The INMARSAT satellite geolocation is over the equator on the Africa side of the center of the Indian Ocean. MH370 was climbing out of Kuala Lumpur towards Vietnam heading North East - away from the satellite. It was in the fringes of the cover of the Pacific INMARSAT satellite but had set up a connection with the Indian Ocean INMARSAT satellite. They only started flying toward the Indian Ocean satellite after breaking radio contact and turning back.

As Robin Clark pointed out in his post on this subject, the aircraft on initial climb out had the satellite at - 73.17° (to the left), which meant that any Doppler shift with reference to the Inmarsat 3-F1 satellite would be increasing.

However, that is not the end of the story, as it is the aircraft that senses the Doppler shift between it and the satellite and shifts its Tx frequency to compensate. If flying toward the satellite it would shift its Tx frequency downward so that any transmission from the aircraft would arrive at the satellite at the assigned and expected Rx frequency. Likewise when flying away from the Satellite, the Doppler shift would be deceasing and the aircraft would adjust its Tx frequency in the opposite direction, i.e. upwards.

The satellite has nothing to do with the Doppler correction; that is assigned to each individual aircraft, otherwise the satellite would have a very large job on its hands when communicating with many aircraft. What the Satellite does do, is transmit a constant carrier on the P Channel which the aircraft uses to sense the received frequency offset from the assigned channel and then adjusts its Tx channel(s) in the opposite direct to compensate.

Any Doppler shift between the the Ground Station (GES) and the satellite is corrected at the earth side, which allows the separate isolation of Doppler between the Sat and the aircraft (AES).

Now start thinking about where this "burst frequency offset" info is coming from.

How will analysis of the last search phase affect the next?
 

Porker1 said and
Heathrow Flyer (quoting an independent Raytheon assessment of Inmarsat data) said:
This would tend to agree with the hypothesis that Ocean Shield's limited and non-continuous pinger detections were genuine (but distant) second or even third "convergence zone" in origin. Sound propagation in deep water tends to be distorted by thermoclines in the depths (abrupt changes of temperature at different levels in deep water (aka layering)). It tends to cause large amplitude focussed ripples in detected acoustics (i.e. an erratically moving incomplete annulus every 30 to 50 miles or so - at or near the surface). Accomplished and experienced sonar operators can tell the difference between first, second and third convergence zones. This effect for a sea-bottomed "pinger" might also be focussed in a particular direction by the ocean bottom's topography (think of MH370's pinger being inside a deep valley with its pinger's acoustics being channelled in the direction of the valley or canyon's mouth). Ocean hydrographers cannot predict or allow for these geomorphic effects and time-warps in detected ensonification. It's at the same time both diffraction and refraction via its water-routing - and for a bottomed object, reflection. The phenomenon is a great friend of submariners as it can cloak their actual position even when making noisy high-speed progress. All they need do is "hide" beneath a layer and noisily propagate in all directions. After swamping all listening sonobuoys and fixed arrays with their high-speed sound signature, they then slow and go to a silent undetectable loiter mode. Concentrating an MH370 sonics search within the immediate area of a few pinger detections was overly optimistic and seems to have disregarded the characteristics of sound in deep water.
Add in the Raytheon claimed discrepancy and it's still a much larger and indefinite search area than those you see on the AMSA/JACC charts. I guess we will see whether this will be reflected in the next phase search zone prognostications.

It's worth noting as far as range of detection goes, that a blackbox pinger's job is to be heard, whilst a submariner's intention is never to be heard - yet, in my experience, long distant detections during aerial ASW sonics search was the norm for conventional snorting and/or nuclear boats (particularly the Soviet Echo II boats and later).

Desert 185 said: Fair comment for a mere nasty brown smell, but does not reflect what an injured crew might do after/during a sudden oxygen flare. There's no oxy on/off valve on the flight-deck and someone blinded by flame might just think that instantly killing the electrics might bring relief - from what is essentially a brief explosion and subsequent fireball. You only have to look at the cockpit photo of the MS 777 "Nefertiti" scorching to grasp that. I'd be surprised if they (or any surviving "he") didn't bale to the cabin immediately. But I'd also not be surprised if they were cooked where they sat.

Unctuous... Excellent post.

I was wondering if the depth etc of the actual receiver would negate any of these anomalies.

I would have thought that if the receiver was virtually on top of the pings then that should give an accurate position. I would assume that IF the boxes are actually there then the receiver would have passed pretty close to one of them.

Of course does not necessarily mean the aircraft is there as well, which is probably why the are having so much trouble locating them with the UAV.

Point taken on the deep water ping progagation.

But the spectrum/waterfall of the recorded pings looked pretty clean. Except that the detected frequency was so offset by 4.2 kHz from 37.5 kHz.

The area searched was about 15 NM x 15 NM around the ping that was heard for 2 hours and 20 mins(!!). Do you think propagation can go beyond this?

Human intervention ....
 

Quote:
The disappearance of Malaysia Airlines Flight MH370 could have only involved human input, according to two aviation experts.

A senior Boeing 777 captain and a former crash investigator agreed that given the information outlined in the preliminary report issued late last week, "human input was essential" for the MH370 to end its flight in the Indian Ocean.
Comments from qualified airline pilots (preferably Boeing 777) would be welcome, please.



Val'd....
For the jet to leave it's programmed waypoint to waypoint flight plan to China;stored in the flight management system; under normal conditions would have required a human to intervene.
1. A pilot could take the controls manually, click the autopilot off and hand fly the jet whatever direction or altitude.
2. A pilot could modify the upcoming legs in the active flight plan to have the autoflight system steer the jet to new waypoints.
3. A pilot could build a second flight plan in the second flight management computer, activate and execute the second flight plan and either hand fly or direct intercept a waypoint on the new flight plan.
4. A pilot could leave the autopilot on, but steer to whatever heading via heading select mode.

On "my" 777's, the company dispatch center can send via data link; new flight plans, to the inactive flight management computer system. To my knowledge, only a human actually on the flight deck of the aircraft could activate and execute such secondary flight plan.

Whatever capability an outside agency (human) might have to activate a secondary flight plan is unknown......

The independent report on the ICAO publication
 

Ref: Steve Winter’s report at:

Analysis: MH370 Preliminary Report | Air Traffic Management | Air Traffic Management - ATM and CMS Industry online, the latest air traffic control industry, CAA, ANSP, SESAR and NEXTGEN news, events, supplier directory and magazine

It’s not my job to act as cheerleader for the analysis we have from the International Panel but Steve Winter has raised some points that I believe can be understood on the basis of the published data, so I offer the following comments:

I have edited the extracts of the report for length.
I was also surprised that the turning point could be so tightly defined on the basis of the data to hand; as plotted it may reflect a best fit to all the ping-arc data, with the North/South constraint of the 18.30 ping-arcs. However, as I have proposed in previous posts, it is the value of 00:11 BFO that sets the final red-zone search area along the 00:11 ping-arc, not the overall course/speed from the ping arc data (at least in the case of level flight at 00:11). The 00:11 BFO has a latitude dependence via component D2.Again as I have proposed before, the high-probability area is set on the basis the aircraft was level at 30000ft at 00:11 so there was no effect of descent rate on the BFO. The width of this zone is set by the total statistical uncertainty. The maps show different heights for the three probability areas, the writer seems to have missed this.

There is no marked area to the North of the red-zone as any climb from 30000ft at 00:11 was regarded as improbable (climb shifts the area North).

Further thought since the last post on the differentiation of the red/green/yellow zones. If a descent is allowed as an option at 00:11, then the final BFO value loses most of its value for setting the speed (and hence course) – the mixture of position and descent contributions cannot be disentangled in one measurement. If there was any BFO measurement from the 00:19 partial pings they would be similarly affected as the aircraft would be descending steeply, since 00:19 is generally attributed as the fuel exhaustion point.

With the 00:11 point discounted for this purpose, the selection of the speed/course drops back to the earlier ping-BFO values. The mid-probability/green area may reflect the best fit to these data points only. With a speed/course selected on this basis the 00:11 ping then gives a predicted descent rate. However, the earlier BFO data must have lower value in setting the speed/course in this situation so the predicted area of search might get quite wide (and more so if the initial turning point is not pinned down accurately). The aircraft analysis based on possible autopilot modes/aircraft configuration may constrain the possible descent rates at 00:11 and so still provide some limit in the Southern direction particularly (which corresponds to higher descent speeds).

With thanks to SKS777FLYER:
OK, that takes care of Nav (i.e. path over the ground/sea). The route followed could all have been input to the FMS and handed to the automatic flight system ('autopilot') at the time of deviation from the original route to Bejing. No further human input to Nav would have been necessary. Everyone on board could have been dead from that point onward and the route over the ground/sea would still have been flown.

However, could a vertical 'profile' to match the claimed visits to (lets assume) 39000ft and 5000ft be input to the FMS, handed to the automatic flight system ('autopilot') at (let's assume, for ease of discussion) the time of deviation from the original route to Bejing and thereafter be followed by the aircraft without further human intervention?

In other words, after input of the profile incorporating 39000ft and 5000ft, could everyone on board have been dead from that point and the profile still have been flown by the automatic flight system?

Val
 

It could be programmed in such a way that it could make a series of climbs or a descents but not both (other than the final descent after the fuel runs out!!)

If the aircraft is flown in VNAV (vertical flight path) and is at say 37000 and a way point altitude of 6000 feet is entered at some point down line then the aircraft will descend to reach 6000 feet point in flight idle descent PROVIDED that 6000 has been entered in to the MCP altitude window, the reverse (climb) is true, but unless some one makes a further amendment to the MCP target it won't climb or descend again.

If you were at 37000 feet and selected 43000 at some point down line (even if this was above the aircrafts CRZ capability at that mass) and entered that in the MCP window, then the aircraft would attempt to make that climb, the speed would slowly decline until just above the min speed and VNAV would likely revert to level change with near zero rate of climb, it shouldn't stall, VNAV would not re engage without human input.

In other words it will not leave any FL or altitude in VNAV unless the MCP window is set to a different value, if the FMC constraint was FL100 and you selected 100ft in the MCP it would only descend to FL100, not 100 feet.

You could program multiple climbs or descents, but not both, so starting at 370 you could set WP 1 to 350, WP2 to 300, WP 3 to 270 and so on, BUT the MCP would have to equal the lowest programmed value for it to achieve that, ditto the reverse in the climb, so stepped climbs or descents are of course possible (creating the illusion that someone is in control?) but not up and down without someone in control !!

This flight went where (ever?) it did because someone made it so, not a Li on battery fire or cockpit 02 fire.

The and only saving grace in this event (from an industry point of view) is that it wasn't a dream liner, that could well have resulted in a fleet wide grounding, even if not justified by the facts, but after the previous battery problems one could understand why

Pingology
 

Did you also account for the satellite motion due to orbit inclination?

Good post LNIDA, that should clarify the nav system operation.

The important point bought out is that there are only two possible scenarios. At the southerly turn point there was either a live operator at the controls to manually select a southerly heading with the heading select control or the FMS had been preprogrammed to make the turn and a live operator was not required.

Either option suggests that the person in control of the aircraft had some training on nav. system operation but it does not imply that a crew member was flying the aircraft. The system is simple enough to operate that a lightly trained hijacker could have steered the aircraft - after all hijackers took flying lessons for 9/11.

Could it not have both but at different times :confused:

yes it is curious. if there has been an explanation I have missed it. aircraft carries ulb. its range is 1 to 2 km in normal conditions. ulb signal is heard so the bearing of it from the tpl is known but not the range but hey we dont need that, it says on the tin it is no more than 2 km away. textbook stuff so far but ulb and aircraft not found. it seems ocean shield and echo together could not establish even a position line. how did this or any other ulb get certificated with such a dismal specification? apparently the number of successful ulb locations via pinger, ever, can be counted on one hand. just another piece of kit to add to the list of recommended upgrades. so its got to be the hard way, the ulb is obviously down there somewhere so its now a long slog to find it a la af 447.

ULB propagation different from submarine
 

An important factor differentiating submarine sound propagation from ULB ping propagation is attenuation of the signal. At low frequencies used for submarine detection, there is virtually no attenuation. Signal decrease is due only to distance (r-squared law) and that can be significantly lessened by focusing effects.

At the 33.5kHz frequency of the assumed pinger detection, the ocean absorbs sound at a rate of about 4.5 dB per kM. This is in addition to r-squared loss. This substantially limits detection range, and is no doubt why the original search area was small. It also leads me to believe that, if the original pinger detect was valid (and I've yet to see another explanation for the signals), the wreckage should still be found close to the detect locations.

It will be interesting to see what the reviewers have to say about this, if any is made public.

See this attenuation calculator: Calculation of absorption of sound in seawater

Interesting article on detection of ULBs and their practical operating ranges. Apologies if it has been posted previously.

Deep-water Black Box Retrieval - November 2009, Volume 13, Number 09 - Archive - Hydro International

@Propduffer
Mr Winter seems to have a good Internet presence, e.g.
ATCA Technical Writing Awards

The last radar sighting at 18:22 is shown in the ICAO report maps and in the slides shown at the Beijing meeting a day or so before. Agreed that the lat/long position of the 18:22 sighting seems to have moved West since previous reports.

Thank you!

I'm at a loss why People feel so urged to exclude (or to foster on the other Hand) possible causes in this case?
What's the benefit of that?
Does it help improve safety?
Does it make somebody sleep better at night?

Why can't we just wait until they find the boxes?

Pingology...
 

I have made attempted to account for all motion between the aircraft and the satellite..
The Inmarsat 3F1 was moving very slowly westward during the whole flight , this was not enough to significantly change the 'doppler' shift of the signals routed through it....IMHO...
It was also moving northward away from the equator very slowly , reaching a limit at around 18:41UTC , and then moving back toward the equator again . It is this 'wobble' which Inmarsat suggests could help distinguish between an aircraft route in the northern hemisphere , and one in the southern............via subtle differences....
By far the greatest motion is the up and down direction...........although the figures published are relative to the surface straight down and Kuala Lumpur is part of the way around the world , and so lower values would be experienced there .

In the chart the first section is defined with negative values as the satellite is ascending . The latter part as positive values as the satellite is descending . I converted the speeds to feet per minute , as this helps relate it to aircraft motion............


http://www.14594.mrsite.com/USERIMAGES/satspeedvert.jpg

Australian 14-15 budget
 

With a very hardline Federal budget due out next Tuesday, I wonder how much pressure the Australian government will find themselves under to wrap up/pass on the associated costs?

I'd like to find the answers with MH370 but at some stage the line has to be drawn. It's great publicity for Australia throughout the world but to fund an indefinite search might be close to political suicide (by a government who are about to implement a debt levy and take an axe to government spending).

Will be interesting to see what questions are asked come Tuesday.

Propduffer and Richard,

The last position given for some radar contact, surmised to be MH370, as given to the Chinese families was 200 nm from Butterworth on a 295 radial.

I don't know how you can "move" that point westward unless you possess some additional radar targets that have not been released to the families.