9-11: The Scott Trilogy; an introduction, cutting to the chase

After four attempts to address William B. Scott’s three 2002 articles written for Aviation Week & Space Technology let me cut to the chase and get some things off my mind. Scott wrote at a time when accurate information was not available, the emerging story was incomplete, and the voices of the day had internalized events in their own way. His was an honest effort and is particularly useful to address three barriers, pitfalls if you will, to accurately understanding the events of September 11, 2001.

The barriers are: time compression; event conflation; and the imposition of post-facto understanding and conditions on both facto and pre-facto events.  The Staff of both the 9-11 Commission and the Congressional Joint Inquiry that preceded it grappled constantly with these barriers to understanding.   Most people writing objectively about the events of the day, the work of the 9-11 Commission Staff, or the work of authors, such as Scott, will remember to ‘walk in their shoes.’ And that especially means not looking through a post-facto lens.  Setting aside the post-facto lens barrier lets look a bit further at the other two barriers.

Time Compression

A good example of time compression is the difficulty the well-researched author, Lynn Spencer, encountered. Hers was another honest effort. In Touching History she established a reader’s road map by time-stamping each section; a chronological approach. That worked well until she attempted to tell the Andrews fighter story.  Spencer struggled to get the Andrews fighters back from the skies over North Carolina to Washington D.C. in time and space to deal with UA 93. The compression of time problem was such that Spencer had to abandon her reader’s road map and did not time-stamp that section of her book

In another example, Richard Clarke’s compression of time in “Against All Enemies” is not helpful in understanding the sequence of events that morning.  Clarke’s is also an honest effort, one which highlights the fact that participant recall is much like an eye witness account, helpful but not definitive.  And Clarke’s problems with time compression lead to the second barrier, event conflation.

Event Conflation

Clarke established a time hack for his readers, “It was now 9:28.”  That time was in reference to establishing an air defense “CAP over D.C.”  That meant that by that time, according to Clarke,  General Myers was in the Pentagon, the Air Threat Conference Call was underway, and “eleven aircraft [were] off course or out of communications.” [attributed to Garvey] Actually, Clarke conflated events as he compressed them in time, leaping nimbly over both barriers at once.  He also established that Norman Mineta was not in the loop by 9:28, despite Mineta’s own account.

Mineta, then Secretary of Transportation established in multiple accounts, including testimony before the 9-11 Commission, that he was there before 9:28 and that an aircraft “50 miles out” was AA 77.  Mineta’s story is the definitive example of event conflation.  The plane in question was UA 93.  Mineta and Clarke cannot both be right at the same time in terms of Mineta’s location in space and time.

Even the events of the day were conflated in at least two instances as they occurred; Delta 1989 was conflated with UA 93, and AA 77 was conflated with AA 11.  NEADS perpetuated the first conflation in its own story of the day. The plane that ‘meandered’ that day was not UA 93, it was Delta 1989.  We may never know the details of the second conflation other than the fact that it occurred and that it did prompt the Langley scramble.

Looking ahead to the Scott Trilogy

Researchers, writers, analysts, and investigators have the capability to overcome time compression and event conflation by using primary and secondary source information as an aid to recall during interviews. Spencer and Scott, especially, did not have the primary source information they needed to deal with either barrier. And as we will discover in the first Scott article some conflated events were also time compressed by Scott’s sources.

Okay, with that off my mind we can move on to Scott’s trilogy. And as we do, let me mention that is was from Scott that the Commission Staff obtained the lead about the Air Threat Conference Call. The tape of that conference is the single most important primary source of the day. For now, Commission Report footnotes stand as the best available secondary source of the day concerning that conference call.


Chaos Theory: NEADS and ZBW; strange attractors, indeed

In an initial article we established that the words ‘chaos’ and ‘chaotic’ were often used to describe events on 9-11. In a second article we began an inquiry into whether or not chaos theory could even be used to analyze events of the day. The answer for now is yes, metaphorically, and perhaps, theoretically. In this article we will continue on the metaphorical track and use the language of chaos theory, specifically strange attractors, to begin exploring the nation’s response.

Sensitivity to Initial Conditions and Strange Attractors

Chaos theory tells us we can never know in advance the initial conditions to which events are sensitive. However, in the case at hand we can, retrospectively, describe the conditions that led to NEADS and ZBW being the two focal points—strange attractors–around which key information of the morning flowed. Two organizations and four people—Colin Scoggins, Maureen Dooley, Shelly Watson and Stacia Rountree–became the self-organizing loci for action.

At all higher echelons–all the way to the White House–entities charged with the national level response failed to manage the flow of information.

More linear processes that did not work

In a previous article I listed four linear processes that were ineffective on 9-11.  We can now add four more processes used that day that were also ineffective.  The four are: the Air Threat Conference Call; the FAA’s Primary Net; the NOIWON conference; and Clarke’s own SVTS conference.  Commission Staff notes from an interview with an NMCC staff officer show that Clarke’s SVTS (Secure Video Teleconference System) conference and the Air Threat Conference Call were actually counterproductive.

Richard Clarke’s instinct was to try and command the attention of the highest level person he could find in each relevant organization. It is hard to fault that approach, even in retrospect, but, in hindsight,  it was exactly the wrong way to proceed when managing chaotic events.  The better approach and perhaps an important lesson learned was to find the highest echelon at which real time information was flowing between FAA and DoD; to identify the strange attractors. Enter, stage center, ZBW and NEADS.

ZBW and Colin Scoggins

ZBW had a problem. It knew where AA 11 was geographically, it did not know where it was spatially. When Mohammed Atta turned the transponder off at 8:20 ZBW lost all technical capability to determine the altitude of an aircraft gone astray. From that moment forward the last known altitude, 29,000 feet, would be a given.

Despite air traffic control observations that the aircraft was descending and despite very real information held at American Airlines that the plane was way too low, the much higher altitude would prevail as people sought to understand what happened in the immediate aftermath of the 8:46 collision of AA 11 with the World Trade Center north tower. Both the Otis and Langley scramble orders, nearly 45 minutes apart, specified an altitude of 29,000 feet.

One person at ZBW knew that NEADS could determine altitude from primary only radar returns, Colin Scoggins. Colin was on break off the operations floor (edited Nov 4, 2009) when events first unfolded but as soon as he learned that altitude was an unknown he went immediately to the operations floor; he had important information. Despite Colin’s best efforts and the effort of other personnel at ZBW and at New York Center (ZNY) NEADS was unable to identify and track AA 11.

NEADS and the Identification Technicians

The professional life of identification technicians–in this case Maureen Dooley, Shelly Watson, and Stacia Rountree–is measured in minutes that can be counted on one hand. That is the time they have to identify unknowns when tasked. The NEADS tapes capture them as a non-stop whirlwind as they worked to get a grip on the facts of the day. No one came closer to managing chaos that day. Right or wrong, they shared information they had in near real time with those who they believed needed to know. During a single hour they made or received at least 26 calls to four different air traffic control centers and one to their Canadian counterparts. Not one of those calls was to a third strange attractor, the FAA’s Air Traffic Control System Command Center (ATCSCC or, simply, Herndon Center).

Herndon Center, a Strange Attractor with no partner

The Herndon Center was buried so deep in the FAA organization chart that it was well nigh invisible.  The Center’s job was to manage air traffic flow; it had no role or place in the ossified FAA Region/Headquarter structure to manage everything else, including crises.  Yet most relevant real-time FAA information was held at or flowed to the Herndon Center.   Even when Herndon communicated accurate information concerning UA 93 to FAA Headquarters as it  was receiving a blow-by-blow account from Cleveland Center it made no difference at the national level.  Herndon may as well have been talking to itself.  So what was different here as opposed to ZBW?

Herndon, including the embedded military Air Traffic Services Cell,  had no operational DoD partner to talk to, no one to partner with to fight the battle that morning.  It wasn’t going to make any difference for the northern attack against New York City.  The element of surprise achieved by Mohammed Atta simply overwhelmed NEADS, ZBW and ZNY.  But soon after UA 175 struck the south tower Herndon knew a criticial piece of information:  “he said planes, as in plural.”  Yet the nation’s response to the southern attack against the nation’s capital never organized around the Herndon Center.

Linearity gave way to non-linearity as the southern attack developed and the nation’s response its self  became chaotic.  No extant linear structure was capabable of dealing with the situation.  In the absence of any effective government structure or process the response self-organized and NEADS and ZBW for a short time became  the strange attractors.  No one at any higher echelon had personal, organizational, or situational awareness to capitalize.  Ultimately, the NEADS/ZBW nexus was transient and the nation’s response rapidly devolved and self-organized around the only people that could make a difference, the passengers and crew aboard United Airlines flight 93.

9-11: NORAD briefings prior to May 23, 2003 hearing, some comments

I mentioned two briefings in the article “9-11: NORAD’s Sudoku Puzzle; a briefing askew, an addendum” that NORAD knew about before the May 23, 2003 hearing.  Thanks to some out of cycle work at NARA by paxvector and History Commons those two briefings are now available on the web.  Here is Cheri Gott’s 2002 briefing to the Satellite Toolkit (STK) Conference.  Here is her briefing to CONR to help prepare them for the hearing.

It is clear from both briefings, based on 84th RADES radar data, that NORAD knew the true story of the Otis and Langley scrambles.  They failed to tell that story at the first air defense hearing on May 23, 2003.  We also learn from these briefings that a “scramble” does not mean aircraft were launched.

A scramble defined

NORAD (Gott), on one sllide,  provides researchers and historians the definition of a scramble:  “Scramble = an order to get aircraft airborne as soon as possible.”  The important point here is that a scramble does not mean that planes were necessarily ever launched.

In my own work for the DoD Inspector General we examined the history of SEADS-directed scrambles concerning air activity over the Florida Straits concerning flight activity by the Brothers To The Rescue and Cuban response to that activity.  We determined the universe of scrambles over a several-month period and then further refined the data to focus only on those cases in which air defense fighters actually launched.  The number was quite small in comparsion to the number of “scrambles.”  My recall is that many of the scramble orders were cancelled because the information was determined to be spurious before the air defense pilots could get airborne.

A scramble order was and is a precautionary step, one in a sequence that can be stopped.  In the case of the Mission Crew Commanders at SEADS their task was to issue the scramble order and then quickly seek authority to actually launch.  The sequence of events on the morning of 9-11 was much the same.  The Mission Crew Commander and Colonel Marr scrambled Otis and immediately sought approval to launch.  General Arnold granted approval.

9-11: Delta 1989; an intervening variable, not a hijack

Aug 31, 2009 Addendum. Here is a BTS link to the official record of Delta flights that departed Boston Logan on September 11, 2001.  There is no listing for a Delta flight 89 and no indication that any Delta flight was scheduled Boston to Las Vegas.

Delta 1989 became a plane of interest briefly on 9-11 concurrent with the information that United 93 was hijacked and was presumed to have a bomb on board.  Delta 1989 became confused and conflated with United 93 in real time. and again in NORAD’s attempt after the fact to piece together the facts of the day.  It is worth noting that Delta 1989 was the only plane that NORAD at all echelons knew much about that morning and the only one that they were able to track.   The problem was they were out of air defense fighters.  Their effort to find anyone who could respond is clearly told on the NEADS tapes.  The effort was intense.

There is just one plane

Notations in Commission files, in contemporary documents of the day, and in testimony before the Commission that refer to flight “89” are simply shorthand notations for Delta 1989.  Lynn Spencer in Touching History acknowledges this; her notation style is ‘[19]89.’  To speculate otherwise ignores a simple truth; there is one and only one plane  in the radar files, both FAA and RADES, and in the air traffic control communications.  The primary source documents are definitive and conclusive.

The 9-11 Commission Staff sorted this out in the primary source information–tapes, transcripts, logs and radar.  Concerning the latter, we tailored a radar video, isolating just the two tracks, Delta 1989 and United 93, so that we could demonstrate to NORAD officials at every echelon that their story that the observed United 93 ‘meandering’ in the skies was, in fact, their watching the flight path of Delta 1989.  No one at any NORAD echelon disagreed with our findings.

How did this all come about?

The story starts at Boston Center.  Given the uncertainly that morning and the stark reality broadcast by Mohammend Atta, “we have some planes,” Boston Center saw a pattern of transcontinental flights originating at Boston.  In reviewing what it knew it determined that Delta 1989 was one such flight.  By that time Colin Scoggins had established a constant information flow to the Identification Technicians at NEADS. and by 9:27 NEADS knew that there were three unaccounted for aircraft.  The MCC/T log shows an entry at that time: “Boston FAA says another a/c is missing.”  A subsequent entry at 9:41 shows: “Delta 89 possible hijack Bos/Vegas.”

The NEADS Identification Technicians, whose story is well told by Michael Bronner, made and received multiple calls to five different FAA Centers that morning.  Among them were one to Indianapolis Center and one to Cleveland Center informing them of the hijack status of Delta 1989.  Cleveland Center, in direct communication with the Delta 1989 pilot, confirmed that he was not a hijack and that information was fed back to NEADS.   NEADS, meanwhile, established a track, B-89, on the aircraft which it forward told to NORAD, the only such track forwarded that morning.

Delta 1989, the only plane the NMCC will hear about from NORAD

To “forward tell” is to link a known track to a specific radar in such a manner that the track can be seen by NORAD echelons above NEADS.  NEADS established only two tracks of interest that morning, B-32 for the unknown that was AA 77 and B-89 for the known that was Delta 1989.  Track B-32 faded before it could be forward told.

The Air Threat Conference Call is conclusive concerning what was forward told.  When asked for an update NORAD informed the Conference at 9:44 that the only other hijacked plane it knew about was Delta 1989.  There was no mention of United 93 or any other aircraft.

The Tracking Story

A technician was assigned to track Delta 1989 and her conversation with another technician was recorded; a transcript is available.  The supposition that she was concerned about an airplane squawking mode 3 7112 has no foundation in fact.  Code 7112 is not an emergency code and was used by at least one airplane that morning, according to the 84th RADES radar files.  Here is what happened as revealed by the radar, the transcript, and the tape, examined concurrently.

Spatial Relationship of D1989, UA93 and Code 7112

delta-1989-slide2

Delta 1989 and Code 7112 both took off at 8:30.  Code 7112 departed from eastern Pennsylvania near the New Jersey border and flew northeasterly, as depicted on the linked slide.  The plane had no correlation to events of the day other than it became, briefly, a plane of interest to NEADS tracking technicians.  This slide also depicts the “meandering” path of Delta 1989 as it is vectored away from United 93 by air traffic control.

The technician who tracked Delta 1989 was first assigned to work a target in the Boston area.    At 9:42 her supervisor assigned her a target off of radar site 53, azimuth 288 and range 92 miles.  She picked it up at azimuth 287 and range 97 miles.  The tape  is clear, the transcript standing alone is not. Here is the audio information in graphic form. 

M3 7112

The radar shows two VFR aircraft and a transponding aircraft, Mode 3 7112, in close proximity at the bearing and range assigned to the technician.  She clearly describes this confluence of potential targets in her comments.  Thereafter, there is a gap in the recorded conversations that is not apparent in the transcript.  During that gap the comment Bravo 089 is heard in the background at 9:43:25. It is clear from the tape and the radar that the technician worked a target in the Northeast—-not 7112 by the way—-before she was assigned to track Delta 1989.

Delta 1989, not a hijack

By 9:58 the Identification Technicians understood that Delta 1989 was not a hijack, although the Surveillance Technicians continued the track.  By 10:09, after NEADS found out about United 93, efforts to scramble fighters from the Midwest in response to Delta 1989 were changed to focus on United 93.  An FAA chronology confirms the 9:58 time.  According to the log of the Air Traffic situation room at FAA Headquarters (separate from the WOC): “9:58, DAL1989 not a hijack.”

Nevertheless, Cleveland Center was uncertain and that uncertainty led Cleveland Tower, Cleveland Airport, and the FBI to treat Delta 1989 with suspicion after it landed as documented in a timeline compiled by Cleveland Air Traffic Control Tower.  It is clear from this document that Cleveland, Tower and Center, attributed  the suspicious information concerning Delta 1989 to NEADS.

The story according to Cleveland

“The OM [Cleveland Center] told the ATM [Cleveland Tower] that they had confirmation that the DAL flight could be a hijack and might have a bomb on board.  The OM stated that the pilot and company [Delta] both said it was a precautionary landing and there was no hijack.  The OM then said that the “Hunters” said it was a hijack. The ATM asked who the hunters were.  There was a lot of confusion at the center and the impression the ATM got was the ‘Hunters’ were in receipt of some intelligence that indicated the aircraft was a hijack.”

After all is said and done

NEADS is “HUNTRESS.”  They had no intelligence.  They knew what they knew from the FAA Centers, in this case Boston and Cleveland.  Such is the cloth from which myths are made in real time; proactive thinking, conflated information, and, ultimately, circular reporting. Such is the cloth from which myths are perpetuated; incomplete analysis based on partial information misinterpreted.

9-11: NEADS tapes and trancripts; a tutorial

While working on a Delta 1989 article it became apparent that a tutorial is in order concerning the NEADS primary source information of the day.  Fragmented attempts to describe events using something other than what the intelligence community calls “all source analysis” are prone to error, if not outright failure.  The same can be said for FAA tapes and transcripts, by the way.

The complete set of information needed to attempt any analysis of events in real time includes the radar files and the software to run them, time-stamped tapes, and any transcripts that were made.  It is possible to overcome the lack of a transcript by making your own.  In this article we will focus on NEADS tapes and transcripts and the 84th RADES radar files.  A helpful start is to recount events from a Commission perspective.

The Commission Experience

Early DoD document requests surfaced the RADES files and software and a partial NEADS transcript, the only transcript prepared by NEADS after the events of 9-11.  We understood that on the first trip to NEADS the organization would make tapes available.  When we arrived NEADS was in the process of making digital files and they were fed to us piecemeal as we began the interview process.  It quickly become clear that the partial transcript and provision of piecemeal tapes was not sufficient; we terminated the visit and caused DoD to be subpoenaed for all relevant files.

DoD provided the audio files but none of them had been transcribed.  The Commission Staff determined that the best way to proceed was to farm the audio files out to professional transcribing organizations.  One organization, Alderson Reporting, found the audio to be so confusing as to who was speaking that they opted to identify speaking voices and try and provide continuity of conversation on that basis.  In practical terms than means that no Alderson transcript is time continuous, although Alderson did insert time benchmarks to assist the reader.  The transcripts are helpful, but it takes “all soure analysis” techniques to get at the underlying events.

The technique the Staff used was to listen to the recordings using Adobe Audition so that individual conversations and transmissions could be accurately time stamped.  Alderson was careful to provide a NEADS-recorded time stamp in each of its transcriptions.  Concurrently, the Staff used the RADES RS3 software to display the radar files relevant to each transcript and tape.  In sum, it took then, and it takes now, all three techniques–transcript, tape, and radar–to understand the events of the day as they occurred.

A Specific Example

Currently, for the Delta 1989 article, I am reading the Alderson transcript for NEADS position DRM 1, Channel 19 SD2 OP, the channel for Major Anderson, as depicted on a schematic of who was at what position.  This is where Adobe Audition comes in handy.  It is clear that the recording on the tape is not continuous, although the tape itself is.  Nor is the transcript continuous.  And the obvious question is why does that come about?

There are two reasons.  First, because of the  voice identification methodology,  Alderson grouped together conversational fragments as if they cohered in real time, time gaps were simply omitted.  The duration of the gaps, some in minutes not seconds, can be determined using a program such as Adobe Audtion.

Second, Major Anderson was free to move about and plug into any given console, as needed.  When he unplugged from his primarly console there was no recording on the corresponding channel.  For example, just before 9:31 (1:00:02 tape run time)  a voice asked, “Major Anderson, you got a second?”  And sure enough, Major Anderson unplugged and the recording stopped.

Moreover, certain members of the crew, Major Fox, for example, were free to plug in anywhere they needed.  So, there is no specific channel for Major Fox, but his voice is heard throughout most tapes in at least background.  Further, the MCC, Major Nasypany, was free to “camp” on any channel he wanted to, so his voice is also heard on many tapes.  Even more confusing is that the three DRM “bled over” to each other during the copying process.

The net result on most NEADS tapes is a confusing blend of voices, background and foreground across the four main centers of activity–Surveillance, Identification, Weapons/Senior Director, and MCC.  So, researchers must take the time to become familiar with the SOCC layout, the participant voices, the radar picture, and the tactical situation at any given time.  Concerning the latter, it is also important to not impose post-facto awareness and understanding on facto (and pre-facto) conditions.

A few specifics from the SD2 transcript and tape (times rounded)

9:14:  NEADS to Langley asking how many aircraft they can sortie

9:23: An American Airliner (3d aircraft headed toward Washington)

9:24:  Scramble order heard in background

9:28:  American 11 mentioned

9:42: Delta what?

9:55: Over Lake Erie

10:07: MCC we got an air track…over the White House [radar needed here]

10:09:  ID type and tail

10:14:  Washington [Center] was reporting our guys…no aircraft over Washington

To be continued

I will add to and refine this article as I relearn more of what I thought I knew 5 years ago.


Chaos Theory: 9-11, thinking out loud

Can Chaos Theory Even Be Applied to 9-11?

This is a fundamental question. According to Nina Hall (Introduction to Exploring Chaos, Norton, 1993) “Chaos theory has resulted from a synthesis of imaginative mathematics and readily accessible computer power. It presents a universe that is deterministic, obeying the fundamental physical laws, but with a predisposition for disorder, complexity and unpredictability.” Does that understanding allow us to say as some observers have that events on the morning of 9-11 were chaotic; that ‘it is chaos out there?’ Certainly the language of chaos theory is useful to describe the events of the day, but can the theory, itself, be applied? Let us start by considering the affirmative and the negative as one source has it.

In the broadest sense the affirmative is supported by Ian Percival in his article in Hall’s compendium of a series of articles in New Science, “Chaos: a science for the real world.” Percival says, simply, “The theory of chaos touches all disciplines.” However, Percival later clearly supports the negative. “The state of Eastern European politics may look chaotic, but you cannot study a subject of this type using chaos theory.” Percival minces no words here. The seeming disorder of politics is not chaotic, though it may look so.

And that may be the case for 9-11, except that the events of 9-11 themselves were essentially a military attack and response, almost always ‘chaotic’ by anyone’s definition of the term. Military lore has it that no battle plan survives contact with the enemy. There is an exception if the attacker achieves the military principal of surprise. Mohammed Atta did just that and his battle plan proceeded as planned; it became ‘chaotic’ only in its final moments in the sky over Shanksville. The response, on the other hand, descended into chaos at multiple, discrete, times beginning with Atta’s first transmission, ‘we have some planes.’

So, what to do? We have a situation described in multiple instances as chaotic yet at least one voice in the literature cautions against the use of chaos theory as a basis for study. Reader, be warned, leap in logic coming up. Let’s turn to a self organizing source, Wikipedia to see what we can find. There are several entries but one that looks a bit promising is “Chaos theory in organizational development.”

Chaos as Metaphor

One of the first things the Wikipedia article tells us is that “‘parallels’ between organizations and the sub-atomic particles exist largely in terms of analogy (metaphorically) between two very different domains of activity.” Here we can acknowledge that the domain of activity on 9-11, described as chaotic, is different than, for example, quantum mechanics. Wikipedia introduces us to Charlotte Shelton. Shelton co-wrote “From Chaos to Order: Exploring New Frontiers in Conflict Management” in 2003. Wikipedia credits her in this way: “The introduction of chaos theory brings the principles of quantum physics to the pragmatic world.” This leads to a discussion in the article on self-organization, one of the specific observables when looking back at the events of 9-11. More on that later.

Time for another leap in logic.

A Paper You Never Heard About

In March, 1997, then Major Susan E. Durham, Ph. D. wrote a research paper at the Air Command and Staff College titled, “Chaos Theory for the Practical Military Mind.” Durham is clear that chaos theory is a mathematical theory and acknowledges the difficulty in application to social situations. Yet, despite what we learned from Percival about proceeding along those lines, Durham jumps right in cautioning, “when we don’t recognize the potential in well-behaved systems to deteriorate suddenly into Chaotic behavior, we also risk losing control.

Nothing on the morning of 9-11 was more well-behaved than the system of loading passengers onto commercial airliners and transporting them to their destination. Nothing had been more well-behaved in a decade than the need for a military response to a hijacked aircraft. There hadn’t been any. It was an orderly morning and linear systems were in place to manage the events of the day, or not.

Four Linear Systems (In order of appearance)

The first linear system challenged that morning was the FAA practice for handling planes and pilots who didn’t follow established procedures. On any given day planes and pilots deviate for benign and transient reasons. Controllers exercise various techniques to correct the situation which can take several minutes. When AA 11 went ‘nordo’ and then quit transponding Boston Center went through its checklist of techniques with no success. Its greatest fear was that the plane was experiencing serious mechanical failure and the Center took steps to allow a continued safe passage. At 8:24 what was a linear situation handled by a linear process suddenly became nonlinear. Mohammed Atta announced “we have some planes.”

The second linear system challenged was the airline practice to go into lock down when a plane was in distress. That system was alerted around 8:20 when the AA 11 flight crew reported a hijacking in progress to American Airlines. The debilitating result of lock down procedures is to create a black box, literally the equivalent of a black hole in space, an entity that sucks all available information into a closed system. The system proceeded at American Airlines (and, later, at United Airlines) as planned with the unfortunate result that no one outside of American Airlines knew what they knew. In and of itself the system did not become nonlinear until the company learned about AA 77 and it suddenly had two situations to deal with simultaneously.

The third linear system challenged was the hijack notification protocol. As spelled out in the staff statement at the Commissions June 2004 hearing, the protocol was laborious, unsuited, and never used. It was irrelevant.

The fourth linear system challenged was the search and rescue protocol. Indianapolis Center did not know it had a hijacking; it thought it had a plane down and implemented search and rescue procedures. The Center called the USAF Rescue Coordination Center (RCC) at Langley AFB to report the loss of AA 77. The RCC immediately initiated well established procedures and multiple state law enforcement agencies and the Civil Air Patrol were notified. No one outside the RCC community was notified and there was no apparent feedback loop to the Langley Air Force Base Command Post. This linear system remained stable that morning with the net result being that it was the source of false circular reporting confirming that AA 77 was down.

The missing link, feedback

The theory of chaos has it that feedback, itself, is a contributor to chaos. Percival tells us that “oscillating systems become chaotic because they possess an element of ‘feedback.'” That element “generates complex dynamics in simple systems.” Hall, herself, broadens our understanding. Her summation is that “Chaos also seems to be responsible for maintaining order in the natural world. Feedback mechanisms not only introduce flexibility into living systems, sustaining delicate dynamical balances, but also promote nature’s propensity for self-organization.” And it is, metaphorically, precisely on this point of self-organization that events of 9-11 turned, there was little feedback and some of that which did exist was counter-productive, for example the circular reporting of the crash of AA 77. Now, back to self-organization.

Self-organization

The Wikipedia article definition is: “Self-organization is the result of re-invention and creative adaptation due to the introduction of, or being in a constant state of, perturbed equilibrium.” All emergency response organizations, and 24-hour watch centers in general, live in this constant state. None of them know when the next call is going to come or what it will bring. The one certain thing is that equilibrium is transient and it most assuredly will be perturbed. Here the reference is to Dooley and Johnson (1995 “TQM, Chaos, and Complexity”). “Being ‘off-balance’ lends itself to regrouping and re-evaluating…in order to make needed adjustments and regain control and equilibrium.” Both NEADS and FAA’s Boston Center are organizations that live in a state of potential perturbed equilibrium. How they adjusted is one of the central stories of 9-11.

But, that wasn’t what was supposed to happen

The nation’s response was supposed to organize around set structures, two in particular. First, both the FAA and NMCC had procedures in place to ‘manage’ events that perturbed the equilibrium. Neither was effective, neither could talk to the other; they might as well have been on different planets.

Second, at the national level, things were supposed to organize around the White House Situation Room. The Secret Service removal of the Vice President to the surreal world of the PEOC virtually ensured that he would be out of touch and filtered from reality, not that the Situation Room was a much better place to be, information-wise. However, there at least the Vice President could have heard, perhaps seen, the Langley fighters overhead at 10:00, as captured on video in real time by a CNN camera crew.

Concerning both the PEOC and the Situation Room, I can’t help but recall George Plimpton’s classic description of a golf swing. Time.com has it this way: “His mind invents a nightmarish fantasy in which a team of inept Japanese admirals, located somewhere in his brain, shout useless instructions through the imaginary voice tubes of his creaking body machinery in an effort to help him hit the ball correctly:”

To be continued and a question

What if there had been feedback loops in place that in real time constantly informed FAA’s Herndon Center and Langley’s Command Post of unusual information available at, respectively, FAA’s Great Lakes Region and Langley’s RCC? Herndon knew about “we have some planes.” The Langley Command Post knew in real time that the air defense fighters had been placed on battle stations. Both the Great Lakes Region and the RCC knew that AA 77 had been lost. The time frame is shortly before 9:10, eleven minutes before Colin Scoggins sounded the false, yet oddly appropriate, alarm of an intruder from the north, and twenty two minutes before Danielle O’Brien noticed the real intruder from the west and sounded a second alarm.

9-11: NRO; not a factor, not an issue

Addendum, August 17, 2009. A document concerning the exercise received from NRO by the Commission Staff is now available.  Exercise inputs are included.  The stated purpose of the exercise is “NRO Emergency Response to a Small Aircraft Crash.”  The handwriting on the title sheet is mine.

There are two misconceptions about the NRO. One is that the NRO was actually able to track the hijacked planes on 9-11 by satellite and, by extension, should have been able to share that information immediately with others. Another is that the scheduled NRO exercise was somehow tied to other training and exercises that day. I am the single Commission staffer who worked the NRO exercise issue and I was on the Other Agency Team on the Congressional Joint Inquiry; the NRO was one of the other agencies. I know of no information which supports either misconception.

NRO, a Member of the Intelligence Community

In the words of an NRO contact several years ago; the NRO is nothing more than a long-haul trucking company which happens to build satellites and delivers and maintains them for its customers. According to its current website:

The NRO designs, builds and operates the nation’s reconnaissance satellites. NRO products, provided to an expanding list of customers like the Central Intelligence Agency (CIA) and the Department of Defense (DoD), can warn of potential trouble spots around the world, help plan military operations, and monitor the environment.

As part of the 16-member Intelligence Community, the NRO plays a primary role in achieving information superiority for the
U. S. Government and Armed Forces.

A DoD agency, the NRO is staffed by DoD and CIA personnel. It is funded through the National Reconnaissance Program, part of the National Foreign Intelligence Program.

Its DoD customers include the analytical agencies—DIA, NSA, and NGA (formerly NIMA). None of them or the CIA had the mission or the staff to track national air traffic; the responsibility of the FAA. The only relevant reporting from space on 9-11 was the SIR (Significant Infrared Activity Report) of the impact of each of the four aircraft. The Commission requested and received that data from DEFSMAC (Defense Special Missile and Astronautics Center) .   The SIR times for each crash were consistent with other data sources.

The NRO exercise

The NRO is just one of many large civilian and government offices which lives and works under one of the Dulles flight paths. The agency long knew that it had to schedule an emergency drill to address the possibility of a flight accident. It so happened that they scheduled such a drill on 9-11 unrelated in any way to subsequent events of the day. There was no correlation to any other training or exercises that day. It was the proactive planning of a single agency. The Commission asked for and received the scenario and exercise package. I don’t recall anything unusual about the contents; certainly nothing to pursue further. I will ask NARA if those files will be coming available. One way or another, interested people can file a FOIA request with the NRO if they wish to pursue the topic further.

9-11: The Andrews Fighters; an expeditionary force, not an air defense force

The Pentagon was struck at 9:38; Andrews Air Force Base is a short flight distance away yet its fighter assets were never a factor on 9-11. In the Mystery Plane article I discussed the sequence of planes that took off that morning from Andrews before and immediately after the Pentagon was struck; except the fighters which I said was a separate story. Here is that story.

The Andrews fighters were coming out of a previous day stand down because of a recent return from extended training in Nevada. Nevertheless, they had 5 planes and a like number of pilots scheduled for training. Three of those planes, the Bully flight of three, departed for training over eastern North Carolina skies at 8:36. The Andrews flight strip states, in part: “Bully 1…3/F16/R…ADW DCNG…1236.” The flight was led by, then, Major Billy Hutchison. Ultimately, four additional planes and pilots became available, but it was the Bully flight, specifically Hutchison, that would later be the first Andrews responder.

Before continuing further with the Andrews story we need to go back and revisit the Langley fighters for a moment. In the second Langley article I discussed the scramble procedures extant at the time. Those procedures specified that regardless of flight plan the Langley fighters would proceed on runway heading to an altitude of 4000 feet. That segment of the flight is clearly visible on the ground trace of the Langley fighters as determined from 84th RADES radar data.

And it is at that point, just short of the Delmarva Peninsula, that the Norfolk Tower controller and the Quit 25 pilot decided that the flight plan, 090 for 60 prevails over the scramble order and the Quit flight turned slightly right to its flight plan heading. But what does this mean in terms of the NORAD response to events of the day?

Had Quit 25 turned north at that point his flying time to the NCA was still on the order of 12 minutes or so at maximum subsonic, a rate of progress determined over time to be the most effective air defense solution to account for safety, to provide time on target, and to allow an approaching fighter a reasonable chance to spot his target. The Quit flight was not going to have a chance to intercept AA 77 even if the scramble had proceeded as NEADS intended. That meant that NORAD had no chance to effectively engage three of the four hijacked aircraft, AA11, UA 175 and AA 77. That leaves just UA 93 as the only hijacked airplane that either Langley or Andrews would be able to engage. It also mitigates the two askew segments of the Quit flight path to the NCA; they still arrived in time to guard against an approaching UA93; something the Andrews fighters did not do.

The misinformation that AA 11 was still airborne and headed south was the catalyst that got the Langley fighters airborne as soon as they were. It is only because of a proactive error by the Boston air traffic controller, Colin Scoggins, that the nation’s air defenders had any real chance to defend against UA 93; but the expeditionary force fighters at Andrews did not receive that same warning; their warning would come later as we are about to see.

Even though controllers at the scope level at Dulles TRACON had seen a “no tag” soon after 9:25 that observation did not become actionable until Danielle O’Brien saw the fast mover, checked her observation with her fiancé’ sitting at an adjacent screen, and sounded the alarm at 9:33 to National TRACON. Soon thereafter her supervisor notified the Secret Service and quickly the Service and National TRACON established an open line. The Service also picked up the phone and called the late General Wherley at Andrews.

General Wherley’s notes at the time, as reflected in Commission Records (Team 8 Box 8. History Commons is aware of newly released documents), show a time of 0930L (9:30 EDT). The time was actually closer to 9:35. By that time the Secret Service was also following the fast moving unknown, now tagged with an “S,” by National TRACON, visible to the Service. There is a misconception that because the Secret Service had a working relationship with National TRACON that the Service could see any FAA radar feeds from anywhere in the country. That is not true, the only thing the Service could see was what National TRACON was seeing, nothing more.

So, the Andrews alert came at 9:35. It would take them the better part of two hours to get armed fighters in the air with the authority to act. The Langley fighters, scrambled at 9:24 established a CAP over the nation’s capital 36 minutes later, albeit without authority, despite two askew flight segments. How can there be such a difference?

The difference is in the roles and missions of the day, a specific determination by the Department of Defense in order to most effectively use its resources. Just as the three services and the Marines have specific roles and missions, so do their major components and subcomponents. We can leave a more detailed discussion of that for another day, suffice it to say here that the air defense mission and the role of the Langley (and Otis) fighters is distinctly different from the expeditionary force mission and the role of its fighters, such as the Andrews contingent. Only the CONUS air defenders had the tactics, techniques and procedures in place to respond rapidly, and there were just 14 fighters with that mission at seven locations; just two of those locations and four of the fighters were immediately available to NEADS.

Andrews went to work immediately to upload its aircraft and respond, but that took time, time that they did not have. They also recalled the Bully flight, first Bully 2 and then Bully 1 and 3. Even though the Langley fighters had effectively CAPPed the nation’s capital by 10:00 that was not understood or even known within the NCA, including the White House. So when things turned serious with the approach of a now notional UA93, as seen on traffic situation displays at multiple locations, the pressure on Andrews intensified; they had to do something. None of this had anything to do with AA 77 despite accounts to the contrary.

The Andrews flight strips show that Bully 2, who came back alone and well ahead of Bully 1 and 3, landed at 10:14 out of fuel. Bully 1, a flight of two F-16s landed at 10:35, low on gas; however, Bully 1 had sufficient fuel to take off again in response to an unknown coming down the river. By then, the notional UA 93 had “landed” at National at 10:28. There was nothing to intercept. Here is a ground trace of the flight of Bully 1 based on data from 84th RADES.

Major Hutchison was airborne by 10:39 and flew three specific routes. First, he circled to the east back directly over Andrews climbing to an altitude of 3600 feet, shown in red. He then flew directly to the Pentagon, shown in green, and overflew the building at 600 feet shortly after 10:42, exactly as described by Creed and Newman in Firefight to Save the Pentagon. He then turned south, shown in yellow, and climbed back to 3600 feet before landing back at Andrews a short seven minutes after taking off. He was not given “weapons free” authority by General Wherley. The Andrews flight strips show him cleared for takeoff at 10:36 and back on the ground at 10:47, an eleven minute period. Flight strips are not definitive; the radar shows that, at best, Hutchison was in the air for just over seven minutes.

His presence was no longer needed. A pair of Andrews fighters, guns only, and with only verbal (text modified on July 19 based on Commission work files). “weapons free” authority had taken to the air. A handwritten flight strip shows “CAP 1, 2 F-16, airborne at 10:51. (Here I give the benefit of the doubt, the handwriting is not clear and it could be 10:57.)  Finally, at 11:12, Andrews was able to launch the Wild flight of two F-16, fully armed, and General Wherley did give them “weapons free” authority.  The Wild pilots were the other two scheduled for training that day.

Given that all hijacked aircraft had been accounted for, the only issue left by the time the Wild flight got airborne was command and control of the skies over the nation’s capital. That became a bit contentious between the Langley and Andrews fighters. Lynn Spencer in Touching History has a good account of the ‘battle’ for command and control. Spencer, by the way, is the only person who has told an unsung story in the skies that day, the efforts of civilian pilots to help air traffic control. It is a story well told.

Air defense forces and expeditionary forces compared. It took the expeditionary forces, Andrews, 97 minutes (9:35 to 11:12) to retool and reconfigure and get fully armed planes with shootdown authority into the air over the National Capital Region.  It took the air defenders, NEADS/Langley, 51 minutes (9:09 to 10:00; recall that the Langley fighters were placed on battle stations at 9:09) to get fully armed planes over the National Capital Region, but with the authority to only identify by ‘type’ and by ‘tail.’   In sum, given the attack and response as it unfolded that day, neither force could have done anything to prevent the WTC and Pentagon tragedies.  Only the air defense force was in position to do anything about United 93 had its passengers not taken matters into their own hands.

9-11: NORAD’s Sudoku Puzzle; a briefing askew, an addendum

This article provides information as it was briefed to the Commission on May 23, 2003. It complements the article “NORAD’s Sudoku Puzzle; a failure to tell the truth,” which should be read first. In that article you were introduced to Colonel Scott, USAF, retired. It is Colonel Scott who presented the briefing to the Commission. The video of that briefing is available on the Commission’s web site; it does not, however, clearly show the briefing charts used. The charts have recently been made available by NARA and the purpose of this addendum is to share them.

Colonel Scott made it very clear that times on his charts were derived solely from logs, primarily the NEADS MCC/T log; no other source. It is understandable, then, why he would brief a Pentagon impact time of 8:43, for example, as opposed to the actual time. On the other hand, it does not make clear why he would show UA93 impacting near Pittsburgh, as show on his introductory chart.

More important, however, is his treatment of the flight path of the Otis fighters and to some extent, the flight path of the Langley fighters as they neared the capital. Scott gave  the impression that the Otis fighters hugged the coast and proceeded directly to New York City, consistent with the account given by the pilots during interviews in 2002. When asked during an interview at CONR why he blurred the scramble path Scott claimed limitations of the Powerpoint program, a disingenuous answer, at best.

The rest of Scott’s charts were timelines, included here for the record. Better renditions will become available when paper copies in Team 8, Box 8 at NARA are uploaded.

NORAD Hearing First Time Chart

second-time-chart1third-time-chart1fourth-time-chart1

See the May 23, 2003 hearing article for a discussion of the discrepancies. In sum, NORAD read the MCC/T log wrong, twice; first when they prepared their Sep 18, 2001 press release and again when they prepared for the first air defense hearing. In my interviews with both Michael Bronner and Phil Shenon I attributed this to shoddy staff work, primarily at NEADS, which was not adequately vetted at either CONR (Generals Arnold and McKinley) or NORAD (General Eberhart).

The NORAD staff had clear and explicit information available; the radar files, the tapes, and the logs of the day.  On September 25, 2001, in a memo to the US Space Command Directorate of Analysis the 84th RADES included an analysis of radar data for 11 Sep 2001 which included radar text files and Powerpoint slides showing flight paths. On June 3, 2002, a NORAD analyst, Cheri Gott made a presentation to the annual Satellite Toolkit (STK) Conference which was based on 84th RADES data. Moreover, she followed that with a May 13, 2003, briefing to CONR just 10 days prior to the first Commission hearing on air defense. A purpose of Gott’s staff work was to produce a product for the CINC (Gen Eberhart) to use from a Headquarters perspective. Relevant Gott source material is in Team 8, Box 8 at NARA.

NORAD’s failure to provide an accurate accounting of the day is inexcusable for any staff and particularly for a staff that had been at the air defense business for decades.  NORAD failed to accurately read its own logs, tapes and radar files.  Together with FAA it failed to reach agreement on the basic facts of the day in the immediate aftermath when events were fresh.  The NORAD staff failed to adequately prepare its CINC for questions it knew were coming during General Eberhart’s annual testimony to Congress.  Ultimately, NORAD failed to tell the story of the valiant battle by Alpha and Delta flights at NEADS; a story that General Arnold conceded was better than the one they did tell.  Thanks to Michael Bronner that story has been told.

9-11: The Mystery Plane; not so mysterious

Last year, the researcher and writer, Mark Gaffney, wrote The 9-11 Mystery Plane. He speculated that a “white plane” seen in DC skies was somehow nefarious in ways that simply don’t track with primary source information of the day or with the body of information accumulated by the 9-11 Commission. In part, the author relied on eye witness accounts and post facto media reporting to try and make his case. He did use the radar files of the day and did refer to the NEADS tapes, but he did not put things together correctly. Following is what the primary source information tells us about the flights from Andrews Air Force Base on 9-11. A related Commission work file is at this link.

The Flights

Four aircraft of interest took off from Andrews that morning, excluding the fighters; a separate story. Flight strips generated by Andrews Tower provide this detail:

Word 31: B742 [E4B], a NAOC (National Airborne Operations Center) flight, according to the flight strip, that staged at 1136 hrs (subtract 4 hours from all times for EDT). Word 31 was airborne at 1327.

Venus 22: A Gulfstream 3, airborne at 1316 on a scheduled flight to West Virginia; it landed back at Andrews at 1354

Gofer 06: A Minnesota Air National Guard C130H, airborne at 1333, ultimately an observer to the aftermath of the impacts of AA 77 and UA 93.

Venus 77: B747, airborne under VFR rules at 1345; it became the “white plane.”

In sum, four aircraft departed Andrews Air Force Base between 9:16 (Venus 22) and 9:43 (Venus 77), three on scheduled flights and one, Venus 77, in reaction to events of the day.

Venus 22

There is little remarkable about VENUS 22. It declared for Lewisburg, West Virginia but returned to Andrews after a short 30 minute plus flight. At the moment Dulles TRACON sounded the alarm about the fast moving unknown (AA77), Venus 22 was approaching Waldorf, Maryland, headed for a holding pattern near the Chesapeake Bay, south of Annapolis, to burn fuel. Venus 22 was in that pattern when AA77 impacted the Pentagon. The plane landed at 9:54 and remarked to air traffic control on approach that it looked like something happened at National Airport. Here is a Google Earth trace of the flight path, created from 84th RADES text files.

Word 31

Word 31 took off routinely and proceeded on a route to the Midwest. It was routed around the National Capital Region to the south and, outbound, passed well south of AA77 like a ship in the night.  It is a bizarre fact of 9-11 that a National level airborne command post flew westerly over Virginia at the same time the hijacked AA77 flew generally the I66/Rte29 corridor on its eastern approach to its target. Neither was aware of the others existence.  The 11th slide in this Commission file, a screen print from 84th RADES data, shows the relationship of Word 31 (labeled Sword 31), AA77, Gofer 06 and, according to Andrews flight strips  two military T2 aircraft, Bobcat 14 and Bobcat 17, transiting the area north-to-south, at altitude.  At the time AA77 was approaching DC, Word 31 was being routed around DC to the south, a normal routing according to controllers at Dulles TRACON.  Air Traffic Control tapes confirm that Word 31 was routed around the National Capital Region before anyone knew that AA77 was approaching.  Reagan National tape (big file, 23 MB) is at this link.

1 DCA 105 TYSON 1324-1350 UTC.mp3

Gofer 06

Gofer 06 was in the takeoff queue behind Word 31 and it was delayed briefly because of B747 wake turbulence. Quick action by FAA controllers at National Tower/TRACON resulted in Gofer 06 being the only military aircraft that day to see one of the hijacked aircraft and the only one vectored to follow any of the four aircraft. En route home, it again came to air traffic control attention. Cleveland Center directed Gofer 06 to turn right to avoid oncoming traffic, UA 93. Gofer 06 reported smoke soon after the turn that turned out to be from the Shanksville site, the first such report to FAA.

Venus 77

Venus 77 took off at 9:45 in a hurry, VFR. The pilot declared for Offutt AFB and on its climb-out western leg Venus 77 passed just north of P56, the restricted flying area over the National Capital Region. It then declared for an orbit south of Washington and turned back east in order to proceed south for orbit. It is during that west leg and return east that Venus 77 was noticed by multiple observers on the ground; it became the ‘white plane’ captured on CNN raw footage taken near the White House. Venus 77 then declared for an orbit in the Richmond area where it set up a north-south racetrack holding pattern. It altered the orientation of the racetrack once to point toward Barksdale AFB. It exited the pattern by returning north and then turned west and proceeded out of the area. The flight of this ‘mystery’ plane is consistent with support for the possible return of Air Force One to Andrews and for its actual flight to Barksdale. This video clearly shows the relationship between Venus 77 and Air Force One.  It also shows that Air Force One had no fighter escort until shortly before it arrived at Barksdale when the Texas Air National Guard arrived on the scene.

There is no correlation between Venus 77 and the Langley fighters. There is no primary source information to support the contention that the Langley fighters were vectored to intercept Venus 77.