Hacker News Reader

Comment by hencq | Hacker News Reader

> It does mean that the Inmarsat positional analysis about the final location of the plane is likely correct.

Does it though? I'm not saying the analysis is incorrect, just wondering how predictable these ocean currents are. I understand that the debris surfacing at Reunion doesn't rule out the current search location, but that's different than confirming or proving it. I mean, can they just 'play back' the ocean currents and arrive at a specific spot? I wouldn't be surprised if there's so much uncertainty that all they can say with certainty is that it must have crashed somewhere in the Indian Ocean. Which isn't exactly new information.

View on HN

The Inmarsat data seems pretty solid, but it's not very precise. There was always the possibility that the plane actually went down somewhere on the northern arc, rather than the southern arc as was assumed. The northern arc was never very likely because the plane would almost certainly have been noticed on that path, but this debris finding rules it out entirely.

(In case anyone is unfamiliar with just what the data showed, here's a quick summary image: http://ogleearth.com/wp-content/uploads/2014/03/MH370_Mar17....)

No, it was actually possible to exclude the northern arc based on the satellite data only:

http://www.airtrafficmanagement.net/2014/06/mh370-what-does-...

"the Doppler shift due to the satellite’s motion is not removed by the plane’s terminal when it pre-corrects the frequency of transmission, and thus there is an asymmetry between the BFO which would be produced by a plane flying north of the equator (and thus situated to the north of the satellite) compared to one flying south on a similar track."

Nice! Never heard that stuff. I stand corrected.

It's a perfect example for "science, it works." They managed to conclude quite a lot from a miniscule amount of data with the information that was never planned to be used for such purposes.

Sadly, people will attempt to overthink that. E.g. jeffwise.net was cited earlier as a forum discussing AF447. But read this from Jeff concerning the Inmarsat data:

   case 3. The plane did not go south at all.
   If this is the case, then the satellite
   communications system must have been compromised
   by hijackers who either flew the plane north to
   Kazakhstan or China (if only the BFO values were
   spoofed) or somewhere else within a huge circle
   encompassed by the 7th ping ring (if both BTO
   and BFO values were spoofed).

Yes, it's possible that occurred. But even Inmarsat itself didn't have a good understanding of this data. It's very very farfetched to think that some sinister elements, as part of a nefarious scheme, spoofed the data being sent to the satellite. IMO the probability is less than 1 in a million. So many other scenarios make more sense.

It appears to me not like overthinking, but like completely failing to understand the way the analysis I've quoted was done (if he wrote the details I've quoted were published). Nobody knew, before the analysis was done, that it would be even possible to use the Doppler effects of the communication recording.

This is what happens when you assume the conclusion first, then try to support it.

Normal inference: the plane may have gone north, satellite signals indicate it went south, therefore the plane didn't go north.

Crazy inference: the plane went north, satellite signals indicate it went south, therefore the signals were tampered with.

The Inmarsat position data was very controversial when it was first announced. This was because the flight path, the last radio contact and last radar signal were all recorded over the South China Sea; a completely different ocean to where the debris was found. Indeed, the first search operation was over and around the designated flightpath, which was literally oceans away from where the plane likely came down. Thanks to Inmarsat, we have a much better, though not completely accurate, understanding of what happened.

Note that the plane was tracked on military radar flying west from the South China Sea, across the Malay Peninsula, and over the Andaman Sea before it flew out of range. Despite this fact, the Malaysians allowed search and rescue efforts to remain focused on entirely the wrong ocean for two days before they finally told anybody that they might be looking in the wrong spot, and it was four days before they actually revealed that they had a full hour of additional radar data that showed the plane had definitely not ended up anywhere near its intended flight path.

Of course, even then the search was subsequently done in the Andaman Sea area under the assumption that the plane crashed near the position of its last radar contact, as the Inmarsat data wasn't announced until several days later.

IIRC it　was controversial as it implies that the 777 was in controlled flight for several hours after it lost contact, making it a deliberate act rather than accident. Radar data released by the Malaysian military showed clearly that the plane turned west then south.

All I'm saying is that this might support the Inmarsat conclusions, but that's not the same as proving it.

If my hypothesis is that all swans are white, finding a black swan would disprove it. Instead this is another white swan. It supports the hypothesis, but it doesn't prove it.

Note by the way that I have no reason to doubt the Inmarsat analysis. I think that was truly groundbreaking use of the data. All I'm saying is that this find doesn't really prove the analysis, nor help with narrowing down the search.

Reply on news.ycombinator.com