olasek

Inmarsat vehemently objects to this new area being called a "hotspot", they say it is a pretty large area and the search will be difficult. It may very well be the same area depicted on earlier maps since Inmarsat in this interview says all the relevant calculations were performed within 2-3 weeks of the loss. They also defend actions of the search teams which concentrated on a tiny northern sliver of the area where the pings were believed to be coming from.

Inmarsat: 'No hotspot' in search for Flight MH370 | Asia | DW.DE | 18.06.2014

RichardC10

I was trying to use the simplest possible model that will satisfy the data to the level of error in that data, which I estimate to be ~2Hz, one sigma. So I have used two parameters only, the speed of the first leg and the speed of the rest of the legs. As noted in section 7.9, the speeds of the first and second legs can be traded to move the position of the 18:29UT ping North, without affecting the fit. We are on very dodgy statistical ground here. There are only 4 reliable BFO data points at most, and I am fitting two parameters already. If I fit more parameters there is a danger of 'over-fitting', which means any data can be fitted. What the fitting process must avoid (of course the investigation knows this) is to avoid just joining the data dots with a complex model. Such a solution would definitely not be correct, there is noise in the data and this has to addressed.

I have been concerned that presentations (to the families) have made attempts to bend the path round Indonesia. The ping rings in the slides shown were not those derived from the data log (or from the slide of satellite elevations). The precisely defined 18:29UT turning point in the slides seems to be an assumption about the navigation process of the flight - perhaps correct, but not supported by any fact of which I am aware.

Better in the BFO/BTO analysis to avoid using data that cannot be 100% verified. If the data is consistent with the radar data, a consistent track should appear as a statistical acceptable solution.

Yes, I think a course can be generated that will satisfy some of the data around ~18:29UT, but I am not sure it adds any weight to the fit, that is a degree of freedom has to be used up matching the added data point.

On the time data series I will work on that. It will require an assumption about how the course changes between ping-rings (as a function of time) - at the moment I have just modelled one course per leg. That can be done of course, but it is another assumption.

RichardC10

I will include that in a full run. Effectively that is what I did in section 7.9 of the paper but didn't include much detail. It will need a different average speed from 18:29UT to 19:41UT, but that may be consistent with potential scenarios for the flight. I was trying to keep the model simple, as I said.

I wasn't overwhelmed by the ATSB fact sheet when it was released. The scale is small, but the ping-ring map seems to show the same wrong data as the Malaysian presentations, and there is speculation about possible navigation waypoints on the route. It wasn't clear what level in the investigation it had come from.

oldoberon

I had missed the fact that the small search area was due to signals picked uo by HMS Echo.


Regarding this larger area that Inmarsat have calculated, can someone confirm my understand below is correct.

A) Some time back there was a diagram showing three segments of the southern arc based on the finalpartialping , if i recall they were called something like the northern section, the middle section and the southern section, and calculations favoured the northern section, which was when they moved from the south of Perth way up north.

B) This section then has to be extended south eastward (approx flight path) to account for the final glide and that is the area they will be searching first. It is also the last area they searched for debris (presumably shifted somewhat to allow for ocean currents/time lapse).

Is that correct.

wiggy

Point of order/correction that may or may not be of relevance to the greater debate..the 777 autopilot track/heading function can be referenced (by pilot selection) to either magnetic north or True north.

FWIW looking further back in the thread I see some comments regarding a 14 degree/hour track change and.. ...Don't forget (?) the Flight Management System uses a mix of both ADIRU and GPS data to perform nav functions. In the real world you'll see a small amount of ADIRU "drift" on every sector and the FMC position will be anchored pretty much on the GPS position.

ZOOKER

Having worked nights at an ATC centre for many years, I must point out that during the wee hours, traffic may be light, but ATC, certainly where I worked, were not "half-asleep".
The charts which were used on the Horizon programme showed the AoRs of both Kuala Lumpur ATCC and that of Ho Chi Min ATCC, but also implied there was a geographical 'gap' between the areas of responsibility of the 2 centres. If this 'gap' really exists, what is it's purpose, and where are the transfer of control/communication points between the 2 ATCCs?

Pontius Navigator

@zooker, I think it might have been Singapore FIR. IIRC a chart was published earlier.

HamishMcBush

Also in the Horizon prog it was stated that in the last 8 minutes, a comms message was sent via Inmarsat that was similar to that sent at the start of a flight, implying that the systems had re-booted after power failure due to running out of fuel; plane dived, fuel re-distributed and there was suffficient to run the engines again

Question:
Would this start-up happen automatically, or would there have had to be intervention from someone on the Flight Deck to attempt to get an engine/generator to power up after having stopped due to running out of fuel?

mm43

Singapore leases on a daily basis three small parts of its FIR to Malaysia.

The delegated areas are shown in the Malaysian AIP.

Mesoman

First, kudos to RichardC10 for an impressive job and no doubt a lot of hard work.

One item that I find odd is the 718 km altitude of the virtual terminal (VT). This seems a very unlikely choice for humans to make - an altitude of 0 would be far more likely. I suppose that they use a VT plus another unstated constant time correction, and that appears as the VT in RichardC10's analysis, but it is odd.

Is it possible that the VT is zero and some other factor is hidden in there to make things work out as if it were at 718km?

It's most unfortunate that Inmarsat didn't publish the full analysis.

RichardC10

Mesoman

As you say, there is an additional constant related to the delay in that type of AES (and probably other constants) that cannot be disentangled from the VT virtual altitude with the data available. I don't think a height of zero would be chosen because:
a. it would give negative BTO value for a flight overflying 64.5E,0N
b. they didn't want to log negative numbers even with noise, so applied an offset

They really are just removing an offset, but that offset includes the variable ground station to satellite distance. The explanation does stack up with the graph of satellite elevations that was shown at a briefing, so it must be close.

sky9

It would probably have been the engines running out of fuel and the APU automatically starting up and using the residual fuel in its fuel line for a short period of time.

FoolsGold

Disclaimer: Non-pilot Librarian at aviation firm.

There seems to have been an analysis of the known and suspected tracks that was performed to determine if the plane was under human control inputs or solely under some mode of auto navigation during the initial period of "Somethings Gone Haywire" in hand off and unusual maneuvers phase.

I would question whether disparate systems of military radars would even be operating at the same system clock settings, particularly when it involves competing military officers.

Consider: System clock for an ATM transaction often differs by several minutes from System Clock for an ATM camera network. So a "noon" withdrawal from an ATM and a "noon" camera snapshot from that same ATM can be off by a few minutes. Also security system at major Miami hotel that displayed time frames on surveillance tapes of hallways, elevators and exit doors was found to be "off" by over fifteen minutes from actual event times.

Therefore, I would have doubts as to timing indications on disparate radar systems, particularly when primary returns are from different radar systems under different military jurisdictions. System clocks vary widely in all complex installations that are essentially a hodge podge of separately designed and separately installed systems.

Pearly White

Are not all aviation time-dependent systems for ATC etc. systems synchronised to a single unit time source, e.g. the Rugby Atomic Clock or similar?

Fostex

Most timing signals for this sort of equipment are driven from GPS. In fact by far the biggest use of GPS is not navigation, but rather an accurate time signal.

RichardC10

That works if you are using noiseless data, i.e. exact measurements with no error. If there is an error on the data and a particular model fits the data (with error) exactly (each data point), then you are 'fitting the noise' which must mean the model includes too many parameters than can be justified by the data and the model must therefore be wrong.

Vinnie Boombatz

Update on MH370 Search

"The search area will be confirmed before the end of June, after completion of extensive collaborative analysis by a range of specialists.

It is already clear from the provisional results of that analysis that the search zone will move, but still be on the seventh arc (where the aircraft last communicated with satellite)."

Duncan Steel, Tim Farrar, and associates hypothesize an area well to the SW:

Statement from an Independent MH370 Investigation Team | Duncan Steel

TMF Associates MSS blog » MH370: analysis of where to look?

"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. This location is consistent with an average groundspeed of approximately 470 kts and the wind conditions at the time. The exact location is dependent on specific assumptions as to the flight path before 18:38UT. "

"A report of the assumptions and approaches used to calculate the estimated location is being prepared and will be published to these web sites in the near future. "

A comment on Steel's blog conveniently plots the various search areas:

https://www.dropbox.com/s/zniuw33asz...Comparison.pdf

If the JACC opts for INMARSAT's 28 S location, note that it is very close to their 400 kt curve on their first public release (pg. 4):

http://www.inmarsat.com/wp-content/u...pler-Study.pdf

Data possibly supplied by the US NTSB in March went even further SW (40 S or even 42 S):

http://www.washingtonpost.com/rw/201...SEARCH20-G.jpg

Objects sighted may be linked to Flight MH370 - The Washington Post

p.j.m

I'm guessing the reason no one has spelled out the exact location of this "new hotspot" is because its the same or close to the original area in the South Indian Ocean that was being exhaustively searched before the sudden "discovery" of invalid pings in a totally unrelated area, causing a wild goose chase, that has now been discounted after many millions of wasted dollars and thousands of man hours effort?

Blake777

The comments in this ABC article indicate that Fugro have been tasked with concentrating on an area 1,600 km "west" of Perth. Though area of highest probability said to be 1,800km.

Missing Malaysia Airlines plane: New data prompts MH370 hunt to revisit previously searched area - Australia Network News (Australian Broadcasting Corporation)

Propduffer

The distance traveled from POVUS, (where the plane had to have been at, or very close to at 2:34) to Steel's projected location is 2,535nm. From 2:24 to 8:19 is 5hrs 45 minutes, this equals 440kts average speed for that leg.

The only possible explanation for anything like a 470kt average speed on the southern leg would be if the plane flew far to the north of Banda Aceh. This can't be because of both ping ring data and fuel on board.

270nm Subang to IGARI . . . . . . . . . (405) kts average speed
510nm IGARI to MEKAR +10nm . . . (510) "
104nm MEKAR to POVUS . . . . . . . (520) "
2535nm POVUS to 36.02S, 88.57E . (440) "
-------------------------
3,419nm in 459 minutes = 446.9 kts average speed for the entire flight

Note: Significant tailwinds figured in for the middle legs.

Note: If the last leg also had tailwinds, the TAS for that leg would have been even lower.

Conclusion: Mr Steel's numbers don't make sense.