In the latest in series of aggressive maneuvers by Russian military planes in European airspace, the Financial Times is reporting today that a Russian intelligence plane nearly caused a mid-air collision with a Swedish passenger jet on Friday while flying along a Flight Information Region (FIR) boundary with its transponder turned off.
An SAS jet taking off from Copenhagen on Friday was warned by Swedish air traffic control to change course to avoid a Russian military intelligence flight, said Swedish authorities.
Peter Hultqvist, Sweden’s defence minister, said it was “serious, inappropriate and downright dangerous” that the Russian aircraft was flying with its transponder — used to identify its position — switched off. He told Swedish reporters: “It is remarkable and very serious. There is a risk of accidents that could ultimately lead to deaths.”
The incident is the latest in a series involving Russian military aircraft over the Baltic Sea this year. In March, an SAS airliner came within 100 metres of a Russian military aircraft shortly after take-off from Copenhagen, Swedish television reported.
In the most recent incident, the Swedish and Danish military detected the Russian aircraft in international airspace on radar and warned the SAS flight, said to have been bound for Poznan, Poland.
A story about the incident in WAtoday links to a YouTube clip of ATC audio combined with speeded-up playback the commercial flight from Flightradar24.com, which indicates that the incident took place near the boundary between two FIR zones, Sweden and Rhein-UIR, with the Russian plane flying west to east along the boundary.
As I wrote in an earlier post, military pilots have been known to fly along FIR boundaries with their transponders turned off as a means of escaping detection. In what may or may not have been a coincidence, after it deviated from its planned course to Beijing, MH370 flew along the FIR boundary between Malaysia and Thailand with its transponder turned off. The pilot in Friday’s incident may have been testing NATO air defense systems to see how well the technique might work over busy Europeans airspace.
This article is dated 08 April 2014:
The German Federal Court is planning to make changes to its constitution to allow fighter jets to shoot down passenger planes that have been hijacked by terrorists to prevent them from being used to carry out attacks.
According to the changes that will be made, the Federal Defence Minister will be given the right to order that a plane that has been hijacked be shot down. Federal Interior Minister Gunter Krings said the changes to the constitution were recommended by the Federal Constitutional Court. Krings added that in times that require quick thinking and decisions, it would be impossible for the Federal Parliament to call a meeting to discuss the issue.
Last summer the court decided that for such an intervention by the military, a joint decision had to be made with the approval of the parliament. However, should the decision-making power be given to the defence minister, the previous law will be overruled.
http://www.worldbulletin.net/haber/133197/german-court-to-permiss-shooting-down-of-hijacked-planes
JS,
“We generally assume that a straight line to the SIO is reasonably likely because it’s a route flyable by an autopilot and a incapacitated crew.” [JS]
The simplest available model, providing the basis for this scenario, has the highest Aiken information criterion and as such will tend to offer us precise predictions; the smaller search area can in turn be relatively easily eliminated. The alternative is to construct a complex model which cannot avoid spreading “its predictive capability more thinly over the data space.” (MacKay, 2003).
“Am I wrong? It seems to me that there’s at least some reluctance to examine a northern route because it’s complicated.” [JS]
It isn’t relevant that we subjectively prefer the single geodesic SIO solution, or conversely that we should wish to disfavour convoluted alternatives on purely subjective grounds. The solutions produced by complex models are already each less probable.
MacKay, D. J. C. (2003). Information Theory, Inference, and Learning Algorithms. p 344. http://www.inference.phy.cam.ac.uk/itprnn/book.pdf
On northern solutions, you are free to explore the conditions that are jointly and severally needed: http://www.aqqa.org/MH370/models/aqqaBFOazimuth_v3-5.xlsx
This model presents two azimuth solutions arrays for each BFO given user-inputs for speed, rate of climb, latitude and height above the WGS84 ellipsoid. Longitude is solved for via the BTOs. The cells highlighted in green are user-editable.
The model was set up for use with Palisade’s @RISK package. The Palisade Decision Tools suite is available as a 7-day trial. If you should like simulation reports for specific scenarios feel free to email a request.
Kind regards,
Barry Martin
@nihonmama
I was thinking in the direction of a failed 9/11 type attack, with subsequent improvised claim.
See also http://www.telegraph.co.uk/news/worldnews/asia/malaysia/10700652/Malaysia-Airline-MH370-911-style-terror-allegations-resurface-in-case-of-lost-plane.html
Barry: awesome stuff. Indeed complexity in any system, whether it be global trade grounded in cheap petroleum or a working theory on MH370, is inversely proportional to the probabilty of its validity. In short, complex systems are prone to failure/invalid conclusions.
Niels: the telegraph islamist piece is a great reread. It provides a frame of a Malaysia-centric islamist hijacking with a target (the KL Petronas towers), as well as some color on Najib’s General MO (inertia as a general strategy). This sort of frame is inherently more probable than most and provides a motive as to why Malaysia would obfuscate what is known re MH370.
Now, the only question is, how would the flight come to terminate in the SIO?
JS:
Barry has already answered the questions about northern paths that you addressed to me. I agree with his answers, but I would like to address your underlying assumption that there is “…reluctance to examine a northern route because it’s complicated.” This perception is really not true at all, and it is this perception that I hope to correct. We are very much in the “leave no stone unturned camp”.
In essence, I agree that we should be looking at all possibilities and double checking everything, including northern paths. If you have followed the IG’s evolutionary work from the beginning, you know that we began with the broad assumption that the northern and southern routes were more or less mirror images. However, once the mathematical implications of the 1.7° I3F1 inclination angle became apparent, we had no choice but to acknowledge that southern routes are far more likely. (Inmarsat has not done a very good job of explaining this. Miles O’Brien did a better job explaining this with animation in the October 8th PBS Documentary.) I agree with you that this fact, in and of itself, does not completely eliminate all possibilities of a valid (albeit twisted) northern route. But when faced with finite resources for the search, priorities must be ranked, and the fact is that it is mathematically far more likely that MH370 is on the southern arc. So where would you put your resources and effort?
I would add that from time to time, I have asked other IG members to run test cases for both new northern routes, and so called “low and slow scenarios” ending on the northern end of the southern 7th arc (~20°S to 25°S). Victor has been particularly patient with me. The results rarely surface publically because despite all efforts to come up with any reasonable set of assumptions that produce low errors, none have been found (by the IG).
If that was the end of the story, then it could be argued that we should keep looking in the north. But the flip side of the story is just as important. In contrast to the difficulty of finding any northern routes or “low and slow routes” that fit the data, we (and others) have confirmed that there are many similar routes to the southern end of the southern 7th arc that fit the data with remarkably low errors using a bare minimum of VNAV and LNAV assumptions. Changing slightly the details of the assumed LNAV or VNAV mode, or the speed, only moves the point on the 7th arc back and forth a short distance while maintaining low fit error residuals (compared to errors for any of the northern routes or “low and slow” routes). This cannot be a coincidence. Or, at least, the chances of this being a pure coincidence are so close to zero that we should not be distracted by these cases.
I’ve been trying to follow the many conversations in here for a few weeks. At the end of the day, there still seems to be much room for interpretation, calculation, frustration, etc…..
The northern route still seems alive, various theories based on different motivations are still proposed, while doubts about the veracity/credibility of “official” pronouncements remain extensive and pronounced.
And the SIO search is still a dead end…for now, at least.
I feel it is legitimate to propose a scenario which I haven’t seem eliminated —
Please assume the following:
1. The plane was highjacked by the pilot, or by him with others, or by others. It’s a planned highjacking in any case.
2. The highjackers had a plan to kill (like terrorists) or to take possession of someone/something (like kidnappers/thieves). And let’s assume that the plan actually worked out. If so, by now, the former goal can be ruled out; this is a heist of sorts.
3. Plan — Act normal, then “disappear” during a hand-off, then quietly move away from civil, then military radar. Fly along a normal route, on a normal altitude, at a normal speed…while only being tracked by a single satellite. Nobody will pay attention to the normal commercial jet following a normal route…at this time of day/night.
4. A highjacking requires a plan to off-load the goods and people as soon as possible…without getting attention. The plane must be landed as soon as it is possible…and safe. The southern route for this is maybe Xmas islands or Sumatra, I guess. The western route is the Maldives/DG. I’ve seen the analyses. But what about the northern route? The north, then the east routes?
5. A Chittagong landing fits the BTO data (relative to distance), from what I can tell. The BFO data is so dependent on speed, altitude, temperature coefficients, adjustments, etc that I can’t figure out how to plot a brief layover in Chittagong followed by a straight-ish line westward across to the Golden Triangle.
Jeff himself played down this very scenario. But then again, that was before he released his own northern scenario. So please forgive me if I have overlooked some obvious data which will refute my crazy idea. I’m just curious.
I thought the speed of the route northbound towards Almaty was around 450kts and it would need to clear over the Himalayas.. So I don’t get the low and slow ranking.
@WolfgangNoack, a mathematician and scientist, tweets:
“possibly Nobody at INMARSAT knows the electronic they built in. Need an retired engineer”
@Rand a typical failed hijacking scenario would be failure to gain access to flightdeck (using explosive). How the flight then ended in SIO is still hard to imagine. I would expect attempted landing but probably not in KL.
@WolfgangNoack is probably right! I’ve seen it multiple times in my own businesses. Probably why Inmarsat continues to claim the EAFC hardware was designed to work only in the northern hemisphere. The EAFC guru guys who worked at Perth in 1999, who knew the real story about the EAFC code bug, are probably no longer around, and no one there today has ever been in touch with MITEQ to upgrade the code to V1.27, which would have eliminated the whole “partially compensated” myth. The guys that designed the AES hardware and wrote the codes are probably lost too. Hopefully, their documentation accuracy and completeness is better than mine. Would anyone today know exactly what the CU, AES and EAFC codes do, without tracking down the original team, now transferred and retired?
@Chittagong, your scenario has elements of my scenario and Victor’s. I think the problem that is going to come up is that your route will neither fit the BFO data nor explain why not.
So which flights could be candidate one that MH370 tailed right from IGIRI at about the same time?
ie: transponder turned off, no SATComs and other comms working… then MH370 would be mainly undetectable all the way outside of any radar zones.
maybe if we can get the candidate flights it could help see which way it went.
Has the IG formally asked the ATSB to publicly clarify…
– how it used the “NW point” Niels had M.Dolan confirm last week (incompatible with primary radar)
– why it claims “less available fuel” drove Mar.28 move (confirmed false by M.Dolan on its own blog)
– why it claims western range limit = E88, when its own Fig.3 (18:40 FMT) adapted to an 18:28 FMT = E84
…yet?
If not, why not? The first two go directly to their (bosses’) trustworthiness (key to assessing trustworthiness of signal data), and the last goes directly to the efficacy of the live search.
I.e. all three go directly to unravelling this mystery.
We have a major turn south. The route to the SIO following that turn is based on the assumptions that the crew were incapacitated and the plane was flown by autopilot. The flight from IGARI to the point of the major turn was apparently quite different. Various systems were shut down, the plane was turned around and flown along the Thai Malaysian border. Then the SATCOM was rebooted and the plane flew NW, following waypoints.
Questions:
1. Must we assume that the pilot (the person piloting the plane, not necessarily Shah) became incapacitated after programming the south turn? Why at this point of the flight?
1. Why did the pilot turn south? It is said that the plane presumably followed a track according to the flight manuals after passing the last programmed waypoint. Can this waypoint be identified? It might give us a clue as to the pilot’s intentions before he became incapacitated.
Hello airlandseaman,
Re yours: “The IG, Inmarsat and ATSB have independently confirmed the BTO Bias (ATSB = -495,679 usec) to within a few usec (<1km)".
That is probably correct. I did not check, say 5% confidence limits by myself, but the number appears to be reasonable.
However, you forgot to mention that a single BTO sample may be far less accurate. Take Tab.2, p.55, ATSB report (June). Absolute maximum errors in the distance associated with BTO errors were 17.7 km. That is why, despite bias BTO is probably sufficiently accurate, IG's attempts to achieve accuracy of 1 km in the estimation of the terminal point was always surprising to me.
Regards,
Oleksandr.
@ Brock
May I comment on your first point. SearchWG (and not mr. Dolan himself) has IMO sufficiently clarified the use of the hypothetical NW point. Indeed the 1912 timing leads to low speed to reach the defined little arc (drawn from nwpoint to IGREX), but is that not what one should do to estimate max fuel burn for set distance?
For the second point: can you repeat what you exactly asked to Dolan and what did he answer?
Oleksandr:
The IG has never claimed an absolute 7th arc accuracy of 1km. I don’t know where you got that idea, but it is not the IG position, or my personal estimate. My personal best guess (say, 80% confidence level) at this point is about ±10 km, certainly not ±1 km. To this range, one must add the post 2nd engine flameout path assumption. So, a search width on the order of 50 km (maybe 20 km inside and 30 km outside) is more realistic. But note that even this estimate leaves a smaller (20%?) chance it could be a bit outside that width. No one has exact statistics here. The question is one of priorities. Where are you going to search first? Clearly, you start in the middle of the 80% area, not one edge of a theoretical absolute worst case “100% area”. ATSB seems to be well aware of the issues and choices now.
@Neils – thanks for responding. To your 2 questions:
1) No, that is NOT how you determine max performance for a given distance – you do not have MH370 slow way down to 293KTAS (extremely fuel INefficient) for 50 minutes, then switch to an array of different speeds, and attempt to portray this as a performance LIMIT.
I defy anyone reading this to simulate a plausible flight path which follows both radar tracks, hits each of 7 position constraints (“NW point” & 1st 6 arcs) at each appointed time, and yet does NOT run out of fuel WELL before 0019 UTC. (Remember, this was the path presented to us as the “HIGHEST PROBABILITY” path.)
2) http://www.atsb.gov.au/infocus/posts/2014/cautious-optimism-in-search-for-mh370.aspx#!//infocus/posts/2014/cautious-optimism-in-search-for-mh370/comments/brock-mcewen-2-(4).aspx
Key word: “DESPITE”. The rest of his response (which meets the “A standard” required to qualify the JIT for the Bafflegab Olympics) claims the search move was driven (SECRETLY, because I find no mention of this rationale other than this lone blog answer to an annoying gnat of a private citizen) by the SIGNAL data. (The IG will confirm that this, too, COUNTER-indicates the move.)
(apologies – simulated path above must also end at “7th arc @ s21”; this was inferable from context, but I wouldn’t want to have anyone spend 9 hours proving me “wrong” at s33.)
Hi airlandseaman.
Citation from earlier IG statement (www_duncansteel_com/archives/date/2014/06):
“While there remain a number of uncertainties and some disagreements as to the interpretation of aspects of the data, our best estimates of a location of the aircraft at 00:11UT (the last ping ring) cluster in the Indian Ocean near 36.02S, 88.57E.”
I should note that the pointed accuracy would correspond to 100 NM, -20/+30 km is a practical approach. My main concern is a position along the arc, where the uncertainty is a way higher than 10 km in my opinion. Perhaps a section of the arc up to 17S 105E (that location, btw, is the second point where fuel-range-based curve crosses the 7th arc) cannot be excluded. Or the area around the Broken Ridge, which is a “hot spot” if one discards 18:40 BFO from the analysis. Indeed, that requires other assumptions with regard to the flight mode and/or what happened between 18:22 and 19:41.
Regards,
Oleksandr.
P.S. Replace ‘_’ in links with ‘.’.
Hi airlandseaman.
For some reason parts of the previous post dissapeared (please discard it). Reposting it:
Citation from the earlier IG statement:
“While there remain a number of uncertainties and some disagreements as to the interpretation of aspects of the data, our best estimates of a location of the aircraft at 00:11UT (the last ping ring) cluster in the Indian Ocean near 36.02S, 88.57E”
The pointed accuracy would correspond to less than 1 km northing and 0.5 km easting distance.
Later this statement was softened:
“We reiterate that our original proposed search area centred on 36.0S 88.5E has a higher probability of a successful search result than the presently envisaged official search area.
Endpoints near the final arc starting from about 39.7S 84.4E and moving along the ping arc to about 27.5S 99.9E can be computed using the same data set as we have used, depending on hypothesized flight path scenarios (e.g., very low effective ground speed, or manoeuvring) during the interval between the satellite data obtained at 18:40 UTC and 19:41 UTC. However, unless further information becomes available, we consider such scenarios to be too speculative to pursue as a group.”
However, now you again forgetting about previous BTOs, which may again shift the estimate of terminal point further off the indicated one. In other words, I am trying to say that the indicated location may not correspond to the mean in the statistical sense; it may not be a center of the most probable area. On top of it add other possible flight modes, and you will get a lot higher uncertainty than 10 km.
With regard to “a search width on the order of 50 km (maybe 20 km inside and 30 km outside) is more realistic” – agree. Though ATSB does not exclude gliding more than 100 NM, -20/+30 km is a practical approach. My main concern is a position along the arc, where the uncertainty is a way higher than 10 km in my opinion. Perhaps a section of the arc up to 17S 105E (that location, btw, is the second point where fuel-range-based curve crosses the 7th arc) cannot be excluded. Or the area around the Broken Ridge, which is a ‘hot spot’ if one discards 18:40 BFO from the analysis. Indeed, that requires other assumptions with regard to the flight mode and/or what happened between 18:22 and 19:41.
Regards,
Oleksandr.
Links:
www dot duncansteel dot com slash archives slash date slash 2014 slash 06
www dot duncansteel dot com slash archives slash date slash 2014 slash 08
@ Brock When I read both the blog from M.Dolan and appendix A from ATSB report it gives me the impression that fuel burn analysis and bto/bfo based flight path generation were not (yet) fully integrated.
(And that possibly the JIT messed up things, see below)
For your point one it is good to notice again that: “A line from IGREX to the 1912 point was used as an upper bound for the airplane performance work after loss of radar contact”
“Performance work” and “max. performance” are clearly different concepts.
Personally I don’t see reasons to doubt ATSB intentions. They seem to do what they can and provide reasonable transparency.
If pressing for more transparency I would like to focus on JIT. I have serious doubts that all parties represented in JIT play open cards and are fully motivated to find out what happened.
Oleksandr:
I hope these comments help you understand.
1. Re: “The pointed accuracy would correspond to less than 1 km northing and 0.5 km easting distance.” The number of significant digits in a latitude/longitude point has nothing to do with accuracy, and the number of significant digits in a latitude/longitude point has nothing to do with your original claim that the IG had estimated the satellite arc radius accuracy to be 1 km.
2. Re: “… now you again forgetting about previous BTOs, which may again shift the estimate of terminal point further off the indicated one.” Same answer as in #1 above. The number of significant digits in a latitude/longitude point has nothing to do with your original claim that the IG had estimated the satellite arc radius accuracy to be 1 km.
3. Re: “…On top of it add other possible flight modes, and you will get a lot higher uncertainty than 10 km.” Yes, and that is exactly what we have said all along. You seem to be confusing the BTO Bias accuracy, which in turn effects the range accuracy and ultimately the arc radius accuracy (a one dimensional accuracy) with the accuracy of the two dimensions needed to describe the search area along the 7th arc.
4. Re: “…Though ATSB does not exclude gliding more than 100 NM”. You are taking the ATSB statement completely out of context and implying something never intended by ATSB. ATSB states that the airplane would begin to spiral in after loss of the second engine, IF and Only IF, it had been on AP and no human intervened. It would only have been possible to glide 100nm if a human was flying the airplane by hand after loss of the second engine, all the way down to the water.
A catastrophy will never happen without a trace. There is no secret force hidden in a catastrophy that tries to dissolve the truth carefully split to digestible pieces for human minds to understand. A catastrophy by definition is an event that is far beyond human control. It happens. Nature is not cruel like humanity, A catastrophy is always honest No disguise whatsoever.
In contrary perfection does not happen in nature. It is always proof that an event was not catastrophic but rather manmade.
When we look at what happened in this disappearance, we see no traces whatsoever of a catastrophic event. Nothing at all. But we see the fingerprints of human intervention at any piece we touch. Starting with the perfection of the disappearance , which can possibly only be manmade, over the amount of active desinformation and seeded wrong leads, up to the refusal of cooperation and free scientific discussion, but instead you see the traces of spin doctors all over the place.
There is clearly a very homogenic picture of intent, but no trace of a catastrophy in the disappearance of 9M MRO.
So the odds are, that the flight was intentionally deviated. At that point we look very silly,
Nobody with a sound judgement in a sane mind would look for a missing Beijing bound plane 10000km away in the opposite direction in the SIO if it was not for a very small set of data , in complete, not indenpently confirmed, heavily redacted, published only after 10 weeks! by people who obviously tried to buy time. It seems very safe to say that this obstruction of the legitimate interests of the NoK and the public was completely unwarranted.
The validity of the data is, contrary to some opinions here, not at all established. Science cannot accept information that is not sourced, not independently confirmed, and so not testable and not replicable. There is not a single instance, where the data led to the confirmation of other observations. To the contrary, the data start with the opposite of which was observed by radar. A plane moving on 290 degrees suddenly is moving at 190 degrees. A lot of people will now say that there is consistency because the last radar contact was in reach of the first arc, but it could have well ben in another place. Its very little evidence to suzport the validity. Anyway given the unexplained logon and the high probability of human intervention into the comm electronics, we cannot use the data at face value. I would be happy if they were valid, but before we find one tiny piece of evidence, we cannot assume validity of the data in a scientific context. It is more likely that someone tampered with the electronicas than the other way. And if so, the data are nuts.
The other issue why we look quite silly is, that we already know from the calculations of the IG that the debris field would be somewhat messier than the swissair event : trillions of very small pieces in a depth of up to 5000 Meter. Its nearly 100% safe that we will not find a trace, even if we are in the right location.
So its a waste of time. The plane will not be found in the SIO. It will be found in a criminal investigation. And we should really push the public to demand answers. And if someone comes and says “Let time do its job!” we should know that the scum of society might think like that
@cosmic
Yes, the plane was deliberately diverted. No one would seriously doubt that from early on in the investigation. However, that did not stop the ATSB (and their team of five collaborators) and the IG from assuming flight dynamics appropriate for a routine flight from SFO to Boston.
Despite the lack of debris, the lack of additional primary radar sightings (which should have been made with their route choice), and any consideration of motive/intent, they forged ahead and stuck some pins in a map. The analytics are indeed beautiful with only one serious negative attribute to date – the plane has not been found.
airlandseaman,
1. “The number of significant digits in a latitude/longitude point has nothing to do with accuracy”. Can you tell me what is a purpose to specify a point with the precision of 1 km, and then tell that the terminal point can be elsewhere in the strip of size 1,000+ km x 50+ km (under different assumptions indeed)? So, what does the indicated point represent?
2. Nope. The terminal location along the final arc is affected by all BTOs and all BFOs. If you think that only BFOs control position along the 7th arc – it is wrong.
3. I am certainly not confused about BTO and BTO bias accuracies, and their impact, but now I am confused by yours “You seem to be confusing the BTO bias accuracy… with the accuracy of the two dimensions…” What does this mean?
4. On contrary, ATSB statement is just right in the context if you are talking about the maximum width of the strip along the 7th arc. I agree, 100 NM gliding is a scenario of much lower probability (unless it is a part of the plan), and that is why ATSB’s approach is “-20/+30 km”. I think their approach is reasonable. But whether it is sufficient or not – remains to be seen: recall that the altitude was assumed and the accuracy of the last ping ring can be +-20 km. Even without consideration of gliding, this may already correspond to the inner arc.
Regards,
Oleksandr.
Yes, yes, yes. The satellite data is based on models never used before to reconstruct a flight path. The satellite data may have been altered or perhaps we are misinterpreting it. If the plane flew the route suggested by the satellite data, it should have been detected on Indonesian radar. And no debris has yet been found in the current search area.
To hear the way people here are carrying on here, I would think some of you would propose searching any other area BUT the current search area.
The fact is the satellite data is the most precise data we have. Yes, there are all kinds of uncertainties associated with it. But the simplest scenarios, i.e., the most likely scenarios, end in the current search area. Note that does not mean the plane is likely to be found there. It means it is the MOST likely place to find the plane. Do people here understand the difference between those two statements?
I’ve yet to hear another scenario or search area that is more likely. If after several more months have passed and there is no sign of the plane in the SIO, then the likelihood of the current search area is reduced, and other ones will rise in relative likelihood. Probably first the current search area would be expanded and then other avenues explored, including scenarios that re-interpret or even totally ignore the satellite data.
Based on what we know today, it would be incredibly irresponsible to stop the ongoing search in the SIO. Is that really what some of you are proposing?
That’s not to say that in parallel we should not continue to dig up more clues and better understand the ones we have. Authorities should be pressed to release all data in their possession. Witnesses previously identified and new ones that surface should be interviewed. We should also seriously considered other scenarios, such as the spoofing scenario that Jeff has proposed. As part of these parallel investiagations, we need more and better investigative reporting.
Perhaps we can eventually convince ourselves of another scenario more likely than the one ending in the SIO. I don’t think we are there yet, at least not with the clues in the public domain.
@jeffwise —
One difference between your scenario and mine is that you have the plane flying over land for a long period of time. Land that has military radars and is contested (or closely watched) by India, Pakistan, China, Russia and the US.
In my scenario, the plane flies well away from land and radars until it briefly lands at Chittagong. Afterwards, it flies east/northeastward over unsophisticated, unpopulated land towards and over the Golden Triangle.
The idea is the same as Victor’s, but is less complicated (no second plane).
I am not capable of comparing BFO data to the proposed flight path, but the BTO fits, right? The time line fits, yes?
@Brock —
In response to your dare…how about if the plane refuels (perhaps partially; not fully), then flies in a straight line from Chittagong to the northeast corner of Burma (the “top” of the triangle)?
@Victor
If you have not heard a more likely scenario you have not been listening. The Christmas Island scenario accounts for motive/intent, lack of debris (see ocean current models), and the lack of radar sightings.
BTW, it can be made to fit the BTO and BFO data as I have posted.
You, and other members of the IG, are trapped in an incorrect paradigm. Not saying mine is correct, but at least I am not trapped.
Victor:
Well said, and exactly on target.
Victor,
So, what are these simplest scenarios?
Oleksandr.
@Dennis: Christmas Island as a termination point can only be forced to fit the BFO with climbs/descents at strategic times and by ignoring the 7th arc BTO. Essentially, you are saying a good portion of the satellite data cannot be used to discriminate the path. If you are going to ignore the satellite data, then you have to open the spectrum of possible scenarios to others equally likely, including northern paths, which also would explain the lack of Indonesian radar data, no debris in the SIO, and even provide a motivation. A flight to the Maldives/Diego Garcia is also possible.
Because a landing on Christmas Island is only one of many scenarios that are possible if you ignore the satellite data, I view a landing on Christmas Island less likely than a crash in the SIO. Is there are a reason why it should be preferred to any other scenario that ignores the satellite data?
You might believe that I am stuck in a paradigm. In my opinion, no other scenario yet rises to higher likelihood than a crash in the SIO. I do believe it is possible, though, which is why I am seriously considering the possibility of spoofed satellite data, either in part (the BFO data) or in whole (the BTO and BFO data).
Dennis, are you suggesting that the search in the SIO be terminated at this time and instead focus be placed on a landing at Christmas Island?
@airlandseaman
I am beginning to piece together who the IG postser are.
Sad really.
@Victor
I never said MH370 made it to Christmas Island. My scenario has it running out of fuel South and East of the Island. A scenario supported by totally independent flight simulations,
@Victor
BTW, I provide a BTO/BFO fit for this path at the urging of Flitzer, another IG member obviously, who did not believe I was perfectly capable of making the calculations. I was put off by that initially, but finally attributed it to the academic nature of the IG.
@Dennis: I referred to a landing on Christmas Island because that is the scenario of Ken St Aubin, who references your work as evidence of his scenario. I do recall now that your scenario differs from his in that you predict the plane crashes before actually reaching Christmas Island. Of course, that begs the question as to why debris was not spotted along the 7th arc when searched.
If you allow for climbs/descents and changes of direction and speed along the way, there is a broad range of paths that fit the data, of which an end point near Christmas Island is one. This range of paths includes paths that end near or on the island of Java. Why is there is a preference for a path ending near Christmas Island as opposed to other paths? How about a landing on Java, which would explain the lack of debris? Do you exclude the possibility of western or northern paths?
Dennis, are you suggesting that the search in the SIO should be terminated at this time and instead a focus be placed on a crash near Christmas Island? This is a key question.
So just to ask as I have not found how to get a list of candidate flights that could have acted as shadows to MH370 while MH370 flew out of radar coverage ? I am currently thinking the shadowing started near IGIRI or BITOD.
I would certainly recommend curtailing (abandoning just to be clear) the search in the current search area (which is very improbable for a number of reasons). I would not recommend resuming the search anywhere else without more investigation.
@Dennis: What additional investigation do you propose? Why not investigate in parallel with the current search effort? Why is any other search area more probable than the current?
I am very much in favor of additional investigation efforts that might turn up a more likely search area or scenario. I hope that is clear.
Searching is expensive. Better to stop until better places are determined. The current search area is based on very poor assumptions,
One reason to expand the scenarios is that in my opinion there should have been debris floating or sunk in those high priority areas (just like AF447 debris spotted in the impact area)
As none spotted even during the sonar scans we must find another path not derived from Inmarsat. Maybe it landed else where.
@Dennis, I asked you for some numbers for your track model in order that we might see some evidence that your assumptions were plausible. Surely that does not constitute disbelief in your ability to calculate.
I recall you criticising others [including the IG] for making constant altitude and constant speed assumptions [even though the V13.1 model does not do so], yet you make these assumptions yourself [4000m, and 500Km/hr] without any justification at all. Do you have any evidence for these assumptions or are they just plucked out of the air.
@DL,
In response to your questions, perhaps the activities at 18:25 were a last ditch effort to land the aircraft as oxygen was running out. A fire producing toxic smoke may have caused the flight crew to don oxygen masks at 17:22, when the flight was diverted to Kota Bharu (airport).
I believe the final waypoint entry was Maimun Saleh Airport (WITN). The purpose was the final attempt to land. The aircraft had already made close passes over two airports, but there is no evidence they ever descended then. IMO the biggest obstacle to this scenario is that there was no descent to low altitude, and I have not heard any satisfactory theory on this point. On the other hand, IMO all the hijacking scenarios have even larger hurdles to be overcome such as the motive and the goal.
One minor point is that the NW course was set well before the SDU was powered up (possibly as just one of several pieces of equipment) at about 18:23. The final turn would have been commanded by setting the FMS waypoint to WITN at about 18:27. That would take the aircraft to WITN, passing over it (again without descending) at about 18:34. At WITN the FMS would switch to a true track course at a bearing of about 192 degrees. While many seem to agree that navigation by waypoint was most likely, I am not aware of any routes that fit the satellite data and pass through any other nearby waypoints besides WITN. Richard Godfrey’s V13.1 route goes to ISBIX, but that is much farther south and I can’t see why a pilot near MEKAR would choose ISBIX. WITN makes sense because it is the closest airport at that time. Banda Aceh is quite plausible, but the only way to go there and fit the satellite data is to land there and wait on the ground before taking off again (this scenario was done by Victor Ianello).
@Niels: re: “fuel burn analysis and bto/bfo based flight path generation were not (yet) fully integrated” this is similar to airlandseaman’s “stovepiping” comment. To make sure I understand this argument, let me paraphrase:
“The JIT allowed the feasible zone (and, in the same round of decision-making, the search itself…) to move way up to s21 on the strength of analysis starting at [NW point at 1912] because NOBODY BOTHERED TO TELL THE FUEL ANALYSTS that this starting point was several hundred nautical miles too close to where others in the investigation already KNEW MH370 was pinned by primary radar at 1822.”
Have I got this argument about right? If so, I think its status on the commonsense spectrum is self-evident. Roughly 75% of the JACC’s acronym represent ANTONYMS to “stovepiping” and “not integrated”.
@Cosmic: well said – but 2 quibbles:
1) BTO/BFO & radar tracks conflict re: inferred destination (west v south), but they do dovetail precisely into each other – see Dr. Ulich’s seminal paper, and Duncan Steel’s response on behalf of the IG.
2) The IG’s meticulous analysis of the signal data leaves NONE of us looking silly – if the data was faked, then it follows that the search is, too – in which case the plan involves wreckage that is faked, intentionally corroded, or non-existent. In ANY of those cases, the IG’s work denies the JIT the leeway to FURTHER delay the search by reinventing yet more fanciful interpretations of the data.
@Victor: also very well said. For the record, I do NOT recommend aborting the SIO search (just extend it to E84!) – but investigating the investigators is, at this point, paramount. Please help me improve my #day200audit data request, such that it is of a quality sufficient to convince at least ONE member of the IG to re-tweet it by day 300.
Shifting toward family for a few days – best wishes to all, and thanks for being such a vibrant, courteous, and committed community.
We WILL uncover the truth.
P.S. @Nihonmama: please keep fencing with Rand in my absence. You do it best, anyway.
It would be nice to know if MH370 started to transmit again some time later (after mar8) and if this was captured in the Inmarsat’s logs.
@airlandseaman, Barry, and Victor
Thanks for the explanation. It would seem that whether there is reluctance or not to explore convoluted routes, the SIO still needs to be priority in terms of searching. So in practical terms it probably doesn’t matter whether the northern routes are not explorer because they are unlikely or because they are expensive. It seems that each of you have a more nuanced approach than what may have come across to me before, so thanks for clarifying.
Let me ask a more technical question. I understand that the BTO values are time offsets – but from what? Are they from the beginning of a signal block, or from the time of some other event? What is the width of a signal block?
I was always suspicious of ISAT’s storage of BTO values, because it appears that the values are jammed into an unreasonably small record size (in bits), or they are unreasonably rounded to 20us.
Some quick background – please bear with me:
The maximum possible round trip variation in all BTOs is about 40,000us with 0 being directly under the satellite and 40,000 being at the edge of the satellite’s range. Rounded to 20us means that only 11 bits are required to store any BTO value, and because it’s round trip, it could be further reduced to 10 bits. The next highest boundary of 16 bits would support either the storage of the unrounded BTO, or the time without subtracting the nominal terminal. At 24 bits, the timing value does not need to be offset or rounded at all.
Nevertheless, our understanding is that the values are offset and/or rounded to save space. The entire daily worldwide BTO set would only require a few MB of storage. If saving space was really the rationale, it’s conceivable that the values were further cropped at the 8bit mark – which would still be enough to track any plane that pinged every few hours.
So, that brings me to my question. What if the BTO values later in the flight are offset by one complete signal block width or one complete byte? Is this possible?
If the signal block size is small enough, could there be a bad assumption at work here as far as which block the signals arrived in?
Likewise, could there be an assumption that the BTO was truncated at 8 bits, when in fact it hasn’t been?
As far as I’m aware there is a Dutch company that has a contract to search a big chunk of seabed, and they are being paid by the Australian govt to do it, and noone else was offering. Moving the goal posts now will look like guesswork. If nothing is found the Malaysians will say we tried, too bad, and the Australian govt will wind it up. It sounds like there are some crunchers out there who think their obsession will go on and on but I doubt it. Is there another govt out there that cares where this thing is? If so name them.
@Brock:
Thanks for your vote of confidence.
I’m taking a little hiatus now as well.
So I’ll punt to MattyPerth, who is extremely adept at fencing — and who I’m sure (if he’s around) will answer if/when Dear Rand pops up from his nap.
On the way out, a gem, tweeted tonight by renowned former CSFB trader and investor Aly-Khan Satchu
“If they can get you asking the wrong questions, they don’t have to worry about answers.” ~Thomas Pynchon
http://www.rich.co.ke/rctools/wrapup.php
Happy Holidays everyone.
And say a prayer for the families of MH370.
Hi Bobby,
It looks like we are on the same page.
There is one more runway ‘nearby’ – Car Nikobar. It is longer than Maimun Saleh Airport, and its orientation could explain final heading after a failed landing attempt. In such a case a sequence of BFOs at 18:25-18:27 could possibly indicate descent rather than a turn.
While ISBIX is an interesting waypoint (a number of my trajectories also pass by), I can’t find a reasonable explanation why would somebody program FMC to ISBIX. Except a scenario if landing really took place, and the purpose was to get rid of the aircraft.
With regard to yours “The aircraft had already made close passes over two airports, but there is no evidence they ever descended then” I am not so sure. Ignored radar data (altitude changes) may be such an indication. For some reason members of IG deny any significant changes in the altitude. In addition, I suspect the initial intent was to return back to KLIA, but the plan changed when the aircraft was around Penang.
Regards,
Oleksandr.
The Jitter Report requested from ATSB several months ago has finally been released here:
http://www.atsb.gov.au/publications/2014/mh370-burst-timing-offset.aspx
The new ATSB data and analysis confirms IG expectations. It also confirms the 7th arc absolute accuracy estimates as follows:
• 50% chance it is within ±3.0 km of the nominal 7th arc
• 95% chance it is within ±8.3 km of the nominal 7th arc
• 99% chance it is within ± 10.6 km of the nominal 7th arc
This means the nominal 7th arc is known to ± 5.7 NM at the 99% confidence level, and it is more likely than not within ± 1.62 NM of the nominal 7th arc. This is the best quantitative evidence we have on the 7th arc accuracy. It verifies that the nominal arc is well known, and the remaining uncertainty is driven by the post fuel exhaustion flight path uncertainty. Unfortunately, we do not have precise measurements or statistics on this part. But the scant data we have at 0019 combined with the full motion class D simulations that have been conducted by ATSB (and verified by me personally in a B777 simulator on Nov 2, 2014) provide guidance. Based on those data and simulations, and assuming no human controlled flight, it is more likely than not that the aircraft descended rapidly in series of fairly steep turns and Phugoid oscillations following fuel exhaustion, placing the point of impact within ~5-10NM of the arc. Taking these two estimates together, we have a conservative suggested search width of ~ ± 16 NM. Given the direction of flight at fuel exhaustion, the search width should be biased slightly to the outside, so a better recommendation might be -12/+20 NM relative to the nominal 7th arc.