So how did he manage to work out where we landed and at what time? On both those flights? (I’ll have to take your word for it on the call sign, too long ago now to remember).
That was the point of the flights. There was limited ADSB data available and in parts of them we did NOT follow published air routes.
Given that one of them over flew Sydney, how did he know we didn’t land there?
Any thoughts on the blind test of the RNZAF P3 flight? Details of that aren’t on ADSB or flight tracking websites either and yet the tracking is pretty bloody close to what they flew.
I get being sceptical but this appears to have some legs.
Keg, I can't comment on what I don't know about, specifically the second flight and the latter stages of the QF6036 APW-ADL flight. What I can say about the QF6036 flight is that his methodology never produced an accurate location fix (that is, within 30km) and after 2 hours the error was 70km.
Regards overflying Sydney, he knew that, '
Fuel at takeoff was 48.0 MT of which 4.5 MT was added by the Captain.' That's a lot for the 6.2 hours to Sydney, isn't it? More like what you would need for the nearly 8 hours to Adelaide.
Regards the RNZAF P3 flight, that wasn't a blind test. He had the track data and then mapped it to his WSPRnet links. I'll comment on the sheer number of links that the author proposes shortly but what is notable in the case of the P3 flight is the base rate of anomalous links - nearing 90 percent of the links are anomalous (either frequency shift or SNR shift). According to the author, that indicates that they are being perturbed by something, likely aircraft.
The issue here is that, even if that contention is correct (and it likely is not), the perturbations can't all be down to the target P3. It's a simple statistical matter - put sufficient lines on a map and plot a track in 2 minute increments on the map and at some point the track plot will land on a line. If 90 percent of the lines are anomalous, there's a 9 in 10 chance that you get to say, that anomaly was caused by the target aircraft.
For fear of going on a bit, the history of this "breakthrough technology" is instructive. Using WSPR, that is a weak signal radio connection between two stations, to detect the presence of an aircraft and applying that to MH370 was first proposed Rob Westphal late last year. His work focussed on the short path (that is the most direct great circle propagation path) between the two stations. People were sceptical if WSPR data could be used that way but the notion of relying on the short path was uncontroversial.
The problem that then arose for the current author's work was that there are very few short path links that traverse the Southern Indian Ocean. Undeterred, the author invoked the long path for his work.
As illustrated above the long path is what's left of the entire great circle when the short path is subtracted. Invoking the long paths provides myriad paths that criss-cross the SIO, the problems being that (a) most of them do not exist (long path connections over the distances invoked by the author are rare) and (b) as there is only one limited data set per each WSPR "spot" (Tx-Rx connection) any perturbation of the long path would be unreadable under the short path.
The sheer nonsense of the author's approach is that he claims both sensitivity and specificity. He variously claims that not only an aircraft on the other side of the world can be detected perturbing the long path transmission but that it can also detect "the wake vortex and a series of small Doppler shifts due to water vapour variations". You might well ask yourself how many "water vapour variations" might impinge on a radio signal that is purportedly travelling 40,000 km around the world!
If that's not bad enough, the author then claims that with that extraordinary sensitivity the technology can discern a specific target on the other side of the world. The fact that many of the Tx Rx stations are near airports doesn't seem to be a concern - potential contamination of the data by other than the target is never mentioned. No one with any understanding of physics, radio and/or radar that I have spoken to thinks that what the author proposes is even remotely possible.
But what the use of long paths does for the author is that it baths areas of interest in path lines that he can use. For the QF6036 example there were at times a long path line on average every 5 km. It's then a simple statistical matter - chance of intersecting a path line X chance of path line being anomalous. The author is essentially invoking order out of sheer randomness.