XPost: alt.astronomy   
      
   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   
      
   [continued in next message]   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)