Crash Location - A revisit using OSCAR & Quikscat data
This is a revisit of the data I originally presented on page 210, and something between the original and that presented below may be close to the truth.
After analysing OSCAR/NOAA surface current data for the period 01 June - 15 June 2009, and the Quikscat satellite surface (+10m) winds for the same period for the area encompassed by 5°30'N 32°30'W, 5°30'N 29°30'W, 2°30'N 29°30'W and 2°30'N 32°30'W, I have come to the conclusion some of the "observed data" bears no relation to the "actual data" resulting from the location of bodies and other debris from AF447.
The basic problem is that the North Atlantic Equatorial Current had developed a "blip" in the area of the suspected crash location. This area was in fact the location of the Doldrums at the time of the accident and the current was attempting to flow Northeast to become the Equatorial Counter Current. However, at 3°30'N 30°30'W a very strong pull to the West was developing and the surface current velocities over the period from 07 June to 17 June increased to average 9.35NM per day or 20cm/sec along the general line in which the bodies were recovered. This is nearly twice the rate provided by the satellite data. In the meantime part of the current continued from 3°30'N 30°30'W in a Northeast direction at velocities approaching 1.5 times the published figures.
The anomolies associated with the OSCAR data are probably due to a slightly lower than expected Mean Sea Level barometric pressure in the pivotal area to the West of 3°30'N and 30°30'W, resulting in the actual MSL being slightly higher than the one used in the OSCAR/NOAA calculation of the current vectors and velocities.
The Northeasterly branch of the current from 3°30'N 30°30'W carried very little debris with it - the most notable (that we know of) being the Starboard Outer Spoiler which was recovered on 13 June NNE of TASIL. Looking closely at the photos of the recovered spoiler, I sense that the damage which resulted in its separation was caused by the high velocity vertical impact of the underside of the wing with the sea surface as the wing was rotating clockwise (backwards) horizontally. The clockwise vertical rotating force exerted by the Vertical Stabilizer as the tail entered only added to the forces experienced by the spoiler.
The heading of the a/c at the time of impact was probably at some point in the East quadrant, which resulted in the spoiler departing "explosively" in the same general direction. Once the general disturbance caused to the sea by the impact had subsided, the spoiler was likely located some distance east of the remaining debris. Hence it later missed the anti-clockwise turn near 3°30'N 30°30'W and continued with the NE branch of the surface current, and was also unaffected by the surface wind due to its very low surface profile as noted in recovery photos.
In the graphic below, overlay of data is on the 06 - 18 June 2009 composite showing "positions of bodies and debris recovered" produced on page 37 of the BEA's Preliminary Report (French version) into the accident. You should note that this graphic in the body of the report differs from that shown in the Appendix, and when comparing data attributed to specific days, positions do not always match up. The pressure on the BEA to complete their initial analysis and compile the data in a limited time may well be behind these discrepancies.
Knowing the approximate position of the Vertical Stabilizer when it was recovered on 07 June, and that of debris and bodies recovered on 06 and 07 June were fairly well grouped, has helped tighten down the analysis required to back-track to the possible impact point. South of 3°30'N the current vectors and velocities provided by OSCAR seem to be fairly constant, and provided those velocities are reasonably accurate a recalculated position of the impact point is shown.
The back-track involved equating the distance and bearing that the Vertical Stabilizer was from the general track line of bodies (dashed magenta line) with the mean surface current along with wind vectors and velocities from 0300z on 01 June to 1500z on 07 June 2009 (6.5 days). A light blue (cyan) line marks the general line along which debris with a low profile to windage has been recovered. The total leeway vector and distance of the v/s is 255.3°T x 9.38NM from the mean bodies track line, and the daily factor derived from Quikscat data (modified to 2m above sea level [velocity x 0.75]) has been used along with the appropriate current vector and distance in the back-track exercise. The data used is replicated below.
Referring back to the top graphic, the green (lime) line represents the NE'ly branch of the current, and the area between the light blue (cyan) and red lines is the area in which the majority of bodies and debris were located and recovered. Debris to the West of the red line has been subject to significant windage and consequent leeway. One body was located close to the red line, and the assumption is that it was caught up in other debris. Three debris items near 3°15'N 29°45'W do not appear in the Appendix version and possibly are unrelated to AF447. An item marked "07" near the "last known position", has probably broken lose at depth from the wreckage and made its way to the surface, but how long it was there before being recovered on 07 June is another matter. Water spouts are common in the ITCZ, and odd items of debris may have been moved about through being lifted and later dumped some distance away. The current vectors shown are overlaid from the OSCAR data, and it is fairly obvious that the general drift of the bodies and debris hasn't always followed the expected line.
The accuracy of the information presented is of course reliant on data from a number of sources, but is the best that could be obtained.
mm43