Newsgroups: alt.ufo.reports
From: MrPostingRobot@kymhorsell.com
Subject: cheap ufo detector -- environmental noise

[uploaded 23 times; last 10/09/2024]

I noticed something 5-10 days back that seems to have borne some kind
of strange fruit. I was out watching the sky one night.  As usual
since maybe Sept the sky was at least 1/2 clouded over (normally at
this time of year it would be clear -- we will get back to what is
going on there a bit later when I fully digest what a little AI model
is telling me). I has going the usual looking-in-all-directions at
once thing and just happened to be looking south when a yellow light
popped up in the middle of the sky. Maybe a little less bright that
Sirius, but pretty bright. Within a second, maybe less, some warbling
alarm started going off somewhere to the east.

Surely this could not be a coincidence. I've noticed before when I go
out to watch that oftentimes alarms of different kinds seem to be
going off randomly around the place. You can "often" is seems hear
them off to the E and sometimes to the S which by coincidence is where
a Federal Mint is located.

So I wondered if these alarms -- whatever they are -- might be getting
tripped by stray magnetic fields generated by someone/something.  The
NUFORC keeps track of "sightings" reported by dozens of MADAR units
scattered across the US. The MADAR boxes detect variations in local
magnetic fields and the sightings numbers over the past few years of
operation show a very strong similarity to other UFO sightings
numbers.  Some people that host MADAR stations report they sometimes
go out to look at the sky when the thing goes off and do see unusual
things flying around at least some of the time. My own number
crunching finds there is also a tendency to go off when there is a
major quake in the region, and possible when there are larger than
normal thunderstorms.

Some time during the Big Pandemic Lock-down -- almost 300 days here in
SE Aus -- I cooked up a mike to listen to the local noises. At first I
was following sirens running around the local streets and highways to
try and gauge how bad the situation was. But with a bit of s/w
tinkering I could detect planes and helicopters going over and dogs
barking at night. So I started keeping a data log of these events. And
I added in a couple of the kinda of alarms I've been hearing just for
good measure.

The log of "car or burglar alarms" at month-by-month granularity looks like:

Month   Number of alarms detected
2021.12 1
2021.79 1
2021.88 1
2022.12 1
2022.21 1
2022.79 1
2022.88 1
2023.04 3
2023.21 2
2023.29 2
2023.38 1
2023.46 1
2023.71 1
2023.96 1
2024.04 2

Month	Number of times mass dog barking detected at night
2021.12 3
2021.29 1
2021.46 1
2021.71 1
2021.79 2
2021.96 2
2022.04 3
2022.12 1
2022.29 2
2022.46 1
2022.96 1
2023.04 1
2023.29 1
2023.46 2
2023.54 2
2023.71 1
2023.79 2
2023.88 1
2024.04 1

Date	Number of times prop airplanes going over at night
	close to  the 1/2 hr (aka "patrols")
2021.04 2
2021.54 1
2021.71 1
2021.88 1
2022.21 4
2022.29 2
2022.38 3
2022.46 3
2022.54 2
2022.62 3
2022.71 3
2022.88 1
2022.96 1
2023.04 1
2023.12 1
2023.21 13 <-- don't know what happened here; maybe I adjusted the
		hardware or software and suddenly detections went up
		10x Or maybe "patrols" kicked into high gear for some
		other reason
2023.29 22
2023.38 18
2023.46 20
2023.54 26
2023.62 17
2023.71 48
2023.79 31
2023.88 25
2023.96 25
2024.04 18


Date	Number of choppers going over at night
2021.04 1
2021.21 1
2021.29 4
2021.38 1
2021.54 3
2021.62 4
2021.71 10
2021.79 5 <-- Pandemic lockdown maybe ended around here
2021.88 10
2021.96 20
2022.04 28
2022.12 26
2022.21 45
2022.29 37
2022.38 24
2022.46 21
2022.54 25
2022.62 11
2022.71 24
2022.79 21
2022.88 11
2022.96 12
2023.04 6
2023.12 22
2023.21 18
2023.29 24
2023.38 20
2023.46 22
2023.54 24
2023.62 26
2023.71 42
2023.79 30
2023.88 22
2023.96 16
2024.04 20

And, finally, sounds of bats flying overhead. I live nr a stand of
very tall trees that in the past few years have started hosting a
small group of big fruit bats -- 1m wingspan or more. When these
things fly over your head at night you know about it.  Not only do
they screech at each other, just the "whoop whoop whoop" of their
wings gives them away if closer than 10m or so.

Date    Number of (persistent) bat sightings at night
2021.04 1
2021.38 1
2021.46 1
2021.79 1
2021.96 1
2022.04 2
2022.62 1
2022.96 3
2023.04 1
2023.12 4
2023.46 2
2023.88 3
2023.96 2
2024.04 5

So the question is first -- are these things possibly related to weird
things flying around the sky?

Well of course this wouldn't be a post if there was nothing there.

When we do careful time series regressions of various UFO types as
reported in the (mostly) US we find the best links with the above data
are as follows:

UFOtype       Alarmtype      Lag(m)   R2           Beta      90%CI
			     &Transf
pctDisk       sqrtRp         0        0.95396885   1.33594   0.133174
pctcopper     Rp             1        0.92888913   4683.37   604.414
pctDisk       sqrtp          0  y     0.90843208   0.130055  0.018738
pctDisk       p              0  y     0.90842219   0.260102  0.0374772
dawn          chopperp       0  xy    0.85388418   -0.645978 0.0978846
dawn2020      sqrtchopperp   0  xy    0.85387717   -0.322995 0.0489446
NY2020        sqrtRchopperp  0  xy    0.84997413   -0.391126 0.0601927
NY            Rchopperp      0  xy    0.84994832   -0.782268 0.1204
lapRectangle  vRp            0  x     0.83660733   146.587   28.6434
lappale       sqrtvRbatp     1  x     0.82813429   1.30794   0.419832

The first col gives the name of the UFO data used in the regr.  The
2nd is the codename of the alarm data we're testing.  The procedure
allows the ufo data to be lagged by 0 or 1 month to compare US UFO
sightings this month or prev month against this month's alarm
numbers. And the R2 and Beta show the certainty and nominal strength
of the link.

All results pass 2 statistical tests at better than 90% confidence,
sometimes much better.

Even tho there is only a relatively small amount of "alarm data" (in
the case of the "chopper detector" 35 months) we see many UFO types
ostensibly strongly correlate with the different alarm types.  It
turns out the most common UFO types correlate with all alarm types
beyond a reasonable chance. There is "something going on".

So the next step might be -- can we maybe COMBINE these different
alarm types to predict when UFO reports might come in? Can we cook up
a "UFO detector" from the different kinds of env noise we're seeing here?

Of course. And the top 10 kinds of UFO report predicted by various
combinations and weightings of the above alarm data types turn out to be:

Code    UFOtype    Transf  R2           Beta       90%CI
24      Disk       -y      0.40720192   1.13892    0.531961
118     silver             0.38633714   1.13133    0.551958
59      Cone               0.33462632   0.589844   0.387954
144     Triangle   -y      0.31049905   -0.3035    0.175078
35      red        -xy     0.30806194   0.488629   0.283485
41      Light      -y      0.29431215   0.175976   0.105486
2       blue               0.25588455   -2.79451   1.84476
65      yellow             0.25045518   1.05751    0.753551
156     orange     -x      0.22039845   -3.74794   2.72874
55      cigar              0.12889854   2.48944    2.55548

The "code" here is the method used to combine the various alarm data
types, above. These involve operations like smoothing, interpolation,
adding together and scaling. A simple AI performed this work so many
of the methods are "complicated" to explain in detail.

But suffice it to say e.g. the top link found shows monthly UFO
reports registered by the NUFORC can be "41% predicted" by combining
the various elements in the alarm list. For each "unit" of alarm
created in "dataset 24" there were about 1.14 Disk reports (since the
"-y" flag indicates logs were taken of the Disk report numbers it
turns out this means "14% more") to the NUFORC in the same month
(+-.53 90% CI). The link was checked for statistical significance 2
ways as usual and passed the standard T-test and rank test at better
than 90% confidence.

Finally, we can verify the "24" function is predictive rather than
just a "curve fit" by using data science validation methods.  A simple
method uses part of the data to estimate the model then verifies the
REST of the data also obeys exactly the same model.  It turns out for
this "best link" the model validates easily.

Date                    "24" function   NUFORC Disks  Prediction
2021.04                  0.219608           13      8.66536
2021.21                  0.382002            7      10.1643
2021.29                  0.422627            8      10.5393
2021.38                      0.24           21      8.85359*(too low)
2021.54                   0.40818            5       10.406
2021.62                   0.48832            9      11.1457
2021.71                  0.546761            8      11.6851
2021.79                  0.396078            5      10.2943
2021.88                  0.644634            6      12.5885*(too high)
2021.96                  0.741667            3      13.4842*
2022.12                   1.17321           17      17.4675
2022.21                   1.35268           15      19.1241
2022.29                   1.29066           16      18.5516
2022.38                   1.04451           19      16.2795
2022.46                  0.983311           20      15.7147
2022.54                  0.925058           26       15.177*
2022.62                  0.664053           18      12.7678
2022.71                  0.874595           25      14.7112*
2022.79                  0.855946           28       14.539**(WAY too low)
2022.88                  0.713394            7      13.2232
2022.96                  0.884722            8      14.8046*
2023.04                  0.519444           11       11.433
2023.12                   1.24444           17       18.125
2023.29                  0.918203           15      15.1137
2023.38                  0.916155           10      15.0948
2023.46                   1.01111           16      15.9713
2023.54                   1.02244           19      16.0758

y = 9.23041*x + 6.63829
beta in 9.23041 +- 6.63384 (90% CI)
alpha in 6.63829 +- 5.54164
T-test: P(beta>0) = 0.987282
r2 = 0.18431023
train RMSE=5.14288
val RMSE=7.89199
VALIDATES

The procedure uses the first 1/2 of the data to estimate the alpha and
beta. This part passes a traditional T-test -- not that it's needed.
The "root mean square error" between the model predictions and the
observed number of Disk reports each month turns out to be 5.1.  Then
it runs the same model on the 2nd 1/2 of the dataset and finds the
RMSE is 7.9. This is "statistically the same" as 5.1 so the model
performed just as well on the unseen part of the dataset and is proven
to be predictive as well as statistically significant.

It seems choppers going overhead at night, car alarms and bat squeaks
*do* predict UFO's -- even on the other side of the world.

--
Welcome to the very first official UFO hearing in American history
It's a historic day for everybody who has always wondered if we are alone in
the universe. Although there have already been multiple hearings on UFOs or
UAPs, this is the first hearing in which credible witnesses will testify
under oath in front of Congress. All representatives already offered their
initial remarks and gave all three witnesses the chance to make their oath
before the hearing starts. These witnesses are former Commander David Fravor,
former fighter jet operator Ryan Graves, and former Intelligence Official
David Grusch.
-- Marca.com, Wed Jul 26 10:48:24 EDT 2023

'UFO Revolution' docuseries shows UAP flying over military base, 'blows up
decades of conspiracies': expert
YAHOO!News, 13 Jan 2024 09:36Z
A UFO docuseries "blows up decades of conspiracies," said Jeremy Corbell, an
investigative reporter featured in the series, ...

Floating 'Jellyfish' UFO haunted US military base in Iraq for years, says
former US Marine intelligence analyst shown infrared video by colleagues
Daily Mail, 13 Jan 2024 05:05Z
Marine broke his silence on a 2017 UFO that became 'the ghost story' of his
base. The ex-intel analyst stated that 'the ...

  Video shows jellyfish-looking UFO in the sky over Iraq with troops told to
  "hunt it down"
  Irish Star on MSN, 09 Jan 2024 18:10Z
  This viral video, supposedly recorded by the US military in 2018, shows a
  jellyfish-looking UFO flying through the sky in ...

Inside secret UFO briefing in DC: Congress finds 'many' claims about US
government harboring aliens and spaceships 'have merit' after grilling top
US ...
Daily Mail, 13 Jan 2024 08:04Z
The US Intelligence Community's Inspector General, America's top spy watchdog
, gave Congress a top secret UFO briefing Friday ...

  Classified UFO briefing: House members emerge with mixed feelings
  The Hill, 13 Jan 2024 08:03Z
  House lawmakers left a Friday classified briefing on UFOs, referred to as
  unidentified anomalous phenomena (UAP) by the ...

  Classified UFO briefing provided `validity to what's been said': Burchett
  News Nation on MSN, 13 Jan 2024 04:00Z
  Tenn., said Friday's classified UFO briefing "provided some validity to
  what's been said" in priority testimony about UAPs.

  [NOP!]
  U.F.O.s Remain a Mystery to Lawmakers After Classified Briefing
  NYT, 13 Jan 2024 01:26Z

  Coulthart: Classified UAP briefing yields `significant' results
  NewsNation, 13 Jan 2024 01:02Z
  Thomas A. Monheim, the inspector general of the intelligence community,
  briefed the House Oversight Committee after a surge ...

  Why is America's Congress being briefed on UFOs?
  The Economist, 11 Jan 2024 16:25Z
  AMERICA'S SPIES brief the country's lawmakers on sober topics. In 2023 they
  gave classified briefings on artificial intelligence, Israel, Russian and
  Chinese misinformation, Sudan and Ukraine--in other ...

  Secret UFO meeting prompts speculation over what government is hiding
  Metro, 05 Jan 2024 22:18Z
  Retired Navy pilot Ryan Graves said that UFOs are an `open secret' among
  fighter pilots, and shared a third-party account of ...

UFO community grows rapidly after whistleblower testimony makes national
headlines: 'Watershed moment'
Fox News, 09 Jan 2024 22:13Z
UFOs are at the center of a "watershed moment" in American politics as
discussion of extraterrestrial life becomes more ...