The following correspondence took place between Elaine Scarry, author of “The Fall of TWA 800: The Possibility of Electromagnetic Interference,” published in the April 9, 1998, issue, and James E. Hall, Chair of the National Transportation Safety Board.

March 13, 1998
James E. Hall, Chair

National Transportation Safety Board
490 L’Enfant Plaza East, S.W.
Washington, D.C. 20594

Dear Chairman Hall:

The inquiry into the fall of TWA 800 on July 17, 1996, has centered on three areas: a mechanical accident, a bomb, and a missile or other high-velocity material object (such as a meteorite). Arriving in your office tomorrow morning is an article in The New York Review of Books that will reach the public next week. It describes a fourth possibility: electromagnetic interference from one or more of the ten military craft that were in the area. The article asks that this matter be taken up by the NTSB.

Prior to its publication, the article was read by many experts: electrical engineers, physicists, specialists in avionics and aerodynamics, pilots, several people in the Navy or Air Force. Among them are people with specific expertise in either electronic warfare or electronic interference.

The article is divided into three parts. Part 1 gives a portrait of High Intensity Radiated Field events and shows the efforts made by the military to protect their own craft against inadvertent radio interference from other military craft. Military planes spend more time in the company of other military craft than do civilian planes, and are therefore more likely to be subjected to a High Intensity Radiated Field event. But civilian planes do sometimes (as in the case of TWA 800) end up in the vicinity of military planes or ships; they may, therefore, sometimes be placed at risk. A NASA report that determines occurrence rates of HIRF accidents is included, as are other materials (an FAA document, a statement by Raytheon’s president) to show why inquiries into the unsolved crashes of civilian planes need to look at this possible source.

Part 1 of the article urges that military studies of electromagnetic interference—such as the seven-month study by the Air Force in 1988 and the three-year Pentagon study in 1989-1991—be made available to the NTSB for its continuing inquiry into TWA 800. This request is compatible with your own insistence during the Public Meeting in Baltimore, December 8-12, 1997, that military and civilian worlds need to share research about aviation safety. You are quoted at length on this key matter on page 63 of the article.

Part 2 of the article moves from the background to the foreground. Ten military craft were in the vicinity of TWA 800 at the time it began its fall: one P3 Orion, one C-130, one Black Hawk helicopter, one C-141, one C-10, the Coast Guard cutter Adak, the USS Wyoming, the USS Albuquerque, the USS Trepany, and the USS Normandy. The section examines five of these craft in order to provide a sample of the questions that need to be asked of them all. It specifies the kinds of sophisticated electromagnetic equipment included on such craft. It also indicates the many pieces of information about these craft that have not yet been made available either to the public or to the NTSB. The article urges that a detailed picture of the electronic equipment that was turned on and turned off that night be provided to the NTSB. It also argues that the airmen and sailors themselves (not one of the Chiefs of Staff or the Commander of the Second Fleet) are the people who can best provide the concrete information needed for a minute-by-minute reconstruction.

Part 3 of the article brings the background research on HIRFs together with information known about the fall of TWA 800 to suggest three possible ways in which electromagnetic interference might have played a part in the crash. The scenarios represent just three of many possible paths that a HIRF accident conceivably could have taken. The section draws on the research completed by the NTSB.

Also included in Part 3 are statements on the need for a public inquiry. The statements are made by Rear Admiral (retired) Eugene J. Carroll, Jr., Deputy Director of the Center for Defense Information; physicist and electrical engineer D.V. Giri, a former research associate at Phillips Laboratory, Kirtland Air Force Base, and a designer of both antennas and filters for microwave systems; Martin Shooman, the author of the NASA study on occurrence rates of HIRFs as well as other studies of aircraft and satellites and a book on Probabilistic Reliability; a physicist deeply knowledgeable about electronics; and David Wunsch, electrical engineering professor at University of Massachusetts-Lowell, who for many years has researched radios and antennas. These people (as well as others whose written research is cited in the article) might be useful participants in any NTSB inquiry into electromagnetic interference.

I am unaware of any findings by the NTSB that would rule out the possibility that a High Intensity Radiated Field event played a part in the crash of TWA 800. Since April 1997, when I began the study, all information released to the public by your agency has appeared to me either neutral or positive with respect to the thesis that electromagnetic interference from an external source may have played a part. This includes, for example, your emphasis, beginning in July 1997, on the need to find an electrical source for the ignition of the central fuel tank, and the revelation that it has so far been difficult to find any conclusive source inside the plane itself.


The article stresses the fact that the inquiry into the possibility of a HIRF accident needs to be carried out by the National Transportation Safety Board (and not, for example, any other government body such as the FBI or CIA). The NTSB is the duly constituted agency for such an inquiry and represents not just those who have been directly harmed by the accident but the citizenry as a whole.

Let me add to what the article says by mentioning how impressed I was by your respect for citizens—for their obligation to keep themselves informed about such an event—throughout the December 8-12 public meeting. I refer in part to your insistence that civilian and military realms must more fully work together to ensure everyone’s safety. But I also refer to your continual awareness of what someone looking at one of your graphs or listening to one of your experts might not yet understand. No sooner did a question arise in my mind (“What does ’30 knot target’ mean?”) than I would hear your voice interrupting the proceedings to ask, “What does ’30 knot target’ mean?”

I hope the NTSB’s inquiry will continue to give voice to matters that are of wide concern. Should you have any response to the article, I would be most eager to learn of it, as would Robert Silvers of The New York Review, which would welcome a response from you.

Elaine Scarry
Cambridge, Massachusetts

April 21, 1998

James E Hall replies:

Dear Ms. Scarry:

Thank you for your March 13, 1998, letter regarding the National Transportation Safety Board’s investigation of the accident involving TWA flight 800. In your letter, you provided a summary of your article in The New York Review of Books regarding the possibility that electromagnetic interference (EMI) or High Intensity Radiated Fields (HIRF) could have caused the accident.

I have read your article and found it quite interesting. The Safety Board’s investigative team has considered the possibility that EMI or HIRF was a factor in the accident and has been actively pursuing the events that could have led to the ignition of the fuel/air mixture in the center fuel tank. The investigative team is working with private contractors and the military to determine the effects of EMI or HIRF on Boeing 747s.

If EMI or HIRF had ignited the fuel/air vapor, other systems on the TWA flight 800 airplane would have also been affected to some degree. The examination of the flight 800 flight data recorder and the cockpit voice recorder did not indicate any unusual signals before the end of the data. Additionally, it would be unusual if only one airplane operating in the area was affected by EMI or HIRF. The flightcrews of other airplanes operating nearby did not report any problems with their airplanes. These are just a few of the issues that will be addressed as we continue our investigation. However, please be assured that the Safety Board is considering every possible event that could have led to this accident, including EMI or HIRF.

Once again, thank you for sending me a copy of your article.

Jim Hall

April 30, 1998

Dear Chairman Hall:

I write to thank you for your letter of April 21, responding to my article in The New York Review of Books and my March 13 letter to you. I am glad to learn that the problem of High Intensity Radiated Fields (or High Intensity Radio Frequency as it is alternatively called) is a subject included in the TWA 800 inquiry.

In your letter you name two features of the crash that you believe diminish the probability that HIRF was the cause. I hope before long to write you a letter about these two matters: the first, the question of whether other planes in the area reported any signs of electromagnetic interference; the second, the question of whether the voice and data recordings on TWA 800 contain any sign of interference.

For the present I wish simply to express my gratitude for your attention to the subject of my letter and article. I also have one pressing question. You state, “The investigative team is working with private contractors and the military to determine the effects of EMI or HIRF on Boeing 747s.” Is it possible to give an idea of the time period you anticipate will be involved in this research? Is it (for example) more likely to be several weeks or instead a year-long project? If you have even a very rough idea about the timing, I would be grateful to know it.


Sincerely yours,
Elaine Scarry

May 27, 1998

Dear Ms. Scarry:

Thank you for your April 30, 1998, letter regarding the National Transportation Safety Board’s ongoing investigation of the accident involving TWA flight 800. In your letter, you ask about the time that will be required for research into the possible effects of High Intensity Radiated Fields (HIRF).

The Safety Board has several energy-related activities in progress. The examination of energy induced by one aircraft system into other aircraft systems began in 1997, and more than half of the planned tests have been completed. Work is also underway to examine threats that may have resulted from an external surface or airborne HIRF source. This is a challenging undertaking that involves an understanding of military emitters, as well as a considerable number of civilian transmitters near the accident site.

We hope to complete these activities within the next six months, but we must evaluate test results and refine our test plans as the investigation continues. Thank you for your interest in the investigation.

Jim Hall

June 17, 1998

Dear Chairman Hall:

Thank you for your letter of May 27, 1998, about the inclusion of research on High Intensity Radiated Fields in the TWA 800 investigation. I appreciated your informing me that, in addition to tests begun in 1997 on the effect of one airplane system on other systems inside the same plane, “work is also underway to examine threats that may have resulted from an external surface or airborne HIRF source.” I also appreciated learning that the inquiry, estimated to take six months, will include attention to both military emitters and civilian transmitters in the vicinity of the accident. (In the postscript below, I comment on the points you raised in your letter of April 21.)

Several distinguished scientists—upon learning of the Safety Board’s inclusion of the subject of HIRFs in its TWA 800 inquiry—have recently stated to me their own sense of the importance of including the possible effects of military and civilian electromagnetic transmissions in the inquiry (quite apart from whether a High Intensity Radiated Field event is in the end found to be the cause of the plane’s fall).

Among them are Robert G. Shulman, Sterling Professor of Molecular Biophysics and Biochemistry at Yale University; Robert L. Walker, Professor of Physics (Emeritus) at Caltech; Melissa Franklin, Professor of Physics at Harvard University; Richard Wilson, Mallinckrodt Professor of Physics at Harvard University; and Theodore B. Taylor, Visiting Fellow, Center for Energy and Environmental Studies at Princeton University.

Similar comments have been made by the experts I cited in the New York Review article on the fall of TWA 800, including the electrical engineer Martin Shooman (author of the 1994 NASA study of HIRFs), the electrical engineer and physicist D.V. Giri (designer of microwave antennas and filters), the electrical engineering professor A. David Wunsch (researcher of radios and antennas), and Rear Admiral (retired) Eugene J. Carroll, Jr. (Deputy Director, Center for Defense Information, and former commander of a naval carrier group). When I sent them copies of your April 21 and May 27 letters, they all expressed to me appreciation of the Safety Board’s ongoing inquiry, and look forward to being informed about your findings as your research continues into the nature of military and civilian transmissions in the vicinity.

I admire the thoroughness of the NTSB, and trust that you will be making every effort to discover the cause of the fall of TWA 800 (whatever that cause turns out to be). I will follow with interest the branches of your inquiry that are reported in the press; and will hope to hear from the NTSB about that part of the research devoted to HIRFs.

Sincerely yours,
Elaine Scarry


Since the Safety Board is now proceeding with the inquiry into HIRFs from military and civilian transmissions, it may be that the two factors you cite in your April 21 letter as discouraging this line of inquiry are no longer perceived as impediments. It may also be that in the weeks that have passed, the Safety Board has discovered additional information that now makes irrelevant any counterarguments by me or others outside the formal inquiry.

However, as I promised in my April 30 letter, I offer the following comments on the two points.


You state that the voice recorder contains no evidence of EMI or HIRF. Even if this were true, it would of course not be a reason to abstain from investigating EMI or HIRF. At its public meeting in December, the NTSB stated its judgment that the voice recorders contained no clue to the disaster. Despite the absence of a reference on the voice recorder to mechanical trouble, to a bomb, or to a missile, these three possibilities have been rigorously investigated. A High Intensity Radiated Field event should be given full scrutiny as well, even if absent on the voice recorder.

But is it absent? My essay in The New York Review acknowledges that in the judgment of the NTSB and many others, the voice recorders contain no explicit record of any kind of difficulty; and it acknowledges as well that this judgment may well be right. But I also call attention to the fact that the two small signs of possible trouble—a comment by the pilot about a “crazy” fuel gauge and a small complaint by the pilot about “trimming” the plane—happen to overlap with the very two events that the head of the 1989 Air Force study named as classic HIRF-related events, interruption in the fuel flow and false instruction to control surfaces.

The complaints are themselves so brief and so low-key (and so compatible with a smooth flight, had that been what followed) that they could not in isolation be taken as calling for an examination of the possibility of a HIRF event. But they do not occur in isolation. First, they occur within one minute and fifty seconds of a catastrophe. Second, the catastrophe, after almost two years, has been assigned no definitive cause. Third, what cause there is has long seemed to NTSB investigators to have something to do with electricity. Fourth, the environment was not electromagnetically neutral: there were not only many civilian transmitters in the area, as you recently point out, but many military craft. (My article follows Assistant FBI Director Jim Kallstrom in citing the figure of ten; but I enclose a Navy document that appears to specify a second P3 in the area; if this document is accurate, the number should be corrected to eleven.)

I did not say in the article, nor do I now, that the investigation should be undertaken on the basis of the pilot’s statements (most parts of the article were written long before the Safety Board released the voice recording to the public). I only suggest that if a High Intensity Radiated Field event is investigated, and if it is one day found to have been the cause, then we may listen to the tape and hear in it an acuity of observation that is inaudible to us now.

You write that “the cockpit voice recorder did not indicate any unusual signals.” But the possibility I raised that there may conceivably be recorded there the opening phase of what two minutes later became a severe HIRF event is compatible with the NTSB’s own Systems Group Factual Report. That report describes how a fuel flow gauge works and lists electromagnetic interference (EMI) among the causes that can produce the kind of anomaly the pilot saw:

Fuel to each engine is passed through a fuel flow transmitter that converts the fuel flow rate to electrical pulse signals. These pulse signals are processed in the electronic module and transmitted to the indicators. An interruption or change to these signals caused by component failure, EMI, or wire faults including but not limited to open shields…. (Docket No. A-516, Exhibit 9-A, p. 132)

The Systems Group Report moves away from electrical to mechanical explanations (the “sticky” fuel gauge of the summary given at the Public Meeting) but it observes that its examination of electrical components “was limited to detecting failures which manifest themselves with direct physical evidence such as burning, arcing, or excessive heat”—physical evidence that is often not present in EMI.



The second point you make in your April 21 letter concerns planes flying in the area of the accident: “it would be unusual if only one airplane operating in the area was affected by EMI or HIRF. The flightcrews of other airplanes operating nearby did not report any problems with their airplanes.”

Assessments about what would be “usual” or “unusual” seem difficult to make, since actual pilot-reported HIRF events (electromagnetic interference from sources outside the plane) have in the civilian sector rarely been studied and since electromagnetic interference (whether from sources inside or outside the plane) tends to be underreported by pilots. Martin Shooman, the author of the NASA study, notes that EMI events often fall on the edges of a spectrum: either they are too minor to report (a momentary fluctuation in a cockpit instrument) or too major to report (the plane suffers a catastrophe with no surviving witnesses).

In what follows I address both sentences in your statement, first by arguing that it would not be unusual if only one airplane operating in an area were affected by HIRF, and second by suggesting that at least one other plane operating in the area may have actually reported events deserving of further scrutiny.


Almost all High Intensity Radiated Field incidents now in the public record have affected only a single plane. The 1994 NASA study—which appears to be the only existing field study of the effect of HIRFs on civilian planes—is exclusively made up of reports involving solitary planes.

It may well be that if we could reconstruct what planes were nearby one of the NASA study cases, we might—by consulting the other pilots—discover that some anomalies did in fact occur that affected more than one plane and simply went unreported. (So, too, we might find, if we consulted the crews of the planes near TWA 800, that they actually did experience some small difficulties which they may have considered too minor to report, especially with so stark a tragedy occurring to a fellow pilot nearby. Indeed they might even have assumed that any difficulties they were having with electrical equipment or with radio transmissions were caused not by a HIRF event but by the catastrophe already underway on the fellow plane.)

While, then, it is imaginable that we may someday come to see effects on more than one plane as the norm, the present norm suggests just the opposite: what has so far been documented for civilian aircraft is a string of incidents involving solitary planes.1 The experimental literature on electromagnetic interference reinforces this picture since it repeatedly speaks of how anomalous EMI events are. A given event can be hard to duplicate at will precisely because all the conditions—location, speed, wiring configuration, shielding, such vulnerable conditions as a corroded fastener or exposed wire—must be so exactly reproduced.

The military record—or that part of the military record that is now open to the public—also portrays solitary craft as suffering HIRF events. The Black Hawks that fell in the l980s each appear to have fallen alone. I am not familiar with any report of a second, nearby craft experiencing even small difficulty. The one publicly recorded instance involving multiple planes is the 1986 mission over Libya in which six Air Force F-111s and two Navy A6s were affected. But this was an instance of out-and-out electronic warfare, involving a blizzard of emissions—a level unlikely to be released near Long Island where TWA 800 fell. (For that matter, even the literature on electronic warfare seems to suggest that directionality and precision of targeting are among the goals of successive generations of Radio Frequency [RF] weapons. See, for example, the February 25, 1998, testimony before Congress on Radio Frequency weapons that says they “leave no signature” and have the capacity to single out a specific target.)

My article in The New York Review urged that the 1988 Air Force study and 1989- 1991 Pentagon study be made accessible to the NTSB in part so that an agency responsible for civilian safety could get an accurate picture of what “the norm” is. Perhaps these studies have now been released to the Safety Board; perhaps their findings form the basis of your statement that “it would be unusual if only one plane operating in the area were affected by EMI or HIRF.”

Would it not be possible to tell the public the basis of your statement (or to give a general description of the evidence you rely on if you are not at liberty to describe it in more detail)?

I share your sense that a HIRF event that fatally and promptly harmed one plane should somehow be independently noticed by other planes. But it seemed worth pointing out that in the past, events have taken place involving a solitary plane.


The NTSB graphs of the area during the seconds immediately before and after TWA 800 began its fall show that the nearest planes are USAir 217 and the Navy P3.

Navy records of the P3 have been made available to citizens who have filed under the Freedom of Information Act (one of whom contacted me through The New York Review and forwarded to me a copy of released materials). Most sections of most pages have been deleted because, as the Navy letter accompanying the documents states, they contain information about “mission capabilities,” about “the movement of significant Naval units,” about “Navy assets,” or because they name individual crew members.

But some information survives. It indicates that when the P3 crew returned to base, they filed multiple repair reports, some possibly entailing electrical difficulties: “Copilots turn needle inop[erable]”; “HF 1 failed to couple entire flight”; “#3 shp gauge failed intermittently throughout flight”; “#3 autofeather will not arm”; “Copilot master ICS box, no UHF 1 or 2 lights”; “HF amp causes constant transmitter off light”; “ECA #12 card 12 causes bearing #2 needle to be off 30 degrees”; “I[dentification] F[riend] F[oe] transponder bad, no mode C”; M[agnetic] A[nomaly] D[etector] R-32 black pen is inop fuse checked good.”

In some cases the mechanics report they are unable to duplicate the reported problem on the ground. In response to the pilot’s report “Mode C inop entire flight,” a mechanic writes “could not duplicate on the deck. Sys checked 4.0 on ground test.” In response to the pilot’s report, “#2 powerlever 36 degree switch inop, will not autoshift with powerlevers forward of 36 degrees,” the mechanic responds, “could not duplicate gripe, RPM switch checks good on low power turns.”

The crew also reported difficulty with the control surfaces: “Rudder trim has to be turned several turns before it engaged and trimmed the aircraft. Happened several times throughout flight.” The pilot making the report speculates that the cause is mechanical: “Rudder trim wheel appears to have slipped inflight when trimming to the right.” But the mechanics find that they cannot duplicate the problem on the ground: “Performed Op[erational] check of rudder trim tab system…. Op[erational] check 4.0, could not duplicate gripe.”

When you say that no flightcrew reported any difficulties, is it that the Safety Board has studied the P3 materials and has decided that they do not constitute HIRF effects? Or is it that the Safety Board has not yet seen the P3 materials? Similarly, has the crew of USAir 217 been explicitly consulted on the issue of whether they experienced any instrument or control surface anomalies? (USAir 217 barely appears in the FAA voice record of conversations between nearby pilots and the Boston Air Controller; hence its omission from my account here.)

When you raise the question about reports made by the flightcrews of nearby planes, I assume you wish to direct our attention to whether there is any evidence that may reinforce, or instead conspicuously fail to reinforce, the possibility that a HIRF event took place. I want to close by asking you about a detail in the NTSB’s weather report (a detail called to my attention by a careful reader of all your materials who contacted me through The New York Review). Its meaning—like the meaning of the P3 reports—I am unable to assess.

The east coast sensors that monitor air-to-ground lightning strikes all confirm that no lightning struck anywhere within a 300-mile radius of the accident. But a sensor on Wallops Island that records sky-to-sky events (cloud-to-cloud and intercloud lightning) recorded—according to the NTSB’s Meteorological Factual Report—“an electrical transient about the time of the accident”2 (Docket No. SA-516, Exhibit No. 5-A, p. 5). Would this record from Wallops Island be relevant to an inquiry into HIRFs?

Let me reiterate my appreciation of the Safety Board’s inquiry not only into EMI problems that might have arisen inside the plane but to a HIRF event from civilian or military transmitters outside the fallen plane.

Because the NTSB is now undertaking an inquiry into High Intensity Radiated Field events, I at first felt it might be inappropriate to offer any additional observations such as those presented in this postscript. But it seems to me that it would be equally inappropriate to ignore the problems you pointed out to me in your April 21 letter about the pilot tape, about what is “usual” in a HIRF accident, and about the reports of other pilots.

Thank you for your patient attention to these matters.

This Issue

July 16, 1998