Siesmic Evidence Proves Inside Job?

[R.Mackey]

Follow Up on Propagation Speed

As mentioned above, the seismograph presented by LDEO is consistent with an origin time of 8:46:29, as reported in NIST, if the propagation time from impact to the LDEO location takes about sixteen seconds. In this post we look a bit closer at this issue to verify that our hypothesis makes sense.

From the LDEO chart itself, the distance from impact point to detector is equal to 33.79 km (in the margins of the graph, and see below). For a period of sixteen seconds, this implies that the wave propagation speed must be about 2100 meters per second. Is this possible?

Previously, I remarked that the ordinary propagation speed of seismic waves is about 5 km/s, but that it varied. It actually varies quite a bit, partly on the rock composition, but also on the type of seismic wave. See here for a simplified but illuminating treatment.

The most obvious waves are the so-called "P-Waves," short for plane waves. These are essentially compression waves. Think of an arc moving through the Earth, and when this arc passes it compacts the ground briefly, just like the shockwave from a bomb. P-waves are also the fastest waves, but they are rarely the strongest.

Next up are the surface waves, or "S-waves." Think of these like waves on the ocean. As the wave passes, the surface of the Earth moves vertically. If you could see them from a great height, you would see ripples in the surface moving outward from the source. The important point, however, is that just like waves in deep water are slower than the speed of sound in water, S-waves are slower than P-waves. It's a different physical phenomenon.

There are two other types, and the wave we're looking at is one of these types, similar to S-waves but slightly different. What we have here is called a Love wave. A Love wave is like an S-wave except instead of the ground motion being vertical, it is horizontal. The wave is a side-to-side shaking that moves outward from the source. Love waves are slightly slower than S-waves, and usually at least twice as slow as a P-wave.

We know our wave is a Love wave for two reasons. First, the source of the wave is horizontal in nature, as the structure suffers an impact that causes it to sway, giving rise to a horizontal displacement at the surface. Second, the detector used by LDEO is east-west surface motion, and LDEO is almost directly north of the impact site. Therefore, the LDEO is reporting on Love waves.

There are other reasons why it makes sense to detect Love waves. These waves are often the most strongly felt, and the impact we're talking about was rather small by earthquake standards (about M_R 1.0 or less). Also, Love waves suffer the least dispersion over distance, for reasons that are somewhat complicated*. Thus it makes sense that Love waves would be the dominant, if not only, wave detected from the impact.

It is apparently quite ordinary for Love waves to have a propagation speed of about 2 km/s. For this reason, it makes perfect sense for the sixteen second delay in the LDEO graphs to account for the wave's travel time.

I've annotated one of the LDEO traces below. Note that the arrival time at LDEO is comparable to the 9/11 Commission's incorrect impact estimate, but again, this appears to be merely a coincidence. Click on the image to enlarge.



As I stated above, it appears to me that the LDEO and NIST are fully consistent, and now we have further reason to trust LDEO's estimate of origin time. The 9/11 Commission origin time matches neither of these, and as previously noted, the error appears to be a sloppy curve-fit to the last seconds of primary radar data. Correcting for this curve-fit, we believe the radar data also plausibly agrees with the LDEO and NIST origin time estimate.

*: The reason Love waves disperse less than P-waves or S-waves is that they are almost totally confined to the surface. The group velocity of Love waves is strongly dependent on the modulus of elasticity and density of its medium, but also with depth -- thus it is much more dependent than the other types of wave. This dispersion relation forces the waves to travel along the surface, because it means the interface to lower material is poorly conducting and reflects those waves back to the top. This is the same principle that makes jet planes sound louder when there is a temperature inversion, or why submarines can hide from sonar beneath a thermocline, because the changing propagation speed funnels the waves horizontally rather than across the material gradient.

 

[gumboot]

Has anyone discussed the margin of error in radar data? It seems the big issue here is a discrepancy in radar impact time and seismic impact time.

However depending on the radar that this information is coming from, the margin of error can be anything from 4.8 to 12 seconds (this being the update period for the fastest rotating ASR series surveillance radars and slowest rotating ARSR-3 surveillance radars).

Given that Boston ARTCC reported AA11 fell below primary radar coverage I would have to presume that it was an ASR radar that tracked AA11, but that's still a margin of error of nearly 5 second. I'm not sure what the delay is like between reception from the radar to display on the screen either, but it wouldn't surprise me if it was a second or two (the data isn't just sent directly to a scope, there's some complex processing done first).

 

[R.Mackey]

Seeing as how for AA-11, there's no radar return at all after 8:46:10 (excepting the phantom "final" trace 30 seconds later), the uncertainty in the sweep is probably not significant.

Both the sweep time and the altitude errors are conflated, because we're basically using these to estimate the rate of descent, and thus the intercept at the target. The chart in the NTSB Report shows greater proportional uncertainties in the altitude estimates than the sweeps. Nonetheless, as I remark above, the impact time from radar is probably no better than +/- 7 seconds or so even if the radar returns are all marked precisely.

 

[gumboot]

Seeing as how for AA-11, there's no radar return at all after 8:46:10 (excepting the phantom "final" trace 30 seconds later), the uncertainty in the sweep is probably not significant.

Both the sweep time and the altitude errors are conflated, because we're basically using these to estimate the rate of descent, and thus the intercept at the target. The chart in the NTSB Report shows greater proportional uncertainties in the altitude estimates than the sweeps. Nonetheless, as I remark above, the impact time from radar is probably no better than +/- 7 seconds or so even if the radar returns are all marked precisely.

The matter of altitude finding is an interesting one. On 9/11 only the ARSR-4 (the JSS Long Range Radar used by NORAD and the FAA) had altitude finding capability with an accuracy of +/- 1500ft @ 175nmi (vertical resolution of 0.17 degrees). The ARSR-4 has an update time of 12 seconds.

The ASR-9 conceptually had altitude finding capability with an accuracy of +/-2000ft @ 40nmi (vertical resolution of 1 degree) but this was a conceptual demonstration and not used operationally. The ASR-9 has an update time of 4.8 seconds.

Now the two nearest JSS ARSR-4 sites to the WTC are Riverhead, NY (72mi) and Gibbsborro, NJ (80mi). By my calculations the Riverhead site would have an altitude finding margin of error of 550ft and the Gibbsborro site would have an altitude finding margin of error of about 650ft (based on 1.7 degree resolution producing a resolution of 1100 and 1300ft respectively).

So in short, you're talking significant margins of error for radar data both as far as delays in actual data reporting and then inaccuracies in altitude.

 

[quicknthedead]

Has anyone discussed the margin of error in radar data? It seems the big issue here is a discrepancy in radar impact time and seismic impact time.

However depending on the radar that this information is coming from, the margin of error can be anything from 4.8 to 12 seconds (this being the update period for the fastest rotating ASR series surveillance radars and slowest rotating ARSR-3 surveillance radars).

Given that Boston ARTCC reported AA11 fell below primary radar coverage I would have to presume that it was an ASR radar that tracked AA11, but that's still a margin of error of nearly 5 second. I'm not sure what the delay is like between reception from the radar to display on the screen either, but it wouldn't surprise me if it was a second or two (the data isn't just sent directly to a scope, there's some complex processing done first).

I have looked at the radar question quite a bit through my sojourns on this.

The 84th Rades had AA Flt 11 tracked for its entire flight and also brought into their study all radar returns from all sources for the flight, and what is telling about the final 8:46:40 crash time cited in the Commission Report that Ryan Mackey mentions above (post #460) when he observes correctly it was based upon a radar return at zero altitude on the graph, are the words found in the same report on page 2, 2nd to last paragraph. They shed light on this "zero altitude" issue:
"However, long range radar obtained from the USAF includes (less precise) altitude information based on primary radar returns..."

One aspect of basic Primary Radar is that it does not provide altitude information, but enhanced military versions of Primary Radar available to the USAF are capable of providing altitude, albeit sometimes resulting in "less precise" data. There are several zero altitudes returns seen on the graph provided, and are extreme examples of this lack of precision in altitude. But primary radar returns are viable, even without an altitude component. They are "returns".

Since the USAF did not reject this final return at 8:46:40 (at zero altitude), their acceptance of it demonstrably indicates it was considered viable, a good return (even without an altitude component). I am not a radar expert, but I believe this is what experts on the subject would basically state.

What Ryan writes in his post #459 sums it up best:

For me, the ultimate proof that there was no paired explosive event was the fact that, on the seismograph, there's only one event. Had there been explosives and a plane impact, there would have been two. This observation is totally independent of any timing mismatch.

We may never know the truth on this discrepancy of times (seismic vs radar), but as Ryan points out, there was only one seismic event. Finis.

Ergo, there has to be error in the data, either in radar or seismic or both (this is possible too).

What do we know for sure?

Somebody's made an error.

But who made it, is this important?

(For the advancement of science, yes, but) In the final analysis, probably not.

Compared to the trauma of 9/11 that everyone has gone through one way or another, this difference in seconds doesn't mean much (at least to me it doesn't anymore).

But hopefully we keep learning. I think I now understand better the seismic aspects of the strikes on WTC1 and WTC2.

Although UA Flt 175 struck about the 81st floor at 540 mph--but was not a dead-center hit, while AA Flt 11 did hit dead-center but at a lesser speed of 440 mph and hit higher up {about the 93rd floor}, the WTC1 seismic event was longer in duration and stronger in magnitude than WTC2's seismic signature.

Those planes hitting those cores was comparable to "ringing a bell", and the "ring" in WTC1 was experienced by many in the basement levels as an enormous, destructive energy transmitted down through the structure's steel core that literally "bounced" off the bottom fixed to bedrock.

Then again I could be wrong. What do I know?
But was 9/11 an inside job? I don't think.

 

[gumboot]

I have looked at the radar question quite a bit through my sojourns on this.

The 84th Rades had AA Flt 11 tracked for its entire flight and also brought into their study all radar returns from all sources for the flight

The 84th RADES is a Radar Evaluation Squadron (that's what RADES stands for) and they're responsibly particularly for taking care of USAF radars and they often assist in investigations. They didn't track AA11 on radar themselves, as a unit (although 84th RADES personnel are located at all NORAD centers) The study for 9/11 was conducted by the National Transportation Safety Board in support of the FBI. The NTSB utilised both military and civilian radar to plot its track, and the military radar used, as I've discussed, is the ARSR-4 which comprises the Joint Surveillance System (JSS) utilised by both DOD and the FAA.

The ARSR-4 is was the only altitude-finding radar in operation at the time of the attacks, so we can say with a reasonable degree of certainty that the radar data presented in the NTSB report came from JSS sites. Further, because we can find out the location of radar sites in the USA via the FAA's National Airspace System Architecture site, we can determine which radar sites would have been used, and how accurate they would have been.

We know from specs that the ARSR-4 has an update rate of 12 seconds - that gives us a sweep margin of error. We know it has a vertical resolution of 0.17 degrees. Combine that with the range from the site to the WTC and we can get the altitude finding margin of error.

It's worth noting that the NTSB report cites the impact time of AA11 as "approximately" 08:46:40EDT. This is more than likely because the NTSB recognises that the radar data has a margin of error as I've discussed above.

Since the USAF did not reject this final return at 8:46:40 (at zero altitude), their acceptance of it demonstrably indicates it was considered viable, a good return (even without an altitude component). I am not a radar expert, but I believe this is what experts on the subject would basically state.

It wasn't the USAF that did or did not reject the final return - it was the NTSB. Their times are an approximation, thus any claim that they are precise must be dismissed.

The final return did have an altitude component - it was 0. All of the returns graphed are from the JSS and all are altitude returns.

For what it's worth the 0ft altitude reading is outside the margin of error of either of the two closest JSS sites.

Ergo, there has to be error in the data, either in radar or seismic or both (this is possible too).

What do we know for sure?

Somebody's made an error.

This does not follow. All data of this type has a margin of error. We've demonstrated that the margin of error for the radar data is well within the acceptable range to match with the seismic data. Therefore we do not have to conclude that human error is involved.

 

[pomeroo]

I have looked at the radar question quite a bit through my sojourns on this.

The 84th Rades had AA Flt 11 tracked for its entire flight and also brought into their study all radar returns from all sources for the flight, and what is telling about the final 8:46:40 crash time cited in the Commission Report that Ryan Mackey mentions above (post #460) when he observes correctly it was based upon a radar return at zero altitude on the graph, are the words found in the same report on page 2, 2nd to last paragraph. They shed light on this "zero altitude" issue:
"However, long range radar obtained from the USAF includes (less precise) altitude information based on primary radar returns..."

One aspect of basic Primary Radar is that it does not provide altitude information, but enhanced military versions of Primary Radar available to the USAF are capable of providing altitude, albeit sometimes resulting in "less precise" data. There are several zero altitudes returns seen on the graph provided, and are extreme examples of this lack of precision in altitude. But primary radar returns are viable, even without an altitude component. They are "returns".

Since the USAF did not reject this final return at 8:46:40 (at zero altitude), their acceptance of it demonstrably indicates it was considered viable, a good return (even without an altitude component). I am not a radar expert, but I believe this is what experts on the subject would basically state.

What Ryan writes in his post #459 sums it up best:


We may never know the truth on this discrepancy of times (seismic vs radar), but as Ryan points out, there was only one seismic event. Finis.

Ergo, there has to be error in the data, either in radar or seismic or both (this is possible too).

What do we know for sure?

Somebody's made an error.

But who made it, is this important?

(For the advancement of science, yes, but) In the final analysis, probably not.

Compared to the trauma of 9/11 that everyone has gone through one way or another, this difference in seconds doesn't mean much (at least to me it doesn't anymore).

But hopefully we keep learning. I think I now understand better the seismic aspects of the strikes on WTC1 and WTC2.

Although UA Flt 175 struck about the 81st floor at 540 mph--but was not a dead-center hit, while AA Flt 11 did hit dead-center but at a lesser speed of 440 mph and hit higher up {about the 93rd floor}, the WTC1 seismic event was longer in duration and stronger in magnitude than WTC2's seismic signature.

Those planes hitting those cores was comparable to "ringing a bell", and the "ring" in WTC1 was experienced by many in the basement levels as an enormous, destructive energy transmitted down through the structure's steel core that literally "bounced" off the bottom fixed to bedrock.

Then again I could be wrong. What do I know?
But was 9/11 an inside job? I don't think.

It takes intellectual integrity to abandon an erroneous opinion. You are to be commended.

 

[Pookster]

Excellent posts by Mackey, Quick, and Gumboot. It's cleared up much of the issue for me.

 

[R.Mackey]

Since the USAF did not reject this final return at 8:46:40 (at zero altitude), their acceptance of it demonstrably indicates it was considered viable, a good return (even without an altitude component). I am not a radar expert, but I believe this is what experts on the subject would basically state.

It probably was a good return, but like I said, we can't be sure it was an intact aircraft any longer. The impact blew off a large number of aluminum panels from the exterior, and shredded the plane itself -- just like chaff. It might easily have been the BEST radar return, from the standpoint of signal strength.

We may never know the truth on this discrepancy of times (seismic vs radar), but as Ryan points out, there was only one seismic event. Finis.

Ergo, there has to be error in the data, either in radar or seismic or both (this is possible too).

I agree. Craig, as I go through and correct my whitepaper, I'd like to put in a summary of your revised opinion, if you'd permit me.

Although UA Flt 175 struck about the 81st floor at 540 mph--but was not a dead-center hit, while AA Flt 11 did hit dead-center but at a lesser speed of 440 mph and hit higher up {about the 93rd floor}, the WTC1 seismic event was longer in duration and stronger in magnitude than WTC2's seismic signature.

Those planes hitting those cores was comparable to "ringing a bell", and the "ring" in WTC1 was experienced by many in the basement levels as an enormous, destructive energy transmitted down through the structure's steel core that literally "bounced" off the bottom fixed to bedrock.

This is quite true. NIST explores the various oscillating modes in Appendix K of NCSTAR1-5A. The Towers have several flexing modes, with different periods in N-S and E-W motion due to the different core dimensions, and also several torsional modes. The more off-center a Tower is hit, the more of these modes get excited, and the more coupling between modes. Since the periods are all different and we're dealing with a damped structure, transferring energy from one mode to another damps the motion even quicker. This is indeed why WTC 2 vibrated for a shorter period than WTC 1. Another reason is that WTC 2 suffered more core damage, and so transmission into the upper floors was a lot less efficient.

Anyway, as far as the timing goes, I'm pretty comfortable it was the 9/11 Commission that made the mistake. The LDEO numbers make sense given their charts. NIST also used video records, and those two agree with each other. The only outlier is the Commission, and given how sloppy the radar data is, it's really no surprise.

Again, I would not have used that last radar return value, but instead done a curve-fit of the descent and stopped when it hit the impact altitude, which is known to +/- 1 meter. If I do so, I get a number that agrees with NIST and LDEO. I'm actually quite happy about being forced to go back and re-evaluate it, since this answer is much more consistent.

 

[gumboot]

Anyway, as far as the timing goes, I'm pretty comfortable it was the 9/11 Commission that made the mistake. The LDEO numbers make sense given their charts. NIST also used video records, and those two agree with each other. The only outlier is the Commission, and given how sloppy the radar data is, it's really no surprise.

To be strictly accurate it's the NTSB that has the different impact time - the 9/11 Commission simply reproduced their number.

 

[quicknthedead]

The 84th RADES is a Radar Evaluation Squadron (that's what RADES stands for) and they're responsibly particularly for taking care of USAF radars and they often assist in investigations. They didn't track AA11 on radar themselves, as a unit (although 84th RADES personnel are located at all NORAD centers) The study for 9/11 was conducted by the National Transportation Safety Board in support of the FBI. The NTSB utilised both military and civilian radar to plot its track, and the military radar used, as I've discussed, is the ARSR-4 which comprises the Joint Surveillance System (JSS) utilised by both DOD and the FAA.

The ARSR-4 is was the only altitude-finding radar in operation at the time of the attacks, so we can say with a reasonable degree of certainty that the radar data presented in the NTSB report came from JSS sites. Further, because we can find out the location of radar sites in the USA via the FAA's National Airspace System Architecture site, we can determine which radar sites would have been used, and how accurate they would have been.

We know from specs that the ARSR-4 has an update rate of 12 seconds - that gives us a sweep margin of error. We know it has a vertical resolution of 0.17 degrees. Combine that with the range from the site to the WTC and we can get the altitude finding margin of error.

It's worth noting that the NTSB report cites the impact time of AA11 as "approximately" 08:46:40EDT. This is more than likely because the NTSB recognises that the radar data has a margin of error as I've discussed above.

Thanks, gumboot. I appreciate the information, especially on the ARSR-4 and the JSS.

Margin for error for altitude is one thing, but applying that same margin to time is another. A return's time is logged accurately no matter which station it comes from. Its time is what it is.

The margin for error you mention above is regarding altitude and can not be attributed to return times. This is covered in the NTSB's Feb 15, 2002 study on the recorded radar data...
http://www.911myths.com/Recorded_Radar_Data_Study--all_four_aircraft.pdf
...where they address the topics of position and time separately in two different sections.

In addition, NTSB citing the impact time of AA11 as "approximately" 08:46:40 EDT means something else (it does not mean the time has error due to the data having a margin of error for its altitude component).

Instead, "approximately" is used because it means the plane crashed within a few moments after 08:46:40 EDT due to the sweep of the radar antenna. The NTSB knew with certainty the last time of the aircraft's existence was 08:46:40 {based upon a viable return} and thus stated the crash-time approximately, but meaning fully, "We do not know the exact second it crashed but know it crashed moments after 08:46:40."

They did not know the exact altitude for that final return (it definitely was not zero elevation), but its time was precisely 08:46:40 EDT {UTC}.

This does not follow. All data of this type has a margin of error. We've demonstrated that the margin of error for the radar data is well within the acceptable range to match with the seismic data. Therefore we do not have to conclude that human error is involved.

Considering the above, this mystery still exists with two high-confidence data sets (radar & seismic) yielding two different times for the same unique event. This can not be if the event was only one. Therefore, logically, something is wrong with the data and/or its interpretation.

Data error--human error--a combination thereof?
I haven't a clue, but hopefully an answer will be found someday.

There is an appointed time for everything
And there is a time for every event under heaven
Ecclesiastes 3:1

 

[quicknthedead]

It probably was a good return, but like I said, we can't be sure it was an intact aircraft any longer. The impact blew off a large number of aluminum panels from the exterior, and shredded the plane itself -- just like chaff. It might easily have been the BEST radar return, from the standpoint of signal strength.

Grumpy over at physorgforum told me that a lot also (sorry, Grumpy; now upon further reflection, your idea does have merit). Yes, perhaps this could lead to solving this, if that last return were eliminated and an extrapolation done upon the remaining data. Grumpy would appreciate it (and so would I).

I agree. Craig, as I go through and correct my whitepaper, I'd like to put in a summary of your revised opinion, if you'd permit me.

Sure. Feel free.

This is quite true. NIST explores the various oscillating modes in Appendix K of NCSTAR1-5A. The Towers have several flexing modes, with different periods in N-S and E-W motion due to the different core dimensions, and also several torsional modes. The more off-center a Tower is hit, the more of these modes get excited, and the more coupling between modes. Since the periods are all different and we're dealing with a damped structure, transferring energy from one mode to another damps the motion even quicker. This is indeed why WTC 2 vibrated for a shorter period than WTC 1. Another reason is that WTC 2 suffered more core damage, and so transmission into the upper floors was a lot less efficient.

Thanks for this interesting analysis.

Anyway, as far as the timing goes, I'm pretty comfortable it was the 9/11 Commission that made the mistake. The LDEO numbers make sense given their charts. NIST also used video records, and those two agree with each other. The only outlier is the Commission, and given how sloppy the radar data is, it's really no surprise.

Again, I would not have used that last radar return value, but instead done a curve-fit of the descent and stopped when it hit the impact altitude, which is known to +/- 1 meter. If I do so, I get a number that agrees with NIST and LDEO. I'm actually quite happy about being forced to go back and re-evaluate it, since this answer is much more consistent.

I look forward to your results, Ryan.
Thanks.

 

[gumboot]

Margin for error for altitude is one thing, but applying that same margin to time is another. A return's time is logged accurately no matter which station it comes from. Its time is what it is.

This isn't strictly true. The Radar Data Processor at each ARTCC collects the radar returns and processing them, rejecting some, before displaying them on the radar screens.

The Antennae themselves are of course all turning at different times, so the RDP is receiving returns at different times from various radar sites, however it updates the screen all at once. That means the returns as recorded have a margin of error equalling the update time.

Now, in the case of ARTCC data it's possible (though unlikely) that radar data is recorded before it gets to the RDP.

However the military radar data, which is what was used to determine altitude, is received from the local ARTCC so it has already gone through the RDP before being sent.

 

[quicknthedead]

This isn't strictly true. The Radar Data Processor at each ARTCC collects the radar returns and processing them, rejecting some, before displaying them on the radar screens.

The Antennae themselves are of course all turning at different times, so the RDP is receiving returns at different times from various radar sites, however it updates the screen all at once. That means the returns as recorded have a margin of error equalling the update time.

Now, in the case of ARTCC data it's possible (though unlikely) that radar data is recorded before it gets to the RDP.

However the military radar data, which is what was used to determine altitude, is received from the local ARTCC so it has already gone through the RDP before being sent.

Interesting! Updating the screen "all at once" is something new in my understanding of radar. I thought system time stamping of returns was a function of the sweep, but now I see that it is also dependent upon the refresh of the screen as you have pointed out.

This changes everything in the time arena of seconds. Yes, you are correct -- the returns do have a margin of error that is dependent upon the update time.

Thanks, gumboot -- great stuff as usual from you!