadian province of Ontario. Down. Correlation analysis between the three cases.

Home > A Comparative Analytical and Observational Study of North American Databases on Unidentified Aerial Phenomena