From the book "Incident at Sakhalin" by Michel Brun.


An excerpt which casts enormous doubt on the authenticity of the Flight Data tape presented to the ICAO as "proof" that the Korean plane KAL007 was "shotdown" by the Russians.
_ ~
The tape decoded by the Accident Investigation Bureau of the Flight Test Center in Breteigny-sur-Orge contained 27 hours of recorded material broken down as follows: a 5.5 hour segment for the last flight until the destruction of the aircraft; two complete flights, said to be a round-trip flight from Anchorage to New York and back to Anchorage for a total of 15 hours; and a final 6.5 hour flight supposedly prior to landing at Anchorage. This makes a total of 27 hours of recorded material on a device that can physically contain no more than 24 hours and 48 minutes of data. Clearly something is wrong. Because the tape speed is regulated with extreme accuracy to enable the reconstruction of the flight to better than one-second precision, we know that the tape was 852 inches too long. Was the tape originally longer than the tape on KAL 007, or was the aditional footage added after the fact? In either case the tape could not have come from the DFDR on KAL 007.

Number of Parameters

The number of parameters appearing on the DFDR tape examined by the ICAO does not correspond to the DFDR that was on board KAL007. In particular, page 29 of the ICAO report states that, according to Korean Air Lines, the following parameters, among others, were recorded by the Korean airliner's DFDR:

None of these parameters appear on the ICAO's DFDR. It provides engine-pressure ratios for only two engines instead of four. This implies that the DFDR used by the ICAO is from a two-engined aircraft and not from a Boeing 747, which is a four-engined aircraft. The absence of universal time is also significant. It allows plenty of latitude for synchronizing the DFDR with the most convenient time scheme available and makes it relatively easy to create false data.

Performance and Maneuvering

The aircraft performance and flight maneuvers recorded by the DFDR are not consistent with the behavior of KAL 007. Page 26 of the report details the takeoff parameters at "Anchorage." I have put the word in quotes because nothing in the data tells us at which airport the aircraft carrying the DFDR landed prior to this takeoff. However, the landing sequence shows it was not Anchorage International Airport (though is may have been Elmendorf Air Force Base, which is also at Anchorage.) I have compared the data from the DFDR concerning the aircraft's landing with a detailed map of the Anchorage International Airport in 1983. They do not match. Following the headings indicated by the DFDR, on its way to the gate after exiting its landing runway, the aircraft would have had to pass through a building.

It is also instructive to compare the data we have on the DFDR aircraft's takeoff, alledgedly from Anchorage International, with the data we have on KAL 007's actual takeoff from that field. We can reconstruct the DFDR aircraft's actions from the DFDR data on the graph of page 26 of the ICAO report. "Chart 6. Flight KAL 007. Takeoff at Achorage." At the same time, we know in detail how KAL 007 maneuvered at Anchorage because we have a record of the aircraft's communications with the control tower.

Comparison of Takeoff Maneuvers
DFDR aircraft KAL 007
Begins to taxi without
authorization 5 minutes 30
seconds before takeoff.

Taxis for 4 minutes 45 seconds

Fails to contact the control
tower and does not stop at the
entrance to the runway.

Begins to takeoff without
authorization.

Rolls for 52 seconds before
taking off.
Requests authorization to taxi
4 minutes 53 seconds prior to
takeoff.

Taxis for 2 minutes 54 seconds.

Stops at the entrance to the
runway and announces,
"Ready for takeoff."

Begins takeoff after receiving
authorization at 12:58:13 GMT.

Rolls for 1 minute 47 seconds
prior to takeoff, which takes
place at 13:00.
The above comparison shows that the data are from two different aircraft. The DFDR aircraft turned onto the runway without prior authorization (there is no trace on the DFDR of any communication before initiation the aircraft's runway operations) and rolled for only fifty-two seconds prior to takeoff, which indicates that the aircraft had significant reserve power. The other aircraft, KAL 007, always requested permission before any ground operation, as shown by its communications with the control tower on pages 17 to 19 of Information Paper No. 1. It rolled or one minute forty-seven seconds prior to takeoff, confirming ahat the aircraft was heavily loaded. The DFDR data indicate that the aircraft took off at a speed of 175 knots, which is high for a Boeing 747, whose takeoff speed is normally between 130-140 knots.

Back to main page Go back to "Keppel and Brun on KAL007
No wreckage, no crash?  Another interesting page on the Korean Air Disaster 1983

Wreckage Investigation
_ more later... maybe