Newsgroups: alt.paranet.ufo,alt.ufo.reports
From: kymhorsell@gmail.com
Subject: ghost rockets (2/n)

[uploaded 52 times; last 17/10/2024]

Ghost rockets are back. :)

  Pilot reports UFO, says missile-like object flew over plane during
  flight to Phoenix
  USA Today, 27 Feb 2021 23:26Z
  A pilot at American Airlines radioed Sun that an unidentified
  object flew over their jet during a flight while they were ...

A good link for some back story is:
<https://www.ufoinsight.com/ufos/waves/mystery-ghost-rockets-scandinavia>


While some intel people take the problem of (at least) what seem to be
unauthorized (conventional) missiles shooting around nr the flight
paths of passenger aircraft *very* seriously, authorities have long
dismissed the whole phenomena as mistaken sightings of meteors.

Indeed, the data shows sightings of "long thin missile-like" UFO's tend
to pile up in Nov and Dec aka "meteor season".


From the NUFORC data we can create a list of "ghost rockets" from
approx the late 1930s. For this effort we'll accept as a GR anything
reported as cylindrical or cigar-shaped and also include reports
including at least 2 of the keywords long, lengthy, thin, slim,
slender, missile, rocket, contrail, or exhaust.

Broken down by decades the avg monthly sightings look like:

Decade	Avg GR per mo
(The data has been adjusted to take into account NUFORC's web
reporting protocol that went into effect in early 2006).
1930 	11.95
1940 	11.95
1950 	12.5789
1960 	20.111
1970 	25.0274
1980 	17.1781
1990 	19.5337
2000 	21.7343
2010 	18.3417
2020* 	19.4167
* not complete

Which seems to show a more or less steady drum-beat of sightings
across N America with maybe an up-tick in the 60-70s and another one
around the 00s.

But the confounding part of the problem of GR comes when you line up
the number of sightings per month against the seasonal meteor
activity:

Meteor activity per month (avg numbers per hr):
Month	/hr
1	563
2	569
3	569
4	647
5	650
6	650
7	779
8	788
9	823
10	830
11	853
12	1126
(Source: wikipedia).

Total GR sightings 1930- for each month:

Month	Number of GR sightings since 1930
1	202
2	186
3	210
4	268
5	271
6	427
7	474
8	412
9	336
10	386
11	295
12	245

When we run a regression of monthly meteor activity against GR
sightings we get:

(Log transform for X and Y).
y = 4.84252 * x^0.624746
beta in 0.624746 +- 0.77917  90% CI
alpha in 1.57744 +- 5.13091 
P(beta>0.000000) = 0.911598
calculated Spearman corr = 0.510490
Critical Spearman = 0.504000 2-sided at 5%; reject H0:not_connected
r2 = 0.17436746

Data:
Month                       #meteors      #GR       model-estimated #GR
			    (sightings/hr)
1                             563          202      253.187
2                             569          186      254.869*(est +1sd wrt obs)
3                             569          210      254.869
4                             647          268      276.168
5                             650          271      276.967
6                             650          427      276.967*(-1sd)
7                             779          474      310.133*(-1sd)
8                             788          412      312.367
9                             823          336      320.964
10                            830          386      322.667
11                            853          295      328.224
12                           1126          245      390.398*(+1sd)

Which shows there is a strong link between GR sightings and meteor
activity. Both usual tests -- a T-test on the \beta and a rank test
comparing the ordering of data by X and by Y -- show a better than 90%
probability the link is not due to chance. Together they argue a high
likelihood meteor activity is related to sightings of GR.

Given the regression produces a classic "power law" the model shows
for a doubling of meteor sightings/hr between say Mar and Dec there is
to be expected a 54% increase in ghost rocket sightings.

So the case is convincing that GR sightings involve meteor sightings
in some cases.


But the R2 above should warn stats-heads there are plenty of GR
sightings that seem not to involve mistaken sightings of annual meteor
shows. The R2 is 17% meaning month-to-month changes in number of
meteors/hr track only 17% of month-to-month changes in sightings of
GR. Assuming "all" GR are meteor sightings misses the 83% of GR
sightings that don't seem to be related to meteors.


We can run another test by comparing GR sightings against supposedly
validated sightings of meteors. The American Meteor Society maintains
a list of fireball sightings across N Am starting around 2005.  
(See <https://www.amsmeteors.org/fireballs/fireball-report/>).

We can check month-by-month (or day-by-day if needed) how well
sightings of objects thought to be fireballs track sightings of
objects thought to be cylindrical UFO's with contrails that sometimes
are reported to drop 180 degree turns.

But we have to be careful. Like the UFO data there are various
"problems" with the fireball data.  In the case of NUFORC sighting
data there was a big methodology change in early 2006 when the group
created a web report form.  Suddenly it was much easier to report a
UFO and about 10x more reports came in. We have to allow for this
using a program that finds multiplying UFO reports before Mar 2006 by
9.8 optimizes a certain statistic showing data before 2006 are not
"similar" to data after 2006.

In the case of the AMS fireball data it seems a change happened around
2011. Amazingly the same s/w finds numbers before 2011 need to be
adjusted by a factor of 10.3 to make data before 2011 "look like" data
after 2011. A similar factor. We might suspect the AMS also may have
intro'd a web report form -- this time around 2011.

Making the relevant adjustments to the 2 datasets a time-series
regression finds:

(AUTO CORR CORRECTION; estimated rho = -0.271987)
y = 0.007305*x + 1.70192
beta in 0.007305 +- 0.00611713   95% CI
alpha in 1.70192 +- 0.443871 
P(beta>0.000000) = 0.989885
calculated Spearman corr = 0.190643
Critical Spearman = 0.306000 2-sided at 5%; accept H0:not_connected
r2 = 0.10319132
Durbin-Watson d = 2.546464
d > 4-dL (2.458421):  Negative auto-corr at 5%

Binned data:
Bin label         av #fireballs/mo    av #GR from    model-estimated #GR
		  from AMS            NUFORC 
      2008.79               93.06            3      2.38173
      2008.96               62.04            1      2.15513*
      2010.88               103.4            4      2.45726*
      2011.21                   9            2      1.76767
      2011.46                   6          1.5      1.74575
      2012.29                  12      2.33333      1.78958
      2012.38                   5            2      1.73845
      2012.54              10.255         1.75      1.77684
      2012.62                  15            1       1.8115
      2012.79                  25            1      1.88455*
      2013.38                  33            1      1.94299*
      2013.46             20.5644            3      1.85215*
      2013.62                  24            1      1.87724*
      2014.21                  27            2      1.89916
      2014.29                  19          2.5      1.84072
      2014.54                  26            2      1.89185
      2014.96                  32          3.5      1.93568*
      2015.12                  34            2      1.95029
      2015.21                  46          2.5      2.03795
      2015.29                  30            1      1.92107*
      2015.38                  18            4      1.83341**
      2015.46                  29            3      1.91377*
      2015.71                  45            1      2.03065*
      2015.96             31.0156            3      1.92849*
      2016.38                  16            2       1.8188
      2016.46                  44            1      2.02334*
      2016.54                  58            4      2.12561**
      2016.79                  70          1.5      2.21327
      2016.88                  81            3      2.29363
      2017.29                  35            2       1.9576
      2017.38                  37            1      1.97221*
      2017.54                  36          1.5       1.9649
      2017.71                  48            3      2.05256*
      2017.79                  73            3      2.23519
      2017.88                  99            1      2.42512*
      2018.04                  55            3       2.1037*
      2018.21                  57            2      2.11831
      2018.46                  40            2      1.99412
      2018.62             51.4667      1.33333      2.07789
      2018.71                  56          1.5        2.111
      2018.79                  69            2      2.20597
      2018.96                  54      1.33333      2.09639
      2019.46                  38          1.5      1.97951
      2019.54             41.2291            2       2.0031
      2019.62                  61      1.33333      2.14753
      2019.96                  86            1      2.33015*
      2020.04                   4            2      1.73114
      2020.62                  65          1.5      2.17675
      2020.62                  65          1.5      2.17675
      2020.88                 125            3      2.61505
      2020.88                 125            3      2.61505
      2020.88                 125            3      2.61505


In this s/w the data is binned/combined/smoothed. This has the effect
of losing some of the high-frequency noise in the data.  I.e. any R2
will be an over-estimate of how much of the Y's are explained by the X's.

Months that seem to have similar #fireballs and #GR are put in the
same bin.  The average of the points in the bin is computed and only
these averages are passed to a time-series sensitive regression.

In this case the TS regr found a -ve serial-correlation.  It tends to
be the case when the number of fireballs and/or GR sightings go up in
one month, the next month they "retract" slightly.  This kind of
behaviour is seen in situations where we're sampling from a finite
population. If you shoot and kill a few rabbits one month then the
next month there are slightly less rabbits around.

Again, the procedure is quite sure fireball sightings have *something*
to do with GR sightings. For each additional 100 fireballs sightings
across the US there are around .7 more GR sightings.

But the R2 statistic (which is -- remember -- an OVER ESTIMATE because
of the binning) shows that fireballs only explain about 10% of the GR
sightings. 90% seem to be something else.

Taken together the data seems to be pointing at a connection between
meteors and ghost rockets. It is a "strong" relationship in that we
are quite sure it is there. But we are also quite sure a large
fraction of GR are not related to seasonal meteor shows or fireball sightings.

There is something else there.

In later posts we'll try to track down "affinities" GR have for
certain locations in the world.

It turns out like many other types of UFO long cylindrical objects
sometimes with contrails seem to vary as local conditions in quite
remote parts of the world (usually the polar oceans) on pretty much
the same day.  When "flying conditions" seem to be good in select
locations -- good visibility, thin ice, no hurricanes, etc -- GR
sightings seem to be higher than when flying conditions are not so
good. The same locations turn up again and again in this kind of
calculation and many of the locations also turn up at beyond-chance
levels on lists of locations with a long history of "very odd happenings".

--
New AI `Deep Nostalgia' brings old photos, including very old ones, to life
The Verge, 28 Feb 2021 20:57Z
An AI-powered service called Deep Nostalgia that animates still photos has 
become the main character on Twitter this fine Sun, as people try to 
create the ...

[Breaking Up Is Easy To Do:]
Iceberg larger than NYC breaks off from Antarctica
New York Daily News, 27 Feb 2021 23:55Z
It's a cold world. A giant iceberg broke free from an Antarctic ice shelf 
Fri and is now floating in the far south Atlantic Ocean. 
At an estimated ...

The One Gut Microbe An MD Says Is Crucial For Longevity + 3 Foods It Loves
mindbodygreen.com, 26 Feb 2021
By now, we know that gut health is inextricably linked to the functioning of 
many of your body's other processes-your immune system, mood, skin health, 
you ...

'We tried to warn people': Inside a Brazilian city's COVID-19 crisis
ABC News, 28 Feb 2021 18:46Z
Thought by many to be the first city in the world to overcome the pandemic, 
Manaus has become a cautionary tale of what can happen when politics and 
science fail, leaving the novel coronavirus to spread almost unmitigated.

Russia launches Arctic weather satellite 
Spaceflight Now, 28 Feb 2021 09:59Z

WHO investigation may have finally found coronavirus patient zero
BGR, 27 Feb 2021 21:00Z
The World Health Organization will soon publish the findings of its official 
investigation into the origins of COVID-19. Investigators spoke with the man 
currently ...

This family have been observing rain for 100 years, and they do not have any 
good news
ABC Southern Qld, 27 Feb 2021 at 9:10pm
A Queensland family honoured for 100 years of service to the Bureau of 
Meteorology as rain observers say the past 10 years have been the driest on 
record.

Rooster kills owner with cockfight blade strapped to its leg
ABC News, 27 Feb 2021 20:10Z
Police in southern India briefly held the rooster before sending it to a 
poultry farm after its owner bled to death from knife wounds to his groin.

Robots bring futuristic, 'digital-savvy new agriculture' opportunities to 
farming
ABC Landline, 27 Feb 2021 20:21Z
A Queensland couple's decade-long push to lead the world in the production 
of agricultural robots has received a multi-mn-dollar funding boost.

Even by Pentagon terms, this was a dud: The disastrous saga of the F-35
Salon, 27 Feb 2021 13:50Z
The military-industrial complex spent $2 tr building a "flying Swiss 
Army knife." Now it's been shelved.

Some Earth Life Could Already Survive on Mars
Universe Today, 27 Feb 2021 04:21Z
Mars' surface is a harsh environment for life. But life on Earth is 
notoriously resilient as well. No one is quite sure yet how microbes from 
Earth would fare on the ...

Forecasters are considering an earlier start to hurricane season -- here's why
WKMG News 6 & ClickOrlando, 27 Feb 2021 01:29Z
Next month, the World Meteorological Organization will be holding its annual 
hurricane committee meeting, during which weather officials will review and 
retire ...

Around the world, regions with Mediterranean climates are drying out, and 
making bushfires worse
ABC Weather, 27 Feb 2021 02:16Z
Scientists believe drought and fires hitting south-western Australia are 
connected to drought and fires hitting Mediterranean climates around the world.

[RECORD!]
Watch 2020's hurricane season unfold in a mesmerizing four-minute timelapse
The Verge, 26 Feb 2021 17:52Z
This week, NASA released a grim four-minute timelapse of the 2020 Atlantic 
hurricane season, a mesmerizing display of last year's record-breaking 
string of ...