disappearing stars -- an AI takes a look at the numbers

[MrPosti...@kymhorsell.com]

In my wanderings I came across a couple papers reporting odd things
seen in the sky via ground-based telescopes. The interesting aspect
was the plates being examined dated back to the early 1950s when
things like Starlink or even Sputnik had not screwed up the sky. Yet
what was reported from a couple different groups was lines of "dots"
and other formations that seemed to look like highly reflected
earth-orbiting "satellites" seemed to be spotted from time to time
over the years.

Typical scenarios involved 2 plates where careful examination found
several bright points on the older plate yet the same positions in the
newer plate were empty.

"Careful" includes checking for faults in the emulsion or processing
that may have accidentally drawn a line of dot on the negative.

One paper reported ~100 of these "disappearing stars" and a later
paper reported ~700 of them.

As usual, attempts to contact the authors of any of the papers have
resulted in 0. Some of the papers are a few years old and since the
pandemic I have to presume people have gone onto new careers at
Twitter. Or something.

So I threw the problem of finding the data and explaining what was
going on to a couple of AI-based programs I run from time to time.

Gathering the data was fairly straightforward. The s/w latched on to
the Mt Palomar Sky Surveys. The first one started around 1950 and ran
for 10-20 years. There was a 2nd Survey starting in ~1980 but,
apparently, that was never finished. There was another attempt more
recently, but that seems to have left no imprint on a public server
the programs can find.

But the 1 and 2 surveys allow us to look at mns of stars on 16x16
arc-min images and compare the state of most parts of the sky from the
1950s with the 1980s.

I already have some tools to look for "things" in images from the old
Kaggle days (kaggle.com/kymhorsell1) and using techniques like "strong
statistical features" (aka SURF) the programs managed to isolate about
150 objects that could be seen in the 1950s survey but had apparently
disappeared by the 1980s even allowing for some proper motion that may
have taken a star off the edge of a plate and onto an adjoining one.

The SURF parameters allow the s/w to build up a small expert system
that not only uses the position of the center of the "star" on the
plate but various qualities relating to how the individual pixels in
the blob are arranged -- a kind of optical fingerprint. The s/w can
determine whether there is a density gradient inside the blob, and it
keeps track of regions on the images that are (extra) bright as well
as circular regions that are extra dark. The number of pixels also
gives an indication of the object's brightness and the spacial
distribution from the object's center pixel can give an idea of
whether it is an extended object and/or is interacting with a nearby
object that is too close for the relevant telescope to actually
resolve. It is also possible to determine from the pixels -- given
some of the exposures were upto 90 mins -- whether the object was
moving in a way not entirely consistent with a star. We might go into
that in a later post. :)

As well as "stars" that seemed to appear on earlier plates but later
were not present in the expected positions, the opposite was also
observed -- sometimes objects appeared on later plates that were not
there in the 1950s. Of the couple 100 "differences" noted in the
collection of images snaffled from the Palomar collection about 1/3
were of the "appearing star" type -- the rest were the regular
"disappearing" kind.

So with a collection of dates when an object was seen or not seen, the
location of the star at some point, and the various photometry
parameters the image processing dredged up we can run the usual massive
search of various data-files in my database to see whether any known
weather, planetary, demographic or other data on file varies on a
statistically similar way to any of the "disappearing star"
parameters and, if so, which are closer and maybe the "most likely
explanation" of what is going on.

And the top results off the rank are as follows:

Data-series Lag R2
(y)
Mwind-GA 5 0.88518841
Mmaxwindseg150 5 0.88297516
Msduah_globe6NHLand 3 0.88092115
Mmaxhail-LA 4 0.86669351
Mwind-FL 4 0.86234448
Mmaxwind-GA 5 0.86232464
Msduah_mtSoPol 3 0.85737059
Mmaxuah_mtUSA49 3 0.85393912
Msduah_mtNH 3 0.84748392
Mforestpct 3 0.84629941
Msduah_mtSoPolOcean 3 0.83581867
Mminuah_mtNoExtOcean 5 0.83540929
Mwind-AR 4 0.82180789
Mneptune-zh 5 0.81655511
Mmaxwind-TX 3 0.81102430
Mmaxneptune-r 3 0.80945403
Mmaxhail-NJ 0 0.80611332
Mhail8 4 0.80494563
Mhail-WY 3 0.80367691

The s/w has determined that any data series that has an R2 > .8 is
effectively in the "best explanation" bucket. So the above list is not
really an ordered list -- all items in it are "equally likely
explanations" the predict the coming and going of the objects in question.

We immediately see most of the explanations seem to involve the
weather someplace on planet dirt. The "wind" items are NOAA-reported
windstorms seen over the years across the US. The "max" prefix
indicates the s/w took the maximum of monthly numbers to produce an
"annual maximum" series from the original data. Similar for "min" and
"sd" (the stddev of the months or an est of annual
variability). Series with a numeric prefix refer to a specific
month. I.e. "hail8" is number of August hailstorms reported to NOAA,
year by year.

In the middle of the pact is one oddball -- "forestpct". This is the
surveyed% of the planet that is still covered in
forest. Obviously this number has been going down for several decades,
but after WWII it was going up a little as people relaxed and started
planning their future families.

And, finally -- and the only thing really connected with things "out
there" -- the position of Neptune seems to have some predictive power
for when "stars disappear". The "-r" series is Neptune's distance from
the sun. The "-zh" is Neptune's position in AU above the ecliptic.

So apart from "probably related to earth weather" the AI programs
point out a possible link with Neptune. Other planets and moons also
turn up way down the list (there are 1000s of items in the list) but
Neptune is in the "best explanation" section -- the other things are not.

And finally, finally we might ask are any other (cough) odd things
maybe connected with "stars" popping up on telescope images some times
and not other times. Given some of the objects imaged *seem* to be
moving relative to the stars as indicated by a "directional" parameter
of the arrangement of the pixels on the image.

We can check down the list a bit and find:

Data-series Lag R2
(y)
Mufo-green 4 0.75836594
Mufo-green2006 4 0.75836594
Mcanadaufos 4 0.63167820
Mufo-MT2006 3 0.60122600
Mufo-Flash2006 5 0.55730722
Mufo-notlight 1 0.51173148
Mufo-SC 4 0.49904489
Mufo-OH2006 3 0.49441492
Mufo-Fireball2006 3 0.47760874
Mufopct-CA 4 0.47042070

Well. Interesting, maybe. Some types of unusual objects seen in the
skies over the US and Canada *seem* to predict the dates some of these
stars come and go. But they mostly seem to relate to things that are
likely "normal" astro-related things like fireballs, green "things"
flashing across the sky (maybe fireballs, green lightning, aurorae, e.g).

Destroying that simple notion is the one item tagged "notlight" which
means all kinds of UFOs reported over N America that have nothing to
do with looking like a light, orb, flash or fireball. I.e. black
triangles, spheres, circles, cylinders, cigars, diamonds, ... well you
know the list.

Finally finally finally there is some mention of California where the
observatory that took the images is located. It seems the pct of ufos
of the total reported by Californians also predict statistically
strongly (all the above items are statistically robust in that they
must pass 2 different tests at the 90% level even to get listed) but
the strength of the predictive power is not in the top category like
the position of Neptune or the temperature of the avg monthly atm temp
over North Hemisphere Land (3-10km above MSL).

More tinkering with the s/w will be on-going.

--
"Nothing in life is to be feared, it is only to be understood.
Now is the time to understand more, so that we may fear less."
- Marie Curie

Unidentified aerial phenomena I. Observations of events
B.E. Zhilyaev, V. N. Petukhov, V. M. Reshetnyk
Main Astronomical Observatory, NAS of Ukraine,
Zabalotnoho 27, 03680, Kyiv, Ukraine
[...] We present a broad range of UAPs. We see them everywhere. We observe a
significant number of objects whose nature is not clear. Flights of single,
group and squadrons of the ships were detected, moving at speeds from 3 to
15 degrees per second. Some bright objects exhibit regular brightness
variability in the range of 10 - 20 Hz. Two-site observations of UAPs at a
base of 120 km with 2 synchronised cameras allowed the detection of
a variable object, at an altitude of 1170 km. It flashes for one hundredth
of a second at an average of 20 Hz. [...]
An object contrast makes it possible to estimate the distance using
colourimetric methods. [Objects with 0 albedo] are observed in the
troposphere at distances up to 10-12 km. We estimate their size from 3 to 12
meters and speeds up to 15 km/s. [...]
[Astronomers in Ukraine have undertaken their own independent survey
of objects they see flying over the Kyiv region at speeds around 15
km/sec. They are watching the daytime sky at the zenith and in front
of the moon. They see many objects -- some bright and some dark,
different sizes. They travel often singly but sometimes in large
groups. They report brightness is linked with speed. The spectrum
of bright objects is reportedly not reflected sunlight. Objects
have been spotted inside the atm upto ~10 km but also out to ~1000 km
above the earth, travelling up to ~1000 km/sec. They are not likely
anything sent by Russia or any other country].

How big a deal is NASA's new UFO study?
[image] An unidentified flying object, as seen by a US Navy jet.
Space.com, 18 July 2022
In early June, NASA announced that it's commissioning an independent study
on unidentified aerial phenomena (UAP), as UFOs have recently been rebranded.
The intent is to move the scientific understanding of UAP forward, said
Thomas Zurbuchen, NASA's associate administrator for science.
"NASA believes that the tools of scientific discovery are powerful and apply
here also," Zurbuchen said in a statement (opens in new tab). "We have
access to a broad range of observations of Earth from space - and that
is the lifeblood of scientific inquiry. We have the tools and team who can
help us improve our understanding of the unknown. That's the very definition
of what science is. That's what we do."
The UAP study team will be led by astrophysicist David Spergel, previously
the chair of the astrophysics department at Princeton University.
"Given the paucity of observations, our first task is simply to gather
the most robust set of data that we can," Spergel said in the NASA
statement. "We will be identifying what data -- from civilians,
govt, non-profits, companies -- exists, what else we should try
to collect and how to best analyze it."
[The investigation of why NASA has missed something obvious to many
for the past 65+ years will come later. But it argues that classical
approaches of Organised Science might draw another blank on UFO's].

Nation's Largest Center for Historical Records on UFOs to be Established in
New Mexico
The Debrief, 18 Nov 2022
The largest historical archive of records on unidentified aerial
phenomena in the nation will be unveiled in New Mexico in the coming
years, according to a group of leading UFO historians and archivists.
The National UFO Historical Records Center (NUFOHRC) will be "dedicated
to the preservation and centralization of UFO/UAP information in the
United States," according to a press release announcing the new
non-profit organization.
[Researcher and historian David Marler] expressed appreciation for the
renewed public interest in the UFO subject seen in recent years.
"What I find though is that, despite the ever-increasing amount of
people that are looking at the UFO or UAP subject, not a lot of
people--a very small percentage--are interested in the history," Marler
told The Debrief.
"Everyone wants to know what's new," Marler says, "and part of that is
due, I think, to the framework by which the intelligence community
[and] the media in recent years has framed the discussion; this `new'
discussion."
"Everyone seems to be looking at 2004 moving forward," Marler told The
Debrief, noting that "we know that this history is diverse and rich and
stretches many decades back, if not even further."

Video Shows Navy Pilots' Encounter With Possible UFO
The Daily Beast, 14 Nov 2022 07:45Z
In an op-ed published in The Washington Post Fri, Christopher Mellon, a
former deputy assistant secretary of defense for intelligence in the Clinton
and ...

60 Eerie Declassified UFO Photos From The '60s and '70s
MSN, 08 Nov 2022
UFOs sightings happen fairly frequently (at least according to UFO sighting
witnesses), but in the 1960s and '70s, alien crafts were spotted almost
every...

'World economy would collapse if govt UFO data was released', warns
lobbyist
Daily Star, 07 Nov 2022
Stephen Bassett, a lobbyist in Washington and head of Paradigm Research
Group, reckons the US govt is secretly holding onto UFO data for fears of
the...

Chilling cockpit audio captures moment Brazilian fighter jet encountered UFO
as pilot screamed "what is this?"
The Sun, 09 Nov 2022 15:41Z
CHILLING cockpit audio captures the moment a fighter jet pilot encountered a
UFO as he said "It's not airplane, what is this? ...

[The Aussies are not the only ones to bet on flies crawling up a wall:]
New report shows best dates and times to see a UFO in California and Fresno
KMPH - Fresno-Visalia on MSN.com, 08 Nov 2022 21:14Z
A new report now shows the best dates and times to see a UFO in California
and Fresno.BetCalifornia.com just released their ...