by Robert W. Butsch

Unless we are willing to escape into sentimentality or fantasy, often the best we can do with catastrophes, even our own, is to find out exactly what happened and restore some of the missing parts -- hopefully, even the arch to the sky.
Norman Maclean
Young Men and Fire

All photographs are accurate. None of them is the truth.
Richard Avedon

If you happen to have a copy of the Columbia Accident Investigation Board Report handy — all six volumes, not just the Volume I main report — and if you have a fair amount of time to kill, you can search out the sad markers of Columbia's tragic final re-entry on February 1, 2003. You will discover that they are plentiful in number and well documented, and that their timing often is cited to tenth of a second precision. You also will discover, if you know where to look, that there are two versions of that timing. What follows is the tale of how this little bit of confusion in the official documentation of one of the more important events in the history of space flight came about.

I have written it for a number of reasons. As a life experience it is unique among the many and varied experiences of which I have heard others tell, and was to me particularly interesting if not particularly pleasant. Also, though the events recounted here could be verified easily enough, it seems that I am really the only person who knows the whole story. There would, of course, be nothing wrong with my taking it with my when I go; but likewise, I think, there is nothing wrong with leaving it behind. Perhaps someone else will be interested.

And, finally, I subscribe to the belief that humans comprise a species so — often destructively — dominant on Earth that it has no long term future other than in space. The collection of events comprising the STS-107 mishap represent one of the milestones along the road that will — or, just as importantly, will not — take us to that future.

I wrote this memoir in 2003 shortly after the events it reports occurred. I have redacted some material from this version to respect the privacy of NASA employees and others that I dealt with, and refreshed some of the text to accommodate the passage of twenty years’ time.

At 7:59 a.m. on a lovely Saturday in February of 2003 I sat in the backyard of my home in Duncanville, Texas, eight millimeter video camera in hand, scanning cloudless blue skies for Space Shuttle Columbia. It was at that moment streaking its way home to the Kennedy Space Center in Florida after nearly sixteen days in orbit. Despite my vigilance I was caught off guard when it appeared. It was traveling twelve thousand miles per hour, so I suppose I should not have been so surprised to see it suddenly materialize as if conjured up out of nothing, a tiny, bright white dot in the west. I jerked the camera's viewfinder up to my eye and pressed the "record" button. I stood up — it had been a stupid idea to sit down — and panned rapidly, following the rushing, silent dot as it traced an arc high across the southwestern sky. Early on a new white dot fell away. A few seconds later I can be heard on the recording saying, "Whoa, baby, parts are falling off." This is testament to the fact that so often the words coming out of our mouths do not accurately reflect what is going through our minds. What was going through my mind was that it looked like parts were falling off. I had no inkling of the disaster unfolding before me. And so I continued taping until the orbiter seemed, by the illusory evidence of a prominent smoke trail, to have done nothing less than fall into the morning sun.

Minutes later I was back indoors checking NASA TV via the Internet. By that time, 8:14 a.m. CST, I should have been seeing views of Columbia's circling approach to runway 33 at KSC, but all that was on the computer screen was images of Mission Control in Houston. I realized that something obviously had gone wrong. I turned up the computer speaker volume in time to hear the Flight Director declare a contingency.

After the initial shock subsided, it dawned on me that the video I had been making merely to add a little footage to the family collection was going to be a small piece of evidence in a very big investigation. It was the beginning of an experience far stranger than I could have imagined beforehand.

Despite having been a space nerd for for more than forty years I had payed scant attention to STS-107. Following on a string of expeditions to supply the International Space Station, this mission had been sent up by NASA as a scientific research flight to explore biological and physical effects of microgravity. It carried a diverse crew of seven, one of whom was a veteran of Israel's legendary air strike against an Iraqi nuclear power facility. It was the 113th Shuttle flight and had been showing every sign of becoming the 112th successful one. What most people were not aware of at the time was that eighty-one seconds after liftoff on January 17 the huge external fuel tank had shed a piece of its orange insulating foam, which in turn had disintegrated against the leading edge of Columbia's left wing. What no one at all was aware of was that this had inflicted a mortal wound in the form of a hole in the leading edge's reinforced carbon carbon heat shielding. Engineers reviewing films of the launch spotted the errant foam. However, equipped with what hindsight revealed to be a deficit of good information and a surplus of optimism, they and their superiors deemed the orbiter safe for re-entry. This perhaps was for the best, as from the instant the foam impacted the wing there was no chance for Columbia’s successful return to Earth.

And so on February 1 at the appointed hour over the Indian Ocean Columbia fired its Orbital Maneuvering System rockets, disposing of just enough velocity to fall out of orbit. This still left a significant residual of 17,000 mph or so, and as the orbiter began to encounter a few molecules of the upper atmosphere these were violently compressed out ahead of it. Their temperature soared to thousands of degrees. Normally this super-heated air would remain disconnected from most of the orbiter's surface, reducing the risk of over-heating. However, disruption to the smooth flow of air over the wing caused by the damage created a path through the boundary layer and into the hole in the RCC. Thus the super-heated molecules gained access to the aluminum skeleton underneath which quickly began to melt. At the same time, the increasing drag on the left side produced by the now worsening wing damage tried to pull Columbia out of control. By the time it was within range of my camera it was no small miracle that it was still flying.

It was by chance only that I was watching. At work on January 31 I had stumbled across an Internet news item noting that the Shuttle would be landing the next morning. Knowing that it was in a relatively low inclination orbit since it was not visiting the ISS, on a hunch I checked a NASA Web site and discovered that, for its first re-entry attempt, its path indeed would pass quite close to Duncanville. The weather was predicted to be ideal. It looked like an opportunity worth getting up early for. That evening after charging my camera's battery I turned on its time and date display. I normally did not do this, out of a conviction that it made my videos look like they were shot by an amateur, which, of course, they were. It was opportune that I now chose to ignore this vanity.

Ten hours later I was the same amateur, but one with something more than a home video on my hands. Despite shock and sadness at the loss of seven lives snuffed out many tens of thousands of feet above the ground and the severe — possibly mortal — blow to the Shuttle program, in dispassionate moments I could imagine the accident having potential for becoming something of a minor adventure. It was apparent virtually from day one that the root cause was probably the foam loss at launch, and that images of the actual breakup would not be of help proving this. However, the ensuing investigation would be quite thorough and concerned with far more than just the root cause, and all photographic evidence would certainly be considered of particular importance.

With this last point in mind I returned to my backyard with the camera to document the scene on tape. I taped all sorts of stuff: the spot where I had been standing while recording, all of the ground structure that appeared in the video, which direction was due south. I was thorough, to a fault perhaps; but it seemed to me that this was a once in a lifetime occurrence, which meant only one chance to do things right.

Two hours after the accident I began the task of determining how accurate the camera's clock was. I discovered that, while setting the seldom used time and date display the night before, I accidentally had bumped the year from 2003 to 2004. This would be an embarrassment should the video ever be made public, but the time of day display was what really was important. From my initial analysis (which I also documented with the camera, as described below) it appeared to have been running eleven seconds behind the actual time of day. A little arithmetic produced a start time for the video of 7:59:42 a.m. CST.

That evening I found a NASA Web site for public submission of photographic evidence. However, it was for uploading digital images. My camera was a humble 8mm, not even Hi8 let alone digital. After more searching I found an email address set up for the benefit of anyone who possessed evidence related to the accident. The email that I sent resulted in a phone call the next day from NASA's Johnson Space Center in Houston. I was given an address to which I was asked to send my video — specifically a VHS copy and not the original. I got back on the Internet and went to Terraserver to get a precise latitude and longitude for my backyard. First thing Monday morning I dispatched a dub of the tape and the extra video documentation (including the first pass at a time calibration), plus my latitude and longitude by overnight air freight.

I expected, naively perhaps, a straightforward process: send the tape; be contacted by NASA to verify a few facts; relax and wait to see information from the video expertly interpreted and put to good use; receive a brief acknowledgment in the investigation's final report; end of story. This isn't quite what happened.

One problem was that I found myself obsessed with analyzing the video. Maybe this was not surprising as I was employed as a database analyst. In school, math had been a giant bore possessing the added disadvantage of being difficult. But once it stopped being homework and I stopped being handed mediocre grades, I became fascinated with its tactical employment against real-world problems. Now fate had tossed me a real-world practical math problem on a scale far grander than anything I likely would ever see again. There was no way I could resist the temptation to try to figure out things for myself.

I refined the timing of the video, eventually determining that it actually had started at 7:59:44 a.m. CST, two seconds later than my initial estimate. I accomplished this by using a stopwatch shortly after the accident to check the difference between the camera's clock and the one on my cable TV converter (while videotaping the converter clock), then a few days later checking the converter against a time display on the Weather Channel, and finally cross-checking these measurements against the National Institute of Standards and Technology's short wave broadcast of the time of day (WWV). By this time I had acquired, courtesy of a colleague at work, a digitized version of my video. With the help of Microsoft Windows Movie Maker I was easily able to determine times for individual video frames.

This hip-bone-connected-to-the-leg-bone process was likely a source of problems I was to encounter in the weeks to come. I did discover that the camera's clock was quite unstable. It was good that I had done that original stop watch time check on the morning of the accident rather than putting it off.

On Friday, February 7, NASA finally made public the first bit of information that I could use to place my video in some sort of context. This was the time of what came to be known as "loss of signal" — when the last bit of decent quality data transmitted from Columbia had been received — and it occurred at 7:59 a.m. CST (no seconds were specified). LOS would become a fixed point of reference, Columbia's last contact with the world as a functioning machine controlled by people and computers before turning into pieces of debris controlled by the principles of ballistics. I concluded that the investigators would likely find the video I had made to be of particular interest because it appeared to span this period of awful transition.

For a while, in fact, I thought that it might be the only one that did. I learned later that this was not the case. CBS News actually had a video shot not far from me — in Arlington, Texas — up on their Web site that covered the same time period; but I couldn't tell this because, as I discovered eventually, the beginning had been edited out. CBS (and probably the videographer as well) undoubtedly did not understand the circumstances well enough to realize what they had done. I was to be frustrated continually by this difficulty finding — among friends, family, casual acquaintances, the media — anyone with whom to discuss the accident who possessed appreciable knowledge of and/or interest in its details.

In the middle of the Saturday afternoon one week following the mishap, while working a volunteer job at the Dallas Zoo, I quite suddenly felt dead tired. I concluded that this was a result of having been under abnormal stress for the previous several days. I am told that psychologists recognize a condition, which goes by various names including "conflict anxiety," in which a person feels that he possesses important knowledge but is powerless to act on it. It occurs, for instance, when an employee has a good idea for doing his job better, but does not have channels available for passing this on to higher-ups in positions to implement it. I now found myself in an analogous situation as a result of my compulsion to do my own analysis and lack of any real connection with the investigation.

On February 26 I received a second phone call from JSC. My first thought was that this must be the call of my previous imagining to verify information. Instead I was asked if I would agree to sell my camera. It turned out that I was talking to some very nice people in JSC's procurement office. I liked the camera, venerable as it was, and did not particularly want to let it go. I offered to loan it to them. They said that the investigators were going to be taking it apart and probably would not be able to put it back together again. They offered a fair price, and added an amount for shipping sufficiently generous that in the end it covered all expenses I incurred associated with the investigation.

A third phone call on February 27 and some email exchanges got the details set up. I shipped off the family's Canon ES50 to whatever fate awaited it. That evening I cursed and stomped around the house when UPS's Web site failed to provide me with timely tracking information for the shipment. I needn’t have made such a fuss. The camera did make it, and I calmed down and went back to devoting most of my leisure time to checking my own analysis and prowling the Internet for any scraps of new information that I could find. A few friends pointed out to me the peculiarity of NASA wanting the camera but not the original of my video. The only answer that I could come up with was some vague speculation that perhaps the VHS copy would somehow give them all the information that they needed.

On March 10 the Columbia Accident Investigation Board, which had been convened to investigate the mishap, released officially, in Microsoft Excel spreadsheet form, what was called “Rev 15 of the Baselined Time Line.” Contained in it there seemed to be some information gleaned from videos that had been submitted by the public, although I could make very little sense of most of it. It did, however, provide increased precision for the time of LOS: 7:59:32 CST.

On March 13, while on a business trip to Baltimore, I saw the time line laid out on the CAIB's “Ground Track and Events Summary” map of that date on the Internet. A map made things very much more clear. The positions and viewing directions for various photographers were marked. Start times for their videos — including mine, now identified by the cataloging number EOC2-4-0018 — were noted (approximately twenty-five videos made between California and Texas documented the accident, although not all of them were represented on this map). I was surprised to see that the start time for my video was specified as 7:59:59 a.m. This was fifteen seconds later than what I had calculated, a significant amount of time considering that Columbia had been covering 3.3 miles every second. My location as represented on the map definitely was wrong. All of this left me confused. I wanted to get it figured out. However, being on the road without my notes and records, there was little that I could do. When meeting schedules allowed I tried to relieve the tension with wanderings about the Waterfront area in the bright sunshine and cold temperatures.

After reviewing my own material following my return home on March 17 I began to suspect that my initial line of site to Columbia was misrepresented on the CAIB's map, and subsequently verified this by noting ground structure in the first few frames of the digitized copy of my video. I twice attempted to contact the CAIB through their Web site without result. A few days later, on March 20, I sent an email to the address I originally had used on February 1 to report the video. At this early stage I was hedging my bets a little, claiming less than a fifteen second error ("AOS" here refers to "acquisition of sighting;" here and in all subsequent email and CAIB public hearing transcriptions I have indicated locations where I have edited out text from the copies of those transcriptions that I possess with an ellipsis):

To: columbiaimages@nasa.gov...

I am the individual who made the Columbia re-entry video from Duncanville, Texas, numbered by NASA as EOC2-4-0018. Information from the video has now been placed in the public record as part of the Accident Investigation Summary Timeline, Rev. 15. Therefore I think that it is appropriate to call attention to some apparent inaccuracies.

The latitude and longitude specifying the location from which I made my video are incorrect. More specifically, the triangle representing my location is placed on the wrong side of Interstate 20. The error introduced is small (two or three miles). However, I included my correct latitude and longitude in a cover letter that went with my video tape to NASA JSC on 3 February, 2003. GPS positioning can easily confirm a particular location to plus or minus a few feet. An error like this would seem to be inconsistent with the otherwise meticulous nature of the investigation (it appears that the investigators may have resorted to some standard latitude and longitude for Duncanville in lieu of using the actual coordinates for my location).

The timeline shows the azimuth component of the video's initial line of sight (AOS) to Columbia lying in a generally southwesterly direction. The investigators seem to have somehow gotten the direction at AOS wrong by some 20 - 25 degrees, resulting in significant timing and location errors. The video's actual initial line of sight is due west. Correcting this would alter the timing of the start of the video (and the events it documents) to about 12 - 13 seconds earlier than what is indicated in the timeline. It also would bring it into very much better agreement with the actual time of day when the video started, an accurate determination of which I took pains to make.

Robert Butsch

Although this produced a reply saying that my email had been forwarded appropriately, there was no other response. I could think of nothing else to try.

On April 4 I received a fourth phone call from JSC, this time from the Earth Sciences and Image Analysis Laboratory, which was involved in much of the actual analysis of videos. They had caught up to the fact that they did not have my original video cassette. They would need it, they said, to do a proper evaluation. I took the opportunity to mention my misgivings about the officially released information. They said that they had indeed received a forward of my March 20 email and had replied, but that in the process the reply must somehow have gotten lost. They confirmed that what had been made public by the CAIB was not completely accurate. In particular, my exact location, like the locations of other photographers, had probably purposely been disguised to protect my privacy. Although the person I was talking with seemed to believe that 7:59:43 CST might be the official start time for my video, followup emails forwarded to me by various people at JSC offered arguments supporting the previously-published suspicious timing and direction conclusions.

From one forwarded message I learned that the investigators had timed my video by synchronizing events captured in it with events seen in a video from a U.S. Army Apache helicopter employing GPS satellite timing:

...Sent: Thursday, April 10, 2003 2:28 PM

...As for the timing issue; we have synchronized Mr. Butch's [sic] video to a second video which has real-time GPS clock information. Based on this information, Mr. Butch's [sic] first view of the vehicle occurred at 13:59:59.0 GMT (7:59:59.0 CST)...

The Image Science and Analysis Group
Human Exploration Science Office (SX3)
NASA-Johnson Space Center

I could see the problem that this was going to produce. How could anyone take seriously times calculated using a very round-about, non-GPS procedure by someone who couldn't even set the right year on his camera display?

Another forwarded email suggested that there was serious uncertainty over Columbia's ground track after LOS:

Sent: Monday, April 14, 2003 9:12 AM To:...

...With regards to Mr. Butsch [sic] azimuth concerns: The flight path shown on the maps are the nominal planned entry path, not the actual flight path. By the time he acquired the vehicle, we had not had a GPS vehicle position, via telemetry, for about 20 seconds. The simple fact is really don't [sic] know where the vehicle was at this point as

FAA radar coverage only extends up to about 70,000 feet and the Shuttle was at 200,000+. It is very possible that if the vehicle was out of control, it was also off its nominal entry track. We have strong indications that it may have been north of the planned flight track by the time Mr. Butsch first saw it. The further North the vehicle was, the closer to due West it would have appeared from his position…

The implication was that this uncertainty could account for our differences over time estimates. The reality of a more northerly ground track seemed highly unlikely based on where I had seen Columbia in the sky, my geographic location (known to great accuracy), the orbiter's altitude (also know to great accuracy), and elementary principles of trigonometry. Although I had re-asserted my 7:59:44 start time claim in an April 11 email, I was reluctant to contest the point, and not just because of my low-tech time calibration and the fact that I am by nature a reserved personality. I assumed that the investigators eventually would catch up to their mistake. These were experts who were involved in a major government investigation in which they presumably would leave no stone unturned. In fact, I was confident that the error would have been discovered even if I had never brought it to anyone's attention, because by now I had come to see, from many viewings of my own video and a couple of others, that careful study of them by themselves provided good clues to accurate times and directions. I won’t go into the boring details here (see Appendix linked to below for those), but suffice to say that it’s not something that was particularly difficult to figure out. But then I had the advantage of being at the actual place where the video had been recorded.

True, all this was not easy to appreciate. Columbia had been traveling several times faster than a rifle bullet thirty-eight miles above the ground, and had never been closer to me than about forty miles line-of-site distance. My camera had not even recorded the orbiter — it would have been almost too small to pick out at that range. Instead it had recorded the surrounding envelope of hot plasma. The pieces seen coming off in the video obviously were large or they would not have been recorded, but even this was misleading since their apparent size also was a function of the size and brightness of their plasma envelopes. In an attempt to get my mind around the the scale of the event I modeled in crude fashion Columbia's track through the sky using some paper cut-outs and a map of North Central Texas. I was amazed to see for the first time just how high, fast, and far away from me the orbiter had been.

I sent off my original 8mm cassette which, another email informed me, went astray after it reached JSC. I never found out if it made it to its intended destination (the fact that there is no public evidence that an EOC — Emergency Operations Center — cataloging number was generated for it suggests that it may not have). Weeks went by and, curiously, no corrected time line was released. I mentioned my concerns to friends and family, but they were not in positions to offer advice on the subject. I also corresponded periodically by email with a space journalist for a major television network. He never sounded convinced of my claims but at least made a good sounding board for some of my frustrations.

At the CAIB's penultimate public hearing on May 6 NASA Flight Director Paul Hill, head of the Columbia Early Sighting Assessment Team, the group handling visual and photographic evidence obtained from sources outside of NASA, made a presentation containing the suspect times and directions. He also implied that work on videos essentially was finished:

...ADM. GEHMAN: Board members? Mr. Hill, what do you think is remaining for your working group to do?

MR. HILL: Primarily processing the last handful of videos to calculate relative motion and good footprints on the remaining western debris and then summarize [sic] everything that we've done...

Not only was the faulty analysis yet to be corrected, it looked as if it was on its way to becoming part of the investigation's final report, due to be written over the ensuing several weeks.

Coming on top of everything else this was too much. I did entertain the idea of just waiting for the errors to turn up in the final report and then pointing them out. I decided, however, given what I had seen up to this point, that the investigation would be even less likely to pay attention to my concerns once the report was signed, sealed and delivered. So I again attempted to contact the CAIB via their Web site, and again got no response. I emailed my original contact at the Earth Sciences and Image Analysis Laboratory (which, for purposes of the investigation, operated under the umbrella of the Early Sighting Assessment Team). She did reply on May 23, pointing out, as she previously had done in our April 4 phone conversation, the curious circumstance that what the CAIB released for the public record was known not necessarily to be accurate, while continuing to specify the starting time for the video as 7:59:59 a.m.:

Thanks for contacting us, as [sic] you point out, the information that Paul Hill presented to the CAIB on May 6 did not include the most recent maps and data. Rest assured that the information at the public hearing is not the official documentation of results. The official report that includes our analysis will not be completed for several more weeks.

Here is the data for your video in the official report:

Latitude 32.67
Longitude 96.90
First view of vehicle 13:59:59 GMT
Last view of vehicle 14:00:53 GMT

You should note that some map represenetations [sic] have shown your view angle toward the PREDICTED trajectory of the orbiter had it been on course. This was done because communication with the vehicle had been lost, so an actual GPS trajectory was not available. By the time you saw the orbiter, it was off the predicted track, so these lines would not be a good representation of what you saw.

I hope this helps with your question.

Earth Sciences and Image Analysis Laboratory
NASA Johnson Space Center (Lockheed Martin Space Operations)...
2400 NASA Road 1…
Houston, TX 77058...

I replied on May 27 and told her that this was just plain wrong. It was an audacious claim which I was reluctant to make, but I had come to the conclusion that, as unfortunately so often is the case in life, shouting was the only way to get anyone's attention. I said that I would put my arguments up on a Web page where I could include adequate detail. This I did on the same day (see link to Appendix below), offering there the suggestion that if they had linked my video to the one with GPS timing using events appearing on both videos that they assumed were identical but actually were not, then this might represent a possible source of error. On May 28 she emailed saying that they were suspicious of some of the arguments made on the Web page, but that they were going to re-do the analysis of the video to try and determine why my numbers represented "such a significant difference" from theirs:

I've passed your reckoning of time on to the group that assigned times to your video--the detail there should be useful for them to sort out why you get such a significant difference. I know there are some errors in your assumptions about how we developed the timing for your video. They will look at it all from scratch, however, and get back with you about the issues you raise. We should contact you by the end of next week, hopefully sooner…

This was a weird situation in which to find oneself. By this time I had looked at the problem from so many different angles that I knew my numbers were correct and that a substantial portion of the investigation’s reconstruction of photographic documentation of Columbia’s breakup was incorrect. However, it had become obvious that I was the only one who did. How could such radically different conclusions be drawn from the most objective of evidence?

One reason may have been the attitude of the investigators. The CAIB was essentially non-responsive. Everyone that I communicated with at NASA itself was civil and polite, careful to thank me many times over for my contribution to the investigation. However, they volunteered few details and seemed always a bit aloof. I was never in a position to make a genuine contribution until I insisted on it.

Admittedly, they did have to sort through hoaxers and would-be profiteers, not to mention some legitimate witnesses with bizarre ideas (an April, 2003, Associated Press story reported on one videographer from the DFW area who was claiming that an obvious lens flare in his video of the accident actually was an image of the Shuttle's outline, even though he could not have been closer to Columbia than about sixty miles). Despite these difficulties, though, the investigators probably would have been better off simply getting on the phone early to all of the private citizens who furnished them with evidence, having a chat, and asking some questions.

The main reason, though, was that photographs and numbers were not sufficient. Objectivity is not accuracy, nor are objectivity and accuracy truth. You had to have been there and witnessed the accident truly to understand what happened. This was an advantage that all of us who got up that morning intent on photographing Columbia had over everyone else. We were eyewitnesses, more so than engineers in front of computer screens at Mission Control, or TV viewers, or casual observers who may have wondered at the smoke trail, or heard the rumbling sonic booms tolling Columbia's death, or noticed hardware fall out of an empty sky onto their driveways, or even chanced to see the orbiter fly overhead not knowing what they were looking at. We knew it was coming and we knew where it went. I knew that it had not passed over me — the implication of the suggested more northerly ground track — because I had watched it fly a track south of me. Curiously, the CAIB never asked a single eyewitness to the accident to testify at any of its meetings or hearings. Keeping eyewitnesses at arm's length from the investigation was not good for either the investigation or the eyewitnesses.

On June 9 the Image Analysis group emailed to say, with a minimum of elaboration, that after re-analysis they had corrected the start time for my video to 7:59:43.5 a.m. CST ("LOS" here refers to "loss of sighting;" times are GMT):

Hi Bob,
We recently did a thorough reanalysis of all the late timeline information and videos (the first since late February 2003). We found the error that you noted in your time accounting. The error occurred because we had linked your video to another video on an event that we thought was unique. We later learned that several similar events occurred during the period.

Our data for your tape are now as follows:

AOS 13:59:43.5
LOS 14:00:37

These agree well (within 1 second) with your own estimates of your video timing based on your cable clock.

All this information will be incorporated into the final Baselined Timeline revision (I'm not sure with this [sic] will be incorporated in Rev 18, or if they will call it Rev 19). Please recognize that the CAIB and NASA don't release every timeline update on their websites. But rest assured that the information being used in the investigation is correct.

Thanks again for your perseverance and patience while we looked through additional data. And thank you for your contribution to the Columbia Investigation...

Evidence (see, e.g., the date for "Entry Debris Events Version 7" in the CAIB report, Appendix E2) suggests that the error was corrected on or before June 4. Fixing the timing automatically resolved the direction error. Although you could never tell it from the email, all of this amounted to a significant change in the investigation's findings regarding Columbia’s breakup, since it required timing and direction information also to be corrected for at least two other videos as well as for a number of late accident events. Interestingly, the initial and final viewing directions for the Apache video were revised at about this time (see p 103 of Volume 3 of the CAIB report vs. p 197 of Volume 3).

I did shout a silent "Yes!" at having been proved correct after much effort. Over the course of the investigation various items of photographic evidence periodically had redirected its focus, but mine had resulted in the last minute alteration of some of its final results. In the end, however, this was my only satisfaction. On the whole the affair had been much less of an adventure and much more of an ordeal than I cared for — not something I would want to go through twice.

Eventually I came to feel as if I literally had been snatched up unawares in some small, peripheral eddy of Columbia's wake as it hurtled by to its doom. My video, like all the others, is little more than obscure documentation of remote death and destruction, its creation as much accident as planning, its photographer an innocent bystander. It was not seen publicly until the second anniversary of the accident. The original timing error persists in the only video publicly released by the investigation, a composite that was presented at a CAIB hearing on March 17, 2003 (in October of 2004 I acquired from JSC, via a Freedom of Information Act request, a VHS copy of the final version of this composite, dated June 4, 2003, which incorporates the corrected timing). As a courtesy the Early Sighting Assessment Team sent me a copy of their final report, dated June 13, 2003. It retains all of the erroneous times and directions. The CAIB issued its final report on August 26, 2003, and supplemental material on October 28. The corrected information can be found in Appendices D9 and E2 while the erroneous information lurks in Appendix E4 (see figures in Appendix linked to below). Although it went to the trouble to note their occurrence, the CAIB otherwise was not interested in events spanning the period from LOS to when debris started to hit the ground, except for a circumspect curiosity about how the astronauts died.

I used to occasionally see videos of the accident made by others. They were both sad and interesting — reminiscent of my own. Mine, however, whenever I saw it after having been away from it for a while, never failed to provoke an emotional response that is difficult to describe.

The recording that I made on that day also has provided what for me will remain my most vivid recollection of an often surreal experience. When Columbia reaches a position south by west it is a brilliant, speeding oval trailing a many-hued incandescence and grey smoke, a sort of incredible Roman candle. But against the featureless blue sky the image is cold and sterile, possessing nothing of the poignancy that human tragedy should evoke. Then the orbiter passes behind a few bare branches of a tree, symbols of a precious life and Earth that it and its crew were destined never again to touch. The piece that I initially had observed coming off still can be seen tumbling forlornly toward the ground. Four prominent spikes of light stab out from the oval, created by the camera's optics and indicating an extraordinary brightness. This phenomenon is uncommon to see during the day: one must be photographing the sun, perhaps, or a cutting torch, or the birth — or death — of a rocket flight. Columbia at its end was a new star.