Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

first cut: searching for stellar light curves in you-know-what

5 views
Skip to first unread message

MrPosti...@kymhorsell.com

unread,
Jul 27, 2022, 7:44:01 PM7/27/22
to
The TESS project is yet another space telescope that looked for
various things -- e.g. planetary transits -- in stellar light curves.
Various sub-projects have looked at parts of the enormous data stream that
came off the telescope and I've zeroed in on DIAmante that looked at
mostly dim M-class stars mostly in the S Hem.

The idea of looking at this data is to determine whether the pattern
of lighting and dimming in somewhat remote stars also shows up in
certain phenomena we have been seeing here on Planet Dirt for the
past 70+ years.

If such patterns do turn up it doesn't prove anything necessarily, but
it ("would" :) raises the question why do phenomena seen here flying
inside the atm and in the oceans as well as LEO and other areas of
nearby space (the whole thing now has become "trans-medium"), why do
these things have any imprint of a pattern of stellar magnitude for
some star or stars quite far away? How can this happen?

I've started downloading yet another giant dataset from the STSCI.
Unlike some other projects DIAmante holds a big catalog of dim stars
that have been viewed at 12 frames per second over several weeks and
produced an ASCII output rather than a FITS file. The FITS section of
my various projects is on hold until I finally manage to extract
ASCII tables from a growing list of FITS sources. Direct output as text
pushes that tedious work further back on the stove. Thanks, whoever. :)

My AI assistants have now tinkered up a simple pipeline that processes
light curves from DIAmante and checks whether any "phenomena" contain a
beyond-chance signal that looks the same. Each DIAmante light curve
(each one is actually a set of related light-curves from the same
part of the sky processed in several different ways) is compressed into
daily data and time series of the relevant "phenomena" are searched for
at least 20 repeats of each DIAmante curve.

It was rather a long-shot to start this search but the results to date
show that, yes, the sighting of unusual things in the Earth's atm --
mostly over N America -- do contain large chunks of beyond-coincidence
repeats of light curves from stars that otherwise are not very
remarkable. It turns out the light-curves for stars we all know and
love -- e.g. Proxima Centauri -- are a little harder to come by
anyway. (I'm not up to trying to extract data tables out of PDF
or other plots of datapoints; that kind of exercise is seldom worth the
trouble of digitising the image).

While the processing is still under development initial results are
encouraging. Of the 10 initial DIAmante light curves 1 in particular seems
to crop up many times. A dim 8th mag star at Dec +8 deg and RA 118 deg
was observed by TESS for about 3 weeks and boils down here to a sequence of
23 numbers that show the average normalized flux for consecutive days
sometime around 2015, probably originating from a dim star 10-20 years
before that.

The data looks like:


JD Avg normalised flux
for each 24 hrs period
observed
1468 1.00169
1469 1.00095
1470 0.998064
1471 0.994287
1472 0.99603
1473 0.998925
1474 1.00019
1475 1.00447
1476 1.00472
1477 1
1478 0.998654
1479 0.999023
1480 1.00158
1481 1.0011
1482 1.00204
1483 1.00216
1484 0.996709
1485 0.994412
1486 0.997462
1487 1.00009
1488 0.999736
1489 1.00072
1490 1.00068

With 23 numbers that essentially are a pattern of "above normal",
"below normal" we might expect to find this pattern maybe 1 time in a
mn (i.e. 1 in 2^23) in any other sequence of numbers. Maybe the
odds get shortened because many phenomena are periodic. But it's hard
to explain that certain types of sightings reported over the past 70+
years have this sequence of ups and downs repeated at least 20 times
inside them.

The sightings in question are recorded by the NUFORC. The organisation
makes its dataset totally available to all researchers without fee or
registration. The organisation classifies individual reports by color
and shape, by location and time of day. I have further added some
classifications based on keywords in the summary comment for each
report. This end up producing about 250 different types of phenomena.

We not only find the above unusual pattern at least 20 times in one
type of phenomena, which seems to be highly against the odds. But we
find the same pattern 20 times in many different types of
phenomena. The initial processing finds around 90 types of phenomena
encode the same series of up and down variations multiple times,
apparently highly statistically significant. You can essentially
predict future variations of the phenomena using this past pattern of
variations -- not exactly, but "skillfully" and far beyond anything
that could be expected by just guessing or using a fixed "average" value.

The reason this is so interesting at this preliminary stage of
light-curve examination -- the fact light curves from distant stars
seem to occur in so many different types of sightings fits in exactly
with other work that is at a more advanced stage. Examining planetary
data around the solar system has found many connections between the
position and movement of the various moons, asteroids, comets and
planets with certain sightings -- but most of the major categories of
the phenomena get absilutely no hits. It's as if sightings reported
on earth are either related entirely to earth-bound phenomena --
unlikely given there is data showing it comes and goes from LEO and
further out in space -- or it is more related to objects much further
out in space than e.g. the main asteroid belt between Mars and Jupiter
or the moons of the outer planets.

Why we find stellar light curves inside many types of UFO sightings is
a matter for only speculation. One immed possibility is that the
phenomena in question are somehow related to the goings-on in the
relevant distant solar system(s). Maybe at some point in their past
history they "came through" or "nearby" that system and the variation
in stellar brightness -- possibly related to major flare activity --
left their mark on the movement of the relevant objects. Or maybe the
phenomena for unknown reasons use a space telescope to see the changes
in brightness in that remote system and adopt a timetable of activity
based on that data. For some reason.

But one of the beauty part of using AI-based investigation techniques
is the ability to demand the system tighten up the elastic bands to
make it even more certain the patterns being discovered are really
really beyond a statistical fluke -- or (worse) somehow an artifact of
all the processing that's been carried out on the relevant data -- in there.

The s/w can introduce additional checking that can eventually be
tightened to make the patterns undetectable in the target data. At
that point we can assess whether the checks are justified or are too
extreme. If they filter out other patterns we know we should be
finding then we know the interesting patterns are "actually" present
in the target data.


--
"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

Hunting UFOs for the Government
BBC, 21 Jul 2022 11:37Z
Secretariat (Air Staff) was an official govt position hunting for
UFOs. For his new BBC 5 Live series, Different, Nicky ...

Pentagon establishes office to track UFOs in space
The Dept of Defense has created an office to track unidentified objects in space and air, under water, or even those ...
Space.com19h
[The data shows interactions between UFO's and "military aircraft"
strongly suggest they see them coming from deep space and pre-position
for the chase. Not to mention nightly chases involving light aircraft
and lights in the sky in sem-rural Victoria].

House committee advances legislation including deeper look into govt
UFO reports
Fox News, 22 Jul 2022 01:34Z
The House Intelligence committee voted to send the Intelligence
Authorization Act for a full vote, including provisions that ...

How a UFO `got into a DOGFIGHT with a US fighter plane for 30 minutes in
400mph chase that left his jet radioactive'
The Sun, 21 Jul 2022 16:26Z
A US fighter pilot was reportedly left shaken after he spent 30 minutes
duelling in the sky with a white shape that appeared ...

Pentagon renames UFO office, expands mission to include `transmedium' objects
Defense News, 20 Jul 2022 22:53Z
Congress held its first hearing on UFOs in over a half centure. During the
hearing, lawmakers questioned Pentagon officials ...
[All Domain Anomaly Resolution Office -- AARO. Sounds serious!]

New research ranks Washington as top state for UFO sightings
YAHOO!News, 19 Jul 2022 18:36Z
The team at online jigsaw site looked at data from the National UFO
Reporting Center to find the total number of UFO ...

Will new NASA study move the needle on UFO research?
Space.com, 18 Jul 2022 10:40Z
NASA announced that it's commissioning an independent study on unidentified
aerial phenomena (UAP), as UFOs have recently ...

R Kym Horsell

unread,
Aug 31, 2022, 2:01:19 PM8/31/22
to
In alt.ufo.reports MrPosti...@kymhorsell.com wrote:
> The TESS project is yet another space telescope that looked for
> various things -- e.g. planetary transits -- in stellar light curves.
> Various sub-projects have looked at parts of the enormous data stream that
> came off the telescope and I've zeroed in on DIAmante that looked at
> mostly dim M-class stars mostly in the S Hem.
...

The programs are fishing around in the TESS dataset to find something
interesting. I've put the latest version of the output up on
<kymhorsell.com/UFO/movie.mp4>.

There are now 4 belts showed in parallel each 10 deg wide at
Dec -65 -55 +55 and +65.

The "time" is the usual number of Julian days (in 1/10ths of days)
from the TESS origin at 1 Jan 2015.

Interesting trashing starts around t=1800.

The plot is still very crude and we're iterating to a nice pretty
colored density plot. But a couple weeks to go.

In the meantime I'm starting to learn about the data.

It seems the telescope concentrates on different parts of the sky
at each "sector" of the dataset. A "sector" is essentially some
sequential number of its 2-week orbits between the E and M.

So looking at the movie at this point sees first some activity in
one hemisphere, then the other, then shifting back and forward.
Only occasionally does the scope point in a direction where it gets
stars in each hem so each band of the plot has data at the same time.

But, anyway, it's starting to show patterns that look like shadows
falling first here, then there, then back again which is what
the stats programs are saying when they compare the variations in
brightness of each 10x10 deg patch of the sky TESS has seen in
the past 4y with UFO activity seen in the skies over the US and Canada
over the same time (day by day resolution).

I noticed a few of the old customers managed to pick up the earlier
versions (two so far) of the MP4 but someone got an error.
The server I'm using sometimes has some buffering problem whereby
when a file on the web page part of the disk space gets updated
the server doesn't see it for some hours and maybe reports "file not found"
or similar. Sometimes you have to hit "refresh" a few times and the
server gets the message and re-reads the disk directory and gets the
contents right for a change.

R Kym Horsell

unread,
Sep 2, 2022, 11:52:54 AM9/2/22
to
In alt.ufo.reports R Kym Horsell <k...@kymhorsell.com> wrote:
> In alt.ufo.reports MrPosti...@kymhorsell.com wrote:
>> The TESS project is yet another space telescope that looked for
>> various things -- e.g. planetary transits -- in stellar light curves.
>> Various sub-projects have looked at parts of the enormous data stream that
>> came off the telescope and I've zeroed in on DIAmante that looked at
>> mostly dim M-class stars mostly in the S Hem.
> ...

The AI s/w has been burning the midnight oil and have started producing
movies of density plots of TESS images.

A sequence of 2k x 2k bits by 2.5 hrs TESS images if boiled down to
a small 20 x 20 pixel tile on the final plot and colored in
depending on the avg anomalous flux over the original TESS image.
The "anomaly" is relative to the average of the whole image set
of that particular part of the sky.

The current display shows the whole sky. Each grid square is
a given 10x10 deg section of the sky as seen from TESS at the
time of image capture. The TESS data has been translated back to
a standart earth-centric RA and Dec hopefully independent of the
position or orbit of TESS at the time.

The positions of the outer planets may be relevant to interpretation
of the movie. Neptune over the time range of the movie is at +ve Dec
around RA 40. The other outer planets incl Pluto are bunched
in the range RA 300-360 at -ve Dec with Pluto as usual at the extreme
inclination of maybe -20.

The movie I've uploaded shows some interesting features.
The colors are unfortunately chosen from a default palette in
my old version of GNUPLOT. But white (+1) indicates the avg brightness
of that 10x10 deg part of the sky is 1 sd above the long-term avg value
for that square. Black (-1) is 1 sd below the long-term avg.

Some areas are seen to "twinkle" as if stars in a co-ordinated way
are brightening and dimming around the same section of the sky.
There are also some features where dark lines or dark regions appear
and fade as if a "whole lot of somethings" got between some stars
TESS was watching over a rather large section of the sky.
Much bigger than 1 10x10 deg tile.

Seemingly you can not only see individual tiles vary in brightness
that statistically seems identical to time-shifted observations of
UFO activity reported across the N Am, but we seem also to see
"lines" of dark and bright that suggest the lateral movements
of dark or bright objects across the sky as seen by TESS.

The s/w is set to update the movie on the web page every now
and then as processing proceeds. Even on a pocket supercomputer
it takes a while to crunch down TB's of images into a 5-min vga movie.
Execution, as they say, is proceeding.

The old movie is at <kymhorsell.com/movie.mp4> now showing
line-plot versions for 6 bands in the sky.

The new movie -- being updated as each group of 1000 frames comes
off the assembly line -- is at <kymhorsell.com/movie2.mp4>.

0 new messages