(An archive of prev articles in this series is available at   
   ).   
      
      
   EXECUTIVE SUMMARY:   
   - We find that "fingerprints" of the motion of certain solar system   
     moons are also found inside certain UFO observations made on Earth   
     some days later.   
   - It turns out a small number of moons seem to robustly explain or   
     predict day to day variations in certain UFO activity.   
   - Given the list of R2 "explanation power" statistics for the   
     different cases we can ask what quality of the relevant moons   
     explains the link found. It turns out in a natural way those moons   
     suspected of having oceans are the ones that are found to predict   
     UFO activity on Earth some number of days later.   
   - The speed of propagation of change of position of moon to UFO   
     observation is around 1 AU per day -- a reasonable "average"   
     interplanetary speed.   
   - The moon with the best explanation power is also the one with the   
     largest suspected ocean in the solar system -- Ganymede.   
      
   I re-ran an old study on comparing "fingerprints" of various moons   
   with the noise hidden inside certain UFO reports. The mental model of   
   a "fingerprint" is that traces of one type of data are often visible   
   in a statistically robust way somewhere else if 2 systems somehow   
   "came into contact" at somewhere.  E.g. if an airliner crossed the US   
   and overflew Louisiana then tiny variations in air temperature and   
   precipitation will "imprint" on the flight characteristics of the   
   aircraft and provided measurements can be made precisely enough --   
   e.g. if they can be repeated many many times and averaged out -- those   
   tiny variations should be detectable in e.g.  the arrival time of the   
   aircraft at its various destinations.   
      
   In the case of UFO's we're dealing with reports mostly from average   
   folks who may or may not understand what they are seeing. But given   
   enough data even some tiny part of it that may correspond to what   
   Congress is now tending to insist are "non human created aircraft" (if   
   they are human-created Congress has ruled that case is handled by   
   another department but not the Pentagon's UFO department) should have   
   traces of places they have been and maybe even link back with some   
   kind of "home location".   
      
   With the aid of a fairly straightforward software that has enough   
   smarts to be able to run statistical tests and interpret the results   
   we can check off each "home location" against different types of UFO   
   reports (the NUFORC database can be easily divided up into about 250   
   types of object by shape, color, time and location seen, direction of   
   motion, and certain other keyword-based characteristics e.g. whether   
   they were observed interacting with a military aircraft).   
      
   We can then ensure each type of UFO tested and each possible "home   
   location" assumed result in highly robust predictions by using an   
   appropriate combination of statistical tests that ensures the number   
   of "false hits" is below 1 in 1000. As part of the output from the   
   procedure we not only see that some home location seems to imprint   
   some part of itself on unusual objects that are later seen on Earth,   
   but we can determine what kind of lag (in the case here that is   
   limited to whole number of days) gets the best match, as well as a   
   summary "match percentage" that describes what fraction of the   
   relevant UFO reports seem to vary in direct parallel with a parameter   
   related to the "home location".   
      
   The next step in the process will check whether there is some   
   characteristic of the relevant "home locations" that somehow is   
   peculiar to those that relate strongly with some kind of UFO activity   
   seen in Earth.  The s/w will attempt to explain what is unusual about   
   "home locations" versus all the other places that might have turned up   
   as significant but didn't.   
      
   As usual with these studies there is plenty of noise to go around so   
   we do expect some spurious results that don't mean anything.  But the   
   overall pattern should show up something statistically significant   
   after allowing for remaining noise in the system. A crude method we   
   shall use here simply will discard results that are furthest away from   
   the common baseline of all results, provided that no more than 25% of   
   the results are discarded.   
      
   So the first part of the results will line various moons up with   
   various types of UFO activity and determine whether there is a strong   
   statistical link between some parameter of the moon's motion and that   
   activity, what kind of lag/delay best matches the 2 things up, and   
   what% of the activity is finally explained by the relevant   
   motion of the moon in question.   
      
   The summary table I have here looks like:   
      
   Moon	Param	Lag	UFOtype	Filter	R2   
   		(days)		(sd's)	(% y explained by x)   
   503	angsep	11	pale	o1	0.30654435   
   502	defillu	17	pale	o1	0.17788930   
   501	satx	6	pale	o1	0.17229421   
   701	satpang	18	pale	o1	0.14159575   
   504	satx	15	pale	o2	0.13129106   
   801	saty	8	pale	o1	0.12317738   
   601	angsep	2	pale	o1	0.11928895   
   606	saty	2	pink	o1	0.09388933   
   902	satx	12	grey	o1	0.08963227   
   703	angsep	10	gold	o1	0.06135808   
   901	satpang	5	pale	o1	0.06120898   
   704	angsep	7	pale	o1	0.04220981   
   705	angsep	17	gold	o1	0.04151140   
   802	saty	16	grey	o2	0.03456327   
   602	defillu	15	Cross	o1	0.02846995   
      
   The "moons" are numbers from JPL's Horizons system. 5xx are Jovian moons,   
   6xx Saturnian; 7, 8 and 9 resp Uranus, Neptune and Pluto.   
      
   The "Param" column gives the parameter of position or movement that   
   was found to be present in the relevant UFO activity. "Angsep" is the   
   angular separation in arcseconds between the center of the moon and   
   its primary, as seen from Earth. SatX and SatY are the angular distance   
   in the "X" and "Y" directions in arcseconds between the moon and its   
   primary. defillu is the fraction of moon that is not illuminated by   
   the sun. and satpang is the counterclockwise position of the moon with   
   the primary's N pole as 0.   
      
   The "UFOtype" column is the section of NUFORC data -- typically a   
   color or Shape of the objects in question. For this study only the   
   NUFORC data from 2006 onwards was used. This corresponds with the   
   period when a web report system was created leading to a 10x increase   
   in reports as well as a change in the character of some of the objects   
   reported.   
      
   The "Filter" column shows how much outlier data was discarded to   
   obtain a statistically robust result. The robust result must pass a   
   T-test and Rank test both at better than 95% significance.  Items   
   marked with "o1" discarded up to 25% of the data to find a significant   
   result. "o2" lines discarded only 5% of the data to get a significant result.   
      
   Finally, the R2 column estimates what% of the day-to-day   
   variation in UFO activity matches the day-to-day variation in motion   
   of the relevant moon however many days earlier. We see e.g. the   
   "angular separation in seconds" of moon 503 (Ganymede) from its   
   primary matches about 31% of the day to day variation in UFO's   
   described as "pale" 11 days later. Typically it's believed only 1-10%   
   of UFO reports are anything other than mis-identified mundane   
      
   [continued in next message]   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)