I recently obtained some of the seismic data concerning UA 93 and that provides an opportunity to discuss an old issue about 9-11, the impact time..
Readers may recall that there was at one time some speculation that a study based on the seismic data suggested that UA 93 impacted at 10:06, three minutes later than all available primary source information indicated. The Commission Staff treated that speculation seriously and ran the issue to ground.
This article is the story of the data and of how the Staff dealt with one of the 9-11 family questions.
The Incorporated Research Institution for Seismology (IRIS) has protocols for assisting educators, researchers, students, committees, and the public. The link provided is to the protocol for researchers. I found the IRIS staff helpful in obtaining the information available.
To run the data I used GEE, Global Earthquake Explorer, and an IRIS-provided link to AmaSeis. (Amateur Seismometer). Readers with seismology experience will find additional programs available from IRIS at the same link.
My point of contact provided me the files I actually used. He obtained them by submitting a Seismic Query to determine what data were available. He then submitted a request via email using BREQ-FAST. Finally, he converted the files to .SAC format via rdseed.
An undated report, “Seismic Observations during September 11, 2001, Terrorist Attack,” concluded: “we infer that the Flight 93 crashed around 14:06:05 +/- 5 (UTC).” The primary author, Won-Young Kim considered data from four seismographic stations. From closest to Shanksville to the furtherest, the sites were: Mont Chateau, (MCWV); Standing Stone, (SSPA); Soldier’s Delight, (SDMD); and Millersville, (MVL)PA.
In his vertical component assessment, Kim initially considered all four seismic stations. He assessed that the signal to noise ratio was too small for both the closest site (MCWV) and SDMD, on the order of 1:1. Therefore, he reduced his analysis to the Standing Stone station where the ratio was 2.5:1. He also established that the ratio at the most distance site, Millersville, was about 2:1.
Kim based his analysis on the Standing Stone station only, ignoring the fact that the data from the closest station was inconclusive. Further he extrapolated only two minutes of data beginning with an estimated origin time of “14:06:05 (10:06:05)” We have no evidence that Kim examined earlier times, specifically 10:03.
IRIS has available only the data from Mont Chateau and Standing Stone, the two stations closest to Shanksville. My contact provided 10 minutes of data, from 10:00 to 10:10.
I examined the data as far as my expertise would take me (undergraduate geophysics classes) and found nothing anomalous to support a time of 10:06 rather than 10:03. I leave it to practicing seismologists that might be interested to provide a more definitive assessment.
Let’s now return to the issue raised to the Commission.
The Question Posed
A 9-11 Family Member told me via e-mail that seismologists agreed that UA 93 crashed at 10:06, not 10:03 as NORAD said. I took that request for action and we obtained a copy of the report. It was not a U.S. Army document; the Army referred us to the Maryland Geological Survey website for a copy.
We found that the time in the study, 10:06, was at odds with all other primary source information, the convergence of which supported a crash time of 10:03. Despite that convergence we took the time to fully consider an alternative time of 10:06, as requested.
Primary source information
Data from the UA 93 Cockpit Voice Recorder (CVR) and Flight Data Recorder (FDR), air traffic control communications, FAA radar, and Air Force radar established a crash time of 10:03.
The seismic data was also primary source information; the study that inferred at time of 10:06 was not. Statements that the seismic data show that UA 93 crashed at 10:06 are inaccurate. Further, statements that leading seismologists and scientists concluded that UA 93 impacted at 10:06 are disingenuous. The work was that of a single person, as it turned out. Here is how we proceeded.
The study was sponsored by the Department of Natural Resources, State of Maryland under contract number SMGS/AG1-01-075. The authors were Won-Young Kim, Lamont-Doherty Earth Observatory of Columbia University, and Gerald R. Baum, Environmental Geology and Mineral Resources Program, Maryland Geological Survey.
Our first step was a request to the Maryland Geological Survey (MGS) asking them to validate and verify the study done by one of its employees. The MGS declined stating that the work was primarily that of Won-Young Kim and that we should address our request to him. The MGS, literally, walked away from the issue.
The second step, then, was to contact Mr. Kim, which I personally did. Over time, I exchanged a series of emails and phone calls with Mr. Kim. I informed him that multiple other primary sources of information did not support his work. His position was that, if so, he still had confidence in his work. Ultimately Mr. Kim agreed to a statement that the seismic data was not conclusive concerning the crash of UA 93.
Given that statement, we reconfirmed that the preponderance of evidence supported a time of 10:03 and that became our position. There was, however, another avenue we could explore.
Going the extra step
Even though we had a high degree of confidence in the 10:03 time, we decided to treat the two times as equal and sought and obtained additional primary source information, infra-red satellite data. We asked the Defense Special Missile and Astronautics Center (DEFSMAC) for the times it had established for the impacts of all four hijacked planes.
We had noticed in the Cheyenne Mountain Operations Center (CMOC) logs that some “IR [infrared] event” times had been entered that morning and assumed that DEFSMAC had recorded all four “IR events.” They had, and the data provided was consistent not only with the 10:03 time for UA 93 but for the other three impact times, as well. The DEFSMAC response is in the classified Commission files held by NARA.
To put it another way, we gave the two disparate times–10:03 and 10:06– equal weight and went to a tie breaker, even though there was no compelling analytical reason to do so.
A comment on DEFSMAC
Accurately reporting “IR events” is vital to the defense of the nation. Missile launches are serious business and DEFSMAC provides the earliest possible indications and warning of such a potential threat. The DEFSMAC times, alone, are sufficient to establish the impact times of the four hijacked aircraft.
In the end
We spent considerable time running the seismic issue to ground. Yet the answer did not matter, it wasn’t the “right” answer. During the reading of Staff Statement 17 at the final air defense hearing we presented the list of sources for 10:03 time of impact for UA 93. Behind us, someone in the audience commented, what about the seismic data?
Questions answered that don’t provide the expected or wanted answer become…more questions.