IPBFacebook



POSTS MADE TO THIS FORUM ARE THE SOLE RESPONSIBILITY OF THE AUTHOR AND DO NOT NECESSARILY REFLECT THE VIEWS OF PILOTS FOR 911 TRUTH
FOR OFFICIAL PILOTS FOR 9/11 TRUTH STATEMENTS AND ANALYSIS, PLEASE VISIT PILOTSFOR911TRUTH.ORG


DIGITAL DOWNLOADS

WELCOME - PLEASE REGISTER OR LOG IN FOR FULL FORUM ACCESS ( Log In | Register )

2 Pages V   1 2 >  
Reply to this topicStart new topic
"aa77" Final Approach Ground Speed Determination From The 84rades Radar Data, 84Rades and FDR data mutually INCONSISTENT?

tumetuestumefais...
post Jan 2 2009, 10:26 PM
Post #1





Group: Respected Member
Posts: 1,111
Joined: 7-November 07
From: Prague or France
Member No.: 2,452



Introduction

In 2007 The Pilots for 911Truth obtained the Flight Data Recorder (FDR) data of "AA77". In the same year in autumn John Farmer obtained the 84th. Radar Evaluation Squadron, Hill AFB, UT (84Rades) USAF/NEADS radar data from 9/11.
Booth are extensive data sets and booth are provided with flightpath analyses. - In case of FDR data there are the flight simulations which include the graphic projection of the main flight parameters (as apeed, altitude, bank angle, throttle...); in case of the 84Rades data there are then the flightpath studies of all 4 aircrafts (the Official conspiration theory considers being the planes used in 9/11 attacks) - made as projects in RS3 radar data decoding and screening software. Booth data sets were obtained via FOIA request from National Transportation and Safety Board (NTSB) - which is the main federal institution-agency which should investigate all aviation accidents. Both sets therefore could be considered as the official data about 9/11 events and therefore should be considered as a proof which could also serve for law enforcement purposes - as the 84Rades data in fact did, because they were provided for FBI investigation already 2 days after the tragic events of 9/11.

The purpose of this analysis was originaly solely to determine the range of possible speeds the "AA77" could fly during its final approach to the Pentagon - to determine the speeds of the "AA77" final approach from 84Rades data. For the analysis I've used the last 8 blips on The Plains ARSR-3 radar (PLA) which is located approximately 30 nautical miles (nmi) northwestwest from the Pentagon - so the "AA77" was really well in the range of the radar.
(The real range of PLA radar in the direction the "AA77" was flyiing from was empiricaly derived from how far the PLA radar have "seen" other planes in that southwestward direction - which the PLA radar did even in distance of approx. 190 nmi - although the "AA77" reappears - after a period of ~19 minutes of "invisibility" - on the PLA radar at the range of 147.5 nmi - which itself is weird and casts doubts about the 84Rades data credibility. But the 84Rades data credibility was not the main question of the original analysis).

Originally I posted the speed analysis in a thread in the Study, which was about addressing of the GL "arguments" against the "northern path".
However the results of the original analysis seemed to me be in such a contradiction to the results of the NTSB "AA77" flightpath study derived from FDR, that I decided to make a separate topic, to discuss this weird issue.


Last "AA77" blips from 84Rades and the groundspeed derived

Here are the last "AA77" PLA radar data decoded by USAF rs3 software:


[click the upper right corner to enlarge...or SOURCE]

On the screenshot from rs3 software you can see the data of the last 8 blips recorded by PLA radar. In fact there should be not 8, but 9 blips - the sweeping period of the radar is approx. 11.9 seconds - so there is surely one blip missing - which one can maybe expect, because the plane was probably already in a very low altitude above terrain.
We don't see any radar height in the data, because ARSR-3 radars don't read it, and we also don't see any Mode C altitude either, because the PLA radar didn't read any, most probably because the "AA77" didn't transmit any from the supposedly switched-off transponder.

Now we can project the Lat/Lon blip data into GoogleEarth to measure the distance between them, which will help us to determine the range of speeds the "AA77" could fly at its final approach. ( Blips are numbered with the red numbers, the long numbers are the timestamps of each blip - note they're all 4 hours after the actual eastern time when the events happened - which could be attributed to the use of an universal time by the US army - but this doesn't affect my analysis):


[click the upper right corner to enlarge...or SOURCE]

The projection of the blips beggins with the No.8 - it is the end of the big ~330 turn the "AA77" did while descending. Then it continues with 7. and from 6. is the trajectory already quite very straight towards the Pentagon. I've used the Google Earth measurement tool (not the calculator embedded in the RS3 software) to determine the distances between the projections of the blips. From it and the exact radar timestamps (with accuracy of 0.005 seconds) I was able to count the possible groundspeeds estimations:
I use this simple equation:

(number of seconds in hour/numberof seconds of the flight period)*number of miles travelled=speed in miles per hour
distance blip 7. - 1. = 6.83 mi
timestamps:
7. 13:36:24.715
1. 13:37:35.890
6.83 mi in 71.175 sec. = (3600/ 71.175)*6.83 = 345.46 mph

distance blip 6. - 1. = 5,68 mi
timestamps:
6. 13:36:36.555
1. 13:37:35.890
5.68 mi in 59.335 sec. = 344.6 mph

distance blip 5. - 1. = 4.60 mi
timestamps:
5. 13:36:48.395
1. 13:37:35.890
4.60 mi in 47.495 sec. = 348,7 mph

distance blip 4. - 1. = 3.14 mi
timestamps:
4. 13:37:00.275
1. 13:37:35.890
3.14 mi in 35.615 sec. = 317,4 mph

distance blip 3. - 1. = 1.90 mi
timestamps:
3. 13:37:12.190
1. 13:37:35.890
1.90 mi in 23.7 sec. = 288,6 mph

distance blips 8. - 6. - 1. = 1.98 + 5.68 = 7.66 mi
timestamps:
8. 13:36:12.875
1. 13:37:35.890
7.66 mi in 83.015 sec. = 332.2 mph

distance blips 8. - 7. - 6. - 5. - 4. - 3. - 1. = 0.75+1.22+1.07+1.50+1.26+1.90 = 7.7 mi ("average speed for the "straight period")
timestamps:
8. 13:36:12.875
1. 13:37:35.890
7.7 mi in 83.015 sec. = 333.9 mph

The blip X. is quite weird (although included in the 84Rades "AA77" flightpath study which is included in the original set of the disks provided by USAF to FBI). The timestamp is again the increase of ~11.9 sec. - so if we count then the speed would drop even more considerably.

distance blips 6. - X. = 5.7 mi
timestamps:
6. 13:36:36.555
X. 13:37:47.810
5.7 mi in 71.255 sec. = 288 mph
But this blip really - it could be anything but not a "AA77" position at the given time - so we should exclude it from our analysis.
I call this blip X. - because its position is quite inconsistent with the other blips (which look quite consistent with each other). Some say it's the radar echo not from the "AA77" itself, but of the fireball or smoke from the Pentagon explosion.


Average groundspeed for whole the 330 turn

Now for the illustration and further comparison I will plot the whole 330 turn. The blue points are the projected blips from the PLA radar. The numbers are timestamps shortened just to minutes : seconds . thousands of seconds. The yellow points are projections of the Gibbsboro radar (- GIB - the ARSR-4 radar located ~110 nmi northeast from Pentagon in NJ) blips - which are also available in the 84Rades data set for "AA77" in some cases (They're missing in majority of cases, mostly missing also the important information of ARSR-4 radar type - and that's the height data - but GIB search blips are readable just in the case of the part of the 330 turn at least for the positioning data, which are in good accord with the PLA data...although some of the blips - which are in the "raw" dataset - are "ommited" from the "AA77" flightpath study provided for FBI by 84Rades in their "AA77" flightpath study!...). The GIB pattern clearly copies the pattern of the PLA radar blips, although the average distance difference between the PLA and GIB blips projected for a same sweep time is ~0.3 nmi - which would almost exactly fit into the combined accuracy range as specified by 84Rades in the 84Rades cover letter.


[click the upper right corner to enlarge...or SOURCE]

If we measure the combined distance with the Google Earth measurement tool for the whole above marked path, we come to 31.09 miles. The timestamps are:
beginning: 32:26.360 , end: 37:47.810 -> the time difference: 321.45 seconds.
So we can count average speed for the whole projected path (3600/321.45)*31.09 = 348.18 mph


CONCLUSIONS:

A.

The ground speeds for the last 8 sweep periods of PLA are in the range from 288 to 380 mph - or between ~245 to ~330 knots.


B.

We can also conclude that - if the 84Rades data are real and the measuring in the Google Earth is sufficiently accurate - then the ground speed of the "AA77" during the last 1.5 minutes of the flight didn't exceeded ~365 knots and most probably not even ~330 knots. So then the speed of "420 knots"
- proposed by some - (or even 465 knots as is the last speed in FDR) in scope of 84Rades data looks a bit like a nonsense.
(The FDR average speed for last 44 seconds of flight is >400 mph = ~350 knots for last 33 seconds even ~450 mph = ~390 knots)
[this two conclusions A. and B. are dependent on the accuracy of the Google Earth measurement tool and we can't rely on it - I've put it here just because it was the core of my original analysis]


C.

The average ground speed for the last 8 sweep periods of the PLA radar is 333,9 mph or ~290 knots. But the average speed for the whole 330 turn was ~348 mph or ~303 knots - so there is a high probability, that the "AA77" decreased its speed during the last 1.5 minute of the flight NOT INCREASED - which is in direct contradiction with the FDR data decoded both in FDR files and in the NTSB animation. [and this conclusion is based on relative computing - virtually independent on the Google Earth measurement tool accuracy!!]

D.

The inconsistence between speed pattern derived from the 84Rades data and the speed pattern derived from the FDR is inspectional.


Although we can't be exactly sure about the Lat/Lon positions derived - because the accuracy of the radar positioning is according to 84Rades cover letter within range of 1/8nmi (0.144 mi) - which still can't explain so huge difference between the speeds derived from FDR and speeds derived form 84Rades and at all it can't explain the opposite speed increase/decrease pattern - which is opposite in FDR and 84Rades respectively for the last ~5 minutes of the "AA77" flight. This leads me to conclusion that either the FDR or 84Rades data (or booth) are at least partially a fake.


The accuracy assesment of the PLA radar

From the 84Rades data can be derived the theoretical accuracy range of Lat/Lon positioning. For this purpose would be good to describe, how the Lat/Lon data are created from the raw radar data. The Lat/Lon position of the blip is determined by this raw data:
1. Timestamp - accuracy
0.005 seconds
2. Range - accuracy of
0.125 nmi = 0.144 mi = 231 meters
3.
Azimut degree - accuracy 0.001 of one angular degree
(+ 4. Exact Lat/lon position of the radar which is fixed)

So, if we assume an object is flying at maximum speed derived from our analysis - 380 mph then
1.
in - 0.01 seconds can travel 380/(100*3600) miles = 0.001055 miles = 1.697 meter;
2. the range accuracy is 2*0.144 miles = 462 meters (the actual position of the flying object can fluctuate in a circle of the diameter of 462 meters); and
3. maximum range (at the blip No. 1) from the PLA radar was exactly 30 nmi = 34.523 miles = 55559.382912 meters - so the 0.002 angular degree circle section of the circle with the radius of
55559.382912 meters = 0.002*((2π*55559.382912)/360) = 1.939 meter.

We see, that the cummulative accuracy of the radar depends almost solely on the Range data which are the least exact. The cummulative error range for positioning measurement of our "AA77" object is then ~
233 meters = 0.144 miles = 0.125 nmi, which is the same value noted in the 84Rades cover letter (at the page 2, paragraph 2) and it is a distance of aprox. the "radius" of the circle the Pentagon is circumsribed around.

If we then assume a chaotic Gaussian normal distribution scatter of the error, then for the whole 31.09 miles path (with computed groundspeed of 348.18 mph it signifies just:
100*(1/(31.09/0.144)) = 0.0046317*100 = 0.46%! (possible error of computed speed)

for 7.7 miles path (distance measured from blip 8. to 1. (with the computed ground speed of 333.9 mph) then:
100*(1/(7.7/0.144)) 0,0187013 = 1.87%

Cummulatively for the comparison - of the whole 330 path with the last straight segment from blip 8. to 1. - is then the maximum error 0.46%+1.87% = 2,33%. But the difference between 348 mph and 333.9 mph is (348.18-333.9)/(348.18/100) = 4.1% - and I therefore can assume (if I don't make a mistake somewhere) that the conclusions C. and D. are statistically proven - because even if the error of the radar measurement would be at its maximum value, the average groundspeed of the whole 330 maneuvre is here proven to be anyway slightly higher then the average groundspeed in the last straigt section between the blips 8. to 1. - And this, if I understand it well, directly contradicts the FDR data where (in contradiction to this derivations from 84Rades) the speed of the last period is really significantly higher (due to increasing of the throttle to max and a straight flight) then the speed during the previous maneuvre.

But it would be good to corroborate it again for example with the estimated speed between the blips 4. and 1. :
The estimated speed there is 317.4 mph, the distance travelled is 3.14 miles - so the statistically estimated error of position measurement would be: 100*(1/(3.14/0.144))+0.46=0.04586 = 5%. But the difference between 348.18 mph and 317.4 mph = (348.18-317.4)/(348.18/100) = 8.84%

And for the blips 3. to 1. the estimated speed there is 288.6 mph, distance travelled 1.9 miles - so the statistically estimated error of position measurement would be 100*(1/(1.9/0.144))+0.46 = 8% - but the difference between 348.18 mph and 288.6 mph is (348.18-288.6)/(348.18/100) = 17.1% - even more convincing...


Well, I'm not completely convinced about the 84Rades data credibility - in fact I don't the data are untampered - there are too much of it at least "missing" (as the missing period of "UA175" after the start on NOR radar, as the hardly explainable extensive "invisibility" periods on the "AA77" and "UA93"journeys...) but if I would really need (which I don't) to decide which of the two data sets (FDR or 84Rades) are more credible, I would chose the 84Rades - because the data set was supposedly created almost immediately after the 9/11 events and immediately handed to the law enforcement - so the space of time for altering, manipulating or forging of the data would be much shorter (~2. days) and would probably consist just of a "deleting" - than in the case of the FDR data - which rested unpublished nor handed to law enforcement literally for years, so there was plenty of time to forge it. Circumstational reasons for this choice would be also the fact, that the FDR was recovered on unknown location and therefore there is not an undisturbed chain of custody - which would in case of the legal proceedings mean, that such a proof would be inadmissible. But the sole fact, that the NTSB would provide the public with the two mutually contradictory datasets about the most heinous terrorist crimes in the american history would be in my opinion surely a cogent argument for beginning an investigation of NTSB. ...Yes, sure, who would do anything like that...

Positioning source:
The public Google maps source for my private Google Earth research and subsequent drawings

--------------------------------------------------------------------------------------------------------------------------------

NOTE: The 84Rades data originated from NORAD-USAF radars (and subsequently were reported about by 84 Radar Evaluation Squadron at Hill AFB in Utah - to FBI - already on September 13 2001 - and then to NTSB) were obtained on FOIA request last year by John Farmer aka BCR-bluecollarrepublican, former 911Ommissioner, and subsequently published. John Farmer then shortly after even established the public forum dedicated solely to the research of the 84Rades data.
But when I posted there in autumn 2007 serious posts discussing the numerous anomalies in the data (and also my doubts about the overall credibility of the data - arising from the numerous weird issues in it) the forum was almost immediately after my postings deleted as a whole - not just my posts, or the respective thread - the whole forum was wiped from the internet. To this day I don't know if it was deleted by the John Farmer himself or by somebody else. - When I'd then asked John Farmer via email what happened with the 84Rades forum (I naively thought it was moved somewhere) I've received just a laconical reply: "It no longer exists...." After it I didn't get any further communication from him anymore, even I've tryied it several times.
Fortunately I still keep the backup copies of the whole data sets - in the state as it was first published - in the original ISO rar archives - which can prove or disproove the authenticity with numerous timestamps in the many files contained.

I please the reader in advance to forgive me the possible mistakes in the grammar - as I'm not an English native. This analysis is just an inspiration for the deeper study of the 84Rades vs. FDR.
This is not a "finished scientific paper", it's just a suggestion where the search for convincing proofs might be directed. - I think, that if there would be above any doubt proven, that the FDR and 84Rades data significantly differ in the inner consistency of the data patterns - then it would be in my opinion another quite strong proof of the coverup. And who makes a coverup is usually the perpetrator or its accomplice.



This post has been edited by tumetuestumefaisdubien: Apr 4 2015, 09:21 AM
Go to the top of the page
 
+Quote Post
JFK
post Jan 2 2009, 10:48 PM
Post #2





Group: Guest
Posts: 564
Joined: 2-June 08
Member No.: 3,485



You may be interested in comparing your results with mine here - http://s1.zetaboards.com/LooseChangeForums...c/322707/1/#new wink.gif

My plot - http://img501.imageshack.us/img501/6772/masterplot2jn9.png

This post has been edited by JFK: Jan 2 2009, 10:51 PM
Go to the top of the page
 
+Quote Post
tumetuestumefais...
post Jan 3 2009, 04:40 AM
Post #3





Group: Respected Member
Posts: 1,111
Joined: 7-November 07
From: Prague or France
Member No.: 2,452



QUOTE (JFK @ Jan 1 2009, 01:48 AM) *


Very good JFK thumbsup.gif I must look into it after I have a bit of a sleep.
Nice that I'm not alone with the ideas to plot the 84Rades and try to come to some conclusions.
At the first glance I see you come to higher speeds than me. I must emphasize I haven't use the RS3 calculator for purpose - because I'd liked to use something "independent" - so I used the Google Earth measurement tool. But I'm planning to make another version of calculations using the RS3 calculator for comparison.
But the accuracy of distance measurements - in my opinion - doesn't affect the most important conclusions C. and D. - because they're based solely on the relative speeds - on the proportion between the whole maneuvre average groundspeed and the groundspeed of the last straight period to the Pentagon where according to FDR somebody put the throttle on max and the plane according to the both FDR data and derived animations accelerated considerably. (The airspeed in the period of that huge 330 turn would be surely even higher than the groundspeed - because 1. a plane doesn't turn "in angles" and 2. is there quite steeply descending so the proportion betveen the 330 turn real speed and the straight period real speed would be probably even considerably more than the computed +4.1%)
I would like to ask you if you can review it if I don't make a mistake somewhere - because the conclusions are really quite heavyweighted...

just a citation> Reheat Jun 4 2008, 08:19 PM> "The RADES data at altitudes above 2-3000' depending on the range from the antenna should be good, but at low altitude it is virtually worthless."
This is a typical demagogical nonsense of somebody who has no clue about a technology (here the radar) and possibly wants to compromise its credibility. Radars "see" well at low altitudes (especially when 30nmi from target, on the hill approx. 400 feet above a virtually flat table land around 5gon rolleyes.gif - it means in this case direct clear sure visibility of everything which is more than like >100 feet above its roof...and the NEADS PLA radar is there most probably built just for good "seeing the 5gon" and Washington D.C. anyway salute.gif ) - if there are no obstacles in the way - as mountains and like - if not, the radar has exactly the same accuracy as if the object is in the higher altitude - just the obstacles can make some blips unreadable. But if they ARE readable, they're virtually simmilarily accurate even if the object is flying very low - it can even "see" behind the obstacles - if they're smaller than the used RF wavelength...

This post has been edited by tumetuestumefaisdubien: Jan 3 2009, 09:37 AM
Go to the top of the page
 
+Quote Post
dMz
post Jan 16 2009, 10:40 PM
Post #4



Group Icon

Group: Global Mod
Posts: 5,019
Joined: 2-October 07
From: USA, a Federal corporation
Member No.: 2,294



Thanks tume.

JFK and I had a couple of related threads at:

Aa77 Rades - A Different Approach.
http://pilotsfor911truth.org/forum//index....showtopic=13231

Aa77 Rades Data Analysis, Questions and Findings from RADES data
http://pilotsfor911truth.org/forum//index....showtopic=13211

Some information on the various John Farmer(s) can be found at post #19 here:

Questions From Reopen911 France On Dca-tyson Video
http://pilotsfor911truth.org/forum//index....&p=10762234
Reason for edit: Added Farmer clarification link
Go to the top of the page
 
+Quote Post
tnemelckram
post Jan 17 2009, 09:42 PM
Post #5





Group: Contributor
Posts: 767
Joined: 30-January 08
Member No.: 2,690



I thought I'd add to Tumes' RADES Radar Plot a Google Earth plot of the last several IAD, BWI, DCA and ADW returns from the .csv files on the AA77 Website. I hope the online file storage link works because this is the first time I've tried to do that. I tested it in the Preview and it did work.

I'm not sure of the technical merit of my method but only intend this to be a rough approximation and think it is sufficient to suggest the points that follow. What I did was create placemarks by typing the time and decimal latitude and longitude into the placemark window. That only permitted me four decimal places.

This shows a couple of things other than the four airport radar returns that I don't intend to be the focus of this post.

(1) The flight of the Muscle 6 helicopter in green helicopter Icons.

(2) A hypothetical "fly over" path in yellow. It is drawn from DCA returns
without a Beacon Code shortly after 13:37:45. I still think is still a little
half-baked because the yellow path is a lot faster than the IAD, BWI, DCA
and ADW approach path. But then again we don't know what type of aircraft.

So please enlarge the area to the west of the Pentagon. The airplane Icon Tags say the airport code and time.
BWI is in orange.
ADW is in red.
DCA is in dark blue.
IAD is in light blue.

Five things were striking to me.

(1) The last DCA return at 13:37:46 is still a second or two away and thus 2-3
seconds after the alleged impact time.

(2) Whether or not you throw out the last and too late DCA return, the rough visual
average path of all the returns points over the Annex and North of Citgo.

(3) The IAD and BWI returns are several seconds behind DCA and ADW returns at
approximately the same place.

(4) After the IAD and BWI returns end, there are 3-4 more returns from ADW and DCA.

(5) The path of the three BWI returns points significantly more to the northeast than
the other three more easterly return paths. If the BWI path continued that would
be my idea of North of Citgo!

http://www.orbitfiles.com/download/id3810762883.html






This post has been edited by tnemelckram: Jan 18 2009, 01:18 AM
Go to the top of the page
 
+Quote Post
dMz
post Jan 20 2009, 03:02 PM
Post #6



Group Icon

Group: Global Mod
Posts: 5,019
Joined: 2-October 07
From: USA, a Federal corporation
Member No.: 2,294



I split the Google Earth-specific posts to Research:

Google Earth Howto, and a Bit on File Servers- split
http://pilotsfor911truth.org/forum//index....showtopic=16090
Go to the top of the page
 
+Quote Post
tnemelckram
post Jan 20 2009, 10:09 PM
Post #7





Group: Contributor
Posts: 767
Joined: 30-January 08
Member No.: 2,690



Hi dMole!

Whew! Glad that's behind us.

QUOTE
I notice that the BWI track seems to be doing something different from the others-hmmm....


Yep I noticed that too. BWI and IAD cut off early too. See my original post 5 on the other thread for my other comments.

QUOTE
Also, the trends appear to put all these North of the Columbia Pike and headed for that Navy Annex-ish. I think I'll pull the above image into my CAD program and take a few measurements (and "shoot" a few extension lines)- could be interesting.


I did some extensions using GE to plot extensions of what I eyeballed as the average course and time for each airport separately. I eliminated them from what you have because I just wanted to send along facts instead of my projections someone who knows more about the subject to use as a base for projections or whatever else might be helpful and meaningful. My extensions all ended up over Annexxy and on NOC'ey.

Is there some mathematical way to arrive at a line or course that is an average derived from all of the return points shown? I had all of the advanced maths and sciences in HS (through Calc. and Physics) but was an indifferent scholar The main thing I learned as this is what enginers and other subcontractors and specialists were for. I have the sense that such an average is possible. Or even better,the average course of these four, the RADES points, and the DFR points so the OCT'ers can't accuse us of being selective..
(Throwing out the "Chantilly Course" because it's silly).

QUOTE
So what was the data source on these lat/lon coordinates again?


The .csv files for radar data for each airport on the "AA77" website that were released by the FAA in I think August or September 2008. The same ones Craig and I debated.

This post has been edited by tnemelckram: Jan 20 2009, 10:17 PM
Go to the top of the page
 
+Quote Post
tnemelckram
post Jan 20 2009, 10:12 PM
Post #8





Group: Contributor
Posts: 767
Joined: 30-January 08
Member No.: 2,690



Corrected link for Post 5 above, also not showing the yellow and green data to preserve my intended focus:

http://www.orbitfiles.com/download/id3821224891.html
Go to the top of the page
 
+Quote Post
dMz
post Jan 20 2009, 11:56 PM
Post #9



Group Icon

Group: Global Mod
Posts: 5,019
Joined: 2-October 07
From: USA, a Federal corporation
Member No.: 2,294



QUOTE (tnemelckram @ Jan 20 2009, 07:09 PM) *
Is there some mathematical way to arrive at a line or course that is an average derived from all of the return points shown? I had all of the advanced maths and sciences in HS (through Calc. and Physics) but was an indifferent scholar The main thing I learned as this is what enginers and other subcontractors and specialists were for. I have the sense that such an average is possible. Or even better,the average course of these four, the RADES points, and the DFR points so the OCT'ers can't accuse us of being selective..
(Throwing out the "Chantilly Course" because it's silly).

[data source:]
The .csv files for radar data for each airport on the "AA77" website that were released by the FAA in I think August or September 2008. The same ones Craig and I debated.

As position and velocity are both vector quantitities TN, why yes there is: vector analysis, and the resultant of a sum of vectors.

http://www.mathwarehouse.com/vectors/resultant-vector.php

I could take a "half angle" estimate measured between 2 or 3 of these 4 flightpath trends, but the "small angle" approximation appears to have 3 of the 4 (excepting BWI) converging to a common path over the Navy Annex/Federal Office Building (FOB). That is of course if the data "source" can be trusted.

Here are some preliminary screencaps from my CAD work as a "teaser:"

CAD vectors overlaid on original Google Earth image
http://flickcabin.com/public/view/19140

4 aircraft tracks and "resultants" without image, and color coded for TN's plane icon colors
http://flickcabin.com/public/view/19141

Resultant "average displacement" dimensions (in US feet, scaled approximately to the 921 foot Pentagon wall)
http://flickcabin.com/public/view/19142

4 aircraft tracks, with no resultants or dimensions
http://flickcabin.com/public/view/19143

4 aircraft tracks, with a 3-point circle superimposed, with references drawn from the "impact" location, the NW Citgo canopy corner, and the "middle" of the red ADW track. The radius is dimensioned in US feet at the image import scale (921 foot wall). This is a little "wide" for the last P4T presentation (Xmas release), but is entirely possible.
http://flickcabin.com/public/view/19144

More to follow on these 4 tracks.

P.S. TN- I moved your reply over here, as I thought it was more topical here (less GE-specific).

EDIT: You could of course subtract the timestamps and use the dimensions from that 3rd image to find a "net" speed across that section (which should be slower than the "instantaneous" velocity for anything other than a straight line (which the orange BWI track was very close to).

Also, it is very easy for me to calculate "resultant" of the 4 different resultant vectors in CAD, but the "missing" data and different timeframes make this inadvisable IMHO.
Go to the top of the page
 
+Quote Post
dMz
post Jan 21 2009, 12:19 AM
Post #10



Group Icon

Group: Global Mod
Posts: 5,019
Joined: 2-October 07
From: USA, a Federal corporation
Member No.: 2,294



For the 4 resultant vectors:

red ADW= brg 61 degrees (clockwise from GeoN for all angles), length = 14527.5163 [US feet, image scaled to Pentagon wall at 921 units for all linear dimensions]

orange BWI = brg 45 degrees, len = 5275.4682 [ft]

dark blue DCA = brg 61 degrees, len = 18438.0051 [ft]

light blue IAD = brg 56 deg, len = 9197.961 [ft]

NOTE: red ADW and dk blue DCA are parallel courses at brg "061"
--------------------------
EDIT: Here are where the extension lines for these 4 color-coded resultants would fall.

With image underlay
http://flickcabin.com/public/view/19146

With no image underlay
http://flickcabin.com/public/view/19147

With image underlay, zoomed out to see aircraft return locations
http://flickcabin.com/public/view/19148

FWIW, I drew the purple line at the "official" NTSB "AA77_tabular.csv" final true heading of 59.8 deg and a length of 6076.1155 feet (1 US nautical mile) for reference. The orange BWI resultant displacement is extremely close to 1 US statute mile for reference. Strangely, the green line from "impact" to the NW Citgo canopy corner line is also nearly in line with the NW corner of the Navy Annex.
--------------------------
EDIT2: Here are my new images with the last few USAF 84 RADES data points also plotted, but there was a missed return between 09:37:12 and 09:37:36 (should be ~12 second radar "sweep."). I haven't done any CAD work on any of these images yet.

Here is the collection (with various lat/lon and distances marked for the NTSB, FAA, and USAF 84 RADES data sets for "AA77"):
http://flickcabin.com/public/viewset/6228
Go to the top of the page
 
+Quote Post
tnemelckram
post Jan 21 2009, 05:46 PM
Post #11





Group: Contributor
Posts: 767
Joined: 30-January 08
Member No.: 2,690



Hi dMole:

Thank you for your quick work. You did what I hoped you'd do with the plots of the radar data from the four airports.

I'm going to read the math link to the sum of vectors math site and maybe learn or re-learn something. Thus enlightened, I'll look at it again.

I hope that this can be turned in to more support for NOC. The best way might be to do this with the four airports, the FDR positions, and the RADES positions, That way nobody can accuse us of being selective. The point being that all of this was available to the 911 Commission and if it is to be believed, then it trends to NOC.

Of course, the question of whether the "yellow line" data in my "learning curve" .kmz is valid evidence of something flying away post flyover might merit consideration next. I'm still mulling it over.

Thanks again.
Go to the top of the page
 
+Quote Post
rob balsamo
post Jan 22 2009, 09:35 AM
Post #12



Group Icon

Group: Admin
Posts: 9,745
Joined: 13-August 06
Member No.: 1



Interesting work guys. It appears Radar plots from the various facilities are more consistent with the NoC approach.

It wouldnt suprise me if Farmer totally screwed the pooch on his radar analysis. Wouldnt be the first time, and im sure not the last.
Go to the top of the page
 
+Quote Post
dMz
post Jan 23 2009, 02:39 AM
Post #13



Group Icon

Group: Global Mod
Posts: 5,019
Joined: 2-October 07
From: USA, a Federal corporation
Member No.: 2,294



OK, here is the CAD/graphical method I used to "average" 4 of the 5 flight path trends (I threw out the orange BWI resultant trend, as it only had 3 data points, and it appeared to be headed much more "north" than the other 4 and would miss both the Citgo and Pentagon entirely. A logical conclusion could be that the orange BWI track is another aircraft entirely (as "AA77" is alleged to have cut its transponder and disappeared off radar).

First, I picked what appeared to be the "convergence" area of the 4 extension lines (RADES, ADW, DCA, and IAD). I then extended perpendicular lines (of the B757-200 124.833 foot [38.05 m] wingspan length) from the "widest" of whichever were the outermost of 4 resultant lines.

http://flickcabin.com/public/view/19352

http://flickcabin.com/public/view/19353

This gave me a (yellow) rectangle (or perhaps more properly a trapezoid/quadrilateral, but that doesn't really matter for our purposes here). Here is this yellow "convergence rectangle" filled with a pattern:

http://flickcabin.com/public/view/19356

http://flickcabin.com/public/view/19357

You will note that I have used the CAD program to draw a yellow extension line from the precise midpoint of each "end" of the rectangle. (CAD makes this very easy to do). Strangely, this yellow "graphical average" resultant is very close to the purple RADES resultant (except I had also thrown out the final RADES data point previously, as it made a strange "90 degree, southeasterly" bend away from the obvious trend of the other RADES data points used). Just look for the purple line- it is the logical RADES trend.

Here is the overall picture, with all 5 extension lines shown:

http://flickcabin.com/public/view/full/19354

http://flickcabin.com/public/view/19355

The same thing, but zoomed in a little closer to the Annex and Pentagon area:

http://flickcabin.com/public/view/19358

http://flickcabin.com/public/view/full/19359

You will note in that last photo immediately above that the little white line segments "straddling" the NTSB-"official" true heading of 59.8 degrees (dark greenish gray line heading for Pentagon "impact") are scaled to the 124 foot 10 inch B757-200 wingspan. That greenish gray line is still 1 nautical mile long or 6076.1155 feet, it's just a different color in this set of drawings.

Conclusion: My "graphical average," the RADES, and 3/4 of the FAA radar trends all show heading North of the Citgo canopy's NW corner. The blue FAA DCA radar trend goes over the Citgo canopy and is inconsistent with the Ed Paik description and drawing.

Next, I will take a more quantitative vector average, but you can all at least review the CAD drawings now.
-------------------------------
For Tume: here is my newer "AA77" NTSB and RADES Groundspeed (plus NTSB TAS and CAS) chart for the last minutes of the flight (from post #6 on a nearby thread):

http://pilotsfor911truth.org/forum//index....&p=10763025

http://flickcabin.com/public/view/full/18928

Also, we can get "average velocities" from these scaled distances between radar/time coordinate returns if you like- let me know which track(s) you are interested in, and I can post the distances. The timestamps will be the most convenient reference method (although TN used Zulu time and I used EDT, partially to distinguish the RADES data points- EDT is 4 hours "later" than Zulu/GMT).

EDIT: Essentially, I used a graphical "draftsman's compass/welder's tape" method to find the middle of the flightpath "cone angles" to skip scaling/trig issues. I merely found the "half angles" and shot that yellow extension line from those.
Go to the top of the page
 
+Quote Post
dMz
post Jan 23 2009, 03:25 AM
Post #14



Group Icon

Group: Global Mod
Posts: 5,019
Joined: 2-October 07
From: USA, a Federal corporation
Member No.: 2,294



I forgot to post the bearings of those resultant heading lines (to two decimal places, in degrees this time).

red ADW = 60.92 degrees
orange BWI = 45.31
blue DCA = 60.74 [not exactly parallel to ADW, but damn close]
cyan IAD = 56.17

purple RADES = 57.82

yellow "graphical average" = 58.94 deg. (derived from all 4, except BWI)

Now working on a scaled vector resultant of these headings (and will use trig in a spreadsheet to confirm, for an arbitrary, hypothetical 300 kts velocity)

EDIT: Well the CAD vector drawing agrees with Excel/trigonometry out to 6 decimal places (in degrees). The resultant of the above vectors ADW + DCA + IAD + RADES is at bearing 58.912598 degrees clockwise from GeoN. The resultant is (4 * the "unit" length u of either distance or velocity for the 4 identical vector lengths) * a "straightness factor" of (1 - 0.000612376 ) for these 4 particular bearings.

In other words, there is a 0.0612 percent "crookedness" loss of whatever unit vector length u one chooses. If these were 4 parallel vectors (at same bearing angle), then the resultant would just be 4 * the unit vector length u, with 0.000% "crookedness" loss.

If I take the arithmetic mean of these 4 bearing angles, I get 58.91250 deg with S.D. = 2.315504 deg. This is 0.0465126 percent different from the CAD and trig methods.

EDIT: The difference ratio between the graphical method and trig was 5.0925140044e-4, or .051 percent.
Go to the top of the page
 
+Quote Post
tnemelckram
post Jan 24 2009, 05:31 PM
Post #15





Group: Contributor
Posts: 767
Joined: 30-January 08
Member No.: 2,690



Hi dMole!

Thanks for quickly putting the RADES, BWI, DCA, ADW and IAD radar return data into form and order from which conclusions can be drawn.

Here's the links to the four airport .csv data files. Anybody can check the plotting of the points by doing a "find" search for "AAL77" because somebody helpfully labeled them to distinguish them from the mass of other data. Remember that Google Earth only allowed me to input four decimal places. Reducing the eight or nine decimal places in the .csv files to four and accurately typing them is an opportunity for error on my part.

DCA http://www.aal77.com/faa/faadata/dcatracon.csv

ADW http://www.aal77.com/faa/faadata/adwtracon.csv

IAD http://www.aal77.com/faa/faadata/iadtracon.csv

BWI http://www.aal77.com/faa/faadata/bwitracon.csv

It's certainly reasonable to omit the three BWI points (which would push the average path even further NOC) for the facially erroneous last RADES point (which would pull that path to the south).

We are trading three points for one. I know that the issue is not the number of points on each side of the argument but rather the soundness of each particular point. WIth that in mind, the last DCA 13:37:46 return (which also pulls south) looks like it is still a second or two away from the Pentagon and thus 2-3 seconds too late for the official 13:37:45 impact time. Maybe tossing it too can be justified. As a sweeetener, we can add that we are not even going to bother factoring in the "Chantilly Course" shown by yet another data set, which would pull the whole mess westward and support a plane striking something like an auto dealership in Arlington.

Interestingly, furthering the line of the three tosed-out BWI returns offers rich targeting opportunities such as the National Cathedral, Watergate, the State Department, the White House, and the Hays- Adams District. With it, everybody can have their own little private alternative 9-11.

I wonder how a formal conclusion should be stated. Rob seems to like them conservative and measured. His above statement is a good starter:

QUOTE
It appears Radar plots from the various facilities are more consistent with the NoC approach.


How about:
QUOTE
The RADES, BWI, DCA, ADW and IAD radar return data was produced by electromechanical devices. Such data is objective and a generally accepted means for determining actual and projected courses in forensic aviation investigations. In fact, it the primary means. Such data must be examined and averaged by generally accepted methods in that field; the result must be the cornerstone of any determination of a flight path. The 911 Commission failed to do so. When these generally accepted means and methods are used:

(1) the returns position some kind of plane on a flight path heading North of Citgo;

(2) the projected course would continue North of Citgo;

(3) the flight path found by the 911 Commission is excluded; and

(4) the testimony of various witnesses to the North of Citgo flight path is cooborated.


This statement might require an expert or authority in forensic aircraft investigations to confirm that the "means" and "method" are "generally accepted in the field" and "primary", and then were used properly.


This post has been edited by tnemelckram: Jan 24 2009, 06:57 PM
Go to the top of the page
 
+Quote Post
dMz
post Jan 24 2009, 06:39 PM
Post #16



Group Icon

Group: Global Mod
Posts: 5,019
Joined: 2-October 07
From: USA, a Federal corporation
Member No.: 2,294



TN, I get 6 decimal places in this version of Google Earth: 4.3.7284.3916 (beta). (I can also see that you only input 4 decimals of lat/lon for those points). Perhaps I'll whittle down an "AA77 FAA" spreadsheet from those 4 and plot with more decimal places to take some measurements, if I get time.

Of course I also get very consistent GE crashes when I try to add a "path" or use a few other functions too...
Go to the top of the page
 
+Quote Post
tnemelckram
post Jan 24 2009, 06:52 PM
Post #17





Group: Contributor
Posts: 767
Joined: 30-January 08
Member No.: 2,690



Annanudderting:

It looks like it would take a helluva curve to get the plane off of the path shown by these radar returns to get it on to the straight line of the official flight path.
Go to the top of the page
 
+Quote Post
dMz
post Jan 24 2009, 07:13 PM
Post #18



Group Icon

Group: Global Mod
Posts: 5,019
Joined: 2-October 07
From: USA, a Federal corporation
Member No.: 2,294



QUOTE (tnemelckram @ Jan 24 2009, 03:52 PM) *
It looks like it would take a helluva curve to get the plane off of the path shown by these radar returns to get it on to the straight line of the official flight path.

The "curve" could be modeled as 2 circular arcs of equal radius (and would be hella easy for me to draw and measure in that scaled CAD file now). Which trend do we want to use- the yellow "graphical" one, draw the new "trig heading", or ??

I would need to have 2 arbitrary "start points" for the "curve" (on whichever trend line we choose) and to draw a new "trig heading" though. Does anyone have preference(s) there?

http://flickcabin.com/public/view/full/19354

http://flickcabin.com/public/view/full/19358

EDIT: To reiterate, the "trig" version, at 58.912598 deg, was ~=0.03 degrees counterclockwise (or "north") of my yellow "graphical average of 4" heading (at 58.94 deg)- not exactly a huge difference (~1.8 arcminutes).

EDIT2: For the benefit of the "instantaneous velocity" duh-bunkerz:

Last known radar headings for final "segments":

ADW 63.08 deg, len = 3412.2264 US feet
BWI 45.65 deg, 2918.2404 ft.
DCA 64.03 deg, 3805.414 ft.
IAD 52.92 deg, 6853.4868 ft.

It doesn't have RADES, but the "final segments" should be quite similar to this earlier image:

http://flickcabin.com/public/view/full/19143

*** RADES *** 148.39 deg *** 1381.6507 feet


That last purple one looks awfully fishy though... (plus the speed would be ~68 knots whistle.gif ).
Go to the top of the page
 
+Quote Post
tnemelckram
post Jan 24 2009, 11:03 PM
Post #19





Group: Contributor
Posts: 767
Joined: 30-January 08
Member No.: 2,690



Hi dMole!

QUOTE
TN, I get 6 decimal places in this version of Google Earth: 4.3.7284.3916 (beta). (I can also see that you only input 4 decimals of lat/lon for those points).


I was concerned about the margin of error at the fourth decimal place and tried to figure out what it was. So starting with my existing plot of the last DCA 13:37:48 return, I decided to find out the maximum distance change from adding or subtracting one from the number in the fourth decimal place of latitude and longitude. The rolling latitude and longitude numbers for the cursor placement at the bottom of my screen actually go to six decimal places. I zoomed to the minimum eye altitude that Google Earth would allow, which was 159 feet.

Then I added three additional placemarks to my original DCA 13:37:48 placemark. Here are the labels and respective six decimal latitude and longitude for all four of those placemarks in the attached .kmz file:

Existing "DCA 13:37:48" = Lat. 38.865300 Long. 78.072190

New "4th Decimal Exact" = Lat. 38.865300 Long. 78.072200

New "4th Decimal -1" = Lat. 38.865200 Long. 78.072100

New "4th Decimal+1" = Lat. 38.865200 Long. 78.072100


Then, with the loving care of a nun disciplining an unruly student, I took the Ruler to em'. The maximum distances are as follows:

DCA 13:37:48 to 4th Decimal -1 = 44.39 feet

4th Decimal Exact to 4th Decimal -1 =46.27 feet

DCA 13:37:48 to 4th Decimal +1 = 47.75 feet

4th Decimal Exact to 4th Decimal +1 =45.81 feet

DCA 13:37:48
to 4th Decimal Exact = 2.88 feet

I think you'll have to really zoom in on the location of DCA 13:37:48 to get a good look. The points start to separate at about 3000 feet eye level. The measuring is really good below 200 feet. Link:


[older link- left for reference purposes]
http://www.orbitfiles.com/download/id3837657542.html

[updated link]
http://www.orbitfiles.com/download/id3839734731.html

QUOTE
Perhaps I'll whittle down an "AA77 FAA" spreadsheet from those 4 and plot with more decimal places to take some measurements, if I get time.


If you think a 47 foot Margin Of Error does not affect the analysis in a meaningful way and my method of determining it is sound, then maybe we can leave this work to some GL schmuck.


QUOTE
Of course I also get very consistent GE crashes when I try to add a "path" or use a few other functions too...


Get the new, improved Google Earth! Now in six (count em', six) decimal places! Step right up, folks!

This post has been edited by dMole: Jan 26 2009, 01:18 AM
Go to the top of the page
 
+Quote Post
dMz
post Jan 25 2009, 12:54 AM
Post #20



Group Icon

Group: Global Mod
Posts: 5,019
Joined: 2-October 07
From: USA, a Federal corporation
Member No.: 2,294



QUOTE (tnemelckram @ Jan 24 2009, 08:03 PM) *
Then I added three additional placemarks to my original DCA 13:37:48 placemark. Here are the labels and respective six decimal latitude and longitude for all four of those placemarks in the attached .kmz file:

Existing "DCA 13:37:48" = Lat. 38.865300 Long. 78.072190

New "4th Decimal Exact" = Lat. 38.865300 Long. 78.072200

New "4th Decimal -1" = Lat. 38.865200 Long. 78.072100

New "4th Decimal+1" = Lat. 38.865200 Long. 78.072100


Then, with the loving care of a nun disciplining an unruly student, I took the Ruler to em'. The maximum distances are as follows:

DCA 13:37:48 to 4th Decimal -1 = 44.39 feet

4th Decimal Exact to 4th Decimal -1 =46.27 feet

DCA 13:37:48 to 4th Decimal +1 = 47.75 feet

4th Decimal Exact to 4th Decimal +1 =45.81 feet

DCA 13:37:48
to 4th Decimal Exact = 2.88 feet

I think you'll have to really zoom in on the location of DCA 13:37:48 to get a good look. The points start to separate at about 3000 feet eye level. The measuring is really good below 200 feet. Link:

http://www.orbitfiles.com/download/id3837657542.html

Hi TN,

I got an empty folder with no data points for that GE file. Also, that orbitfiles server does not like spaces in the filename, so I had to rename "4th" to "4th.KMZ" to get it to open.

I'll use the following calculator and your last DCA "exact" data point:

Lat. 38.865300 Long. 78.072190

http://williams.best.vwh.net/gccalc.htm

Latitude will have the greatest effect in the 4th decimal place of lat. degrees (as the latitudes don't "narrow" due to spherical projection). I'll use the WGS84 datum and US feet calculator parameters for all.

N (hold longitude constant at 78.072190 deg W)

Lat+1 = Lat. 38.865400
Lat_0 = Lat. 38.865300
Lat-1 = Lat. 38.865200

Lat+1 > lat_0: 180 degrees, 36.42155940950806 feet

Lat_0 > Lat-1: 180 deg, 36.42155877792267 feet
----------------------
Now looking at longitude (hold lat constant at Lat_0 = Lat. 38.865300 deg N)

Lon+1 = 78.072290
Lon_0 = 78.072190
Lon-1 = 78.072090

Lon+1 > Lon_0: 89.99996862645274 deg, 28.474426344909133 feet

Lon_0 > Lon-1: 89.9999745889962 deg, 23.0642853461734 feet
---------------
Next, I think I'll take the "start" and "2nd from last" RADES points from this analysis (as that is the greatest distance found in the analysis on this thread) and evaluate both with the online calculator and my RADES Spherical Law of Cosines approximation for distance/velocity to estimate the error there. I'll also compare to the scaled CAD drawing distance.

EDIT: In retrospect, mine is a beta version of Google Earth (but it does have a much prettier "high resolution" mode that I don't recall in last year's version).
Go to the top of the page
 
+Quote Post

2 Pages V   1 2 >
Reply to this topicStart new topic
3 User(s) are reading this topic (3 Guests and 0 Anonymous Users)
0 Members:

 




RSS Lo-Fi Version Time is now: 22nd September 2019 - 12:13 AM