Content uploaded by Vincent Lyne
Author content
All content in this area was uploaded by Vincent Lyne on Dec 08, 2021
Content may be subject to copyright.
Unravelled Cryptic Pathways of MH370-
Goodbye and Travel-Path Secrecy Reveals
Landing Details and Precise Location
Version 1.0: 3rd February 2021. Version 2.0: Updated Flight Path and Location Precision. 7th February 2021
Version 3.0, Updated Flight Path using corrected arcs and Defence Jindalee Radar Network Range. Reconciling that the Pilot’s home
simulator track crosses precise final location. Further analysis of seafloor sound propagation and aircraft damage to confirm glided landing
and implosion at depth. 2nd March 2021. Updated to illustrate the Batavia Seamount sound emission. 14th March 2021. Identifying and
tracking MODIS anomalies. 18th March 2021. Updated to reconcile the hydroacoustic sounds analyses of Usama Kadri. 8th December 2021.
Vincent Lyne (BEng Hons, MEngSc, PhD)
Retired Scientist
Private Citizen
Hobart. Tasmania. Australia.
Enquiries: Melissa Lyne (BSc)
Email: melissa.lyne@gmail.com
For background, please read three reports at:
https://www.researchgate.n et/publication /348590528_Goodbye_from_MH370_Defines_8th_Arc_First_of_4_I terative_Reports
https://www.researchgate.n et/publication /348677217_Scen ario-free_Evidence-
based_Location_Range_of_MH370_and_Feedback_Assessment_of_Scenarios_Second_of_4_Iterative_Reports
https://www.researchgate.net/publication/348742993_Resurrecting_Abandoned_Hydroacoustic_Evidence_of_MH370_From_Haystacks_to_Haystack_to_Ne
edle_to_the_Remaining_Pinpoint_Third_of_4_Iterative_Reports
Special Acknowledgement: I am very grateful and indebted to Dr Alec Duncan for his
knowledgeable, patient, and thorough responses to my questions on acoustics. Any errors,
omissions and misinterpretations with the final conclusions are mine.
Short Summary:
We conclude our thought review by showing that the bewildering array of MH370 sound anomalies,
including lack of triangulation is however explainable and entirely consistent with deep-water sound
injection from MH370 at a precise location that coincides with our theory—which also reconciles the
detailed divergent hydroacoustic interpretations by Usama Kadri (2019). Triangulation is not possible
as two acoustic stations receive an emitted water-borne signal from a seamount, whilst another
receives a mysterious “seismic” signal. We review the Pilot’s home simulator track and find that it
crosses precisely over our location – indicating that the track was for planning purposes, rather than
the assumed ditching in the Southern Ocean. Recovered debris and their locations on the aircraft
suggest a glided landing with major impact to the right, and later dislodgement of fore and aft
internal parts from possible implosion(s) at depth. Highly unusual anomalies identified in MODIS
images for April and June 2014 are consistent in location, drift direction and drift speed with
debris/material from the final location. All evidence therefore confirms the postulated scenario and
the final location of MH370 near where the longitude of Penang intersects the Broken Ridge. All
other possible locations are ruled out, including the Batavia Seamount, following derivation of a
detailed flight path that matches our scenario and explains the persistent southerly path (to avoid
JORN) that veered eastward before the 7th arc to an attempted glide landing at the final location.
Long Summary:
We wrap up this report series on MH370 by attempting to unravel the cryptic pathways of the final
sound from MH370 which was heard at the hydroacoustic stations at Diego Garcia (at 1:58 UTC), the
Perth Canyon (at 1:33:44 UTC) and Cape Leeuwin (1:34:50 UTC). The Diego Garcia signal pointed to a
source near Madagascar at station HA08s1. The Perth Canyon and Cape Leeuwin stations had
bearings towards Maldives and off north-eastern Africa, but with no corresponding signals at Diego
Garcia2. Our analyses attempt to resolve whether these signals could have come from our
postulated location or a different location.
Discussions on the web suggested the Gulden Draak Rise as the source (which was surveyed3)4.
While the Gulden Draak Rise approximately explains the Perth Canyon and Cape Leeuwin signals, its
arrival time at Diego Garcia isn’t plausible and the bearing at Diego Garcia is inconsistent.
The only plausible precise resolution of all observations involves a deep-water directed injection of
sound from our postulated location that proceeds along the seafloor to the Gulden Draak Rise and
then, after it is intensified, radiates off a slightly deeper northern seamount, the Batavia Seamount,
as a weak water-borne signal before being detected at the Perth Canyon and Cape Leeuwin at the
correct bearing and time difference. This water-borne signal from this seamount is not at the correct
bearing, nor heard in the available record at Diego Garcia at an earlier time—possibly buried in the
noisy record.
The deep injection also sends a signal to the seafloor to the west that travels at approximately
10,665 km/hour (about twice the water-borne sound speed) along the deep-water channel at the
base of the Broken Ridge escarpment. This seafloor signal proceeds from the western end of Broken
Ridge directed towards Madagascar, where it is reflected and heard as the mysterious seismic signal
at the correct bearing and the correct time to Diego Garcia. A water-borne signal from the seamount
near Gulden Draak Rise would arrive at Diego Garcia earlier (by about 6.5 minutes), and at a
different bearing than the seafloor signal, but this appears not to have been observed, or perhaps
buried in the noisy record? Even if it were to be observed (at the right bearing), this does not negate
the explanation for the 1:58 UTC signal as both signals could have plausibly been generated. We
present maps and an elevation profile to assess the plausibility for the Batavia Seamount sound
emission as a “bathymetric radiator” scenario.
To dismiss Batavia Seamount as the landing location, we derive the flight path to match the timings
and location of the arcs, and the need for secrecy in skirting around the Jindalee-Over-the-Horizon-
Network (JORN) range. We find a persistent southerly path from the 3rd arc up to the about the 6th
arc that was responsible for investigators thinking the aircraft was on autopilot and later running of
fuel at the 7th arc. However, the path veers eastward between the 6th and 7th arcs to the final
location, and the 7th arc handshake was most likely from a warm reboot related to preparing the
aircraft for a precise water landing. We also find that the Pilot’s home simulator track crosses
precisely over the final location, indicating that the track was for determining waypoints rather than
the assumed ditching in the Southern Ocean.
1 https://www.mirror.co.uk/news/world-news/mh370-mystery-missing-wreckage-theory-13931848
2 https://www.nature.com/news/sound-clue-in-hunt-for-mh370-1.15390
3 https://www.atsb.gov.au/media/5773565/operational-search-for-mh370_final_3oct2017.pdf
4 https://370location.org/2017/03/mh370-acoustics-a-7th-arc-candidate-near-gulden-draak-seamount/
Analyses of recovered parts and their location on MH370 further supports the glided water landing
scenario, with major impact to the right of the aircraft and later dislodgement of internal parts from
the front and rear most likely due to implosion(s). Extremely high chlorophyl anomalies identified in
MODIS images for April and June 2014 are consistent in location, drift direction and drift speed with
debris and possible dissolved/floating material from the final location, that affected ocean colour
estimates from MODIS.
These final analyses confirm that our postulated location is consistent with evidence: the strict
constraints imposed by the hydro-acoustic signals which past investigations abandoned; the drifter
path analyses (in the earlier reports); and the requirements imposed by the average aircraft speed
and time to sound generation in the deep ocean. Circumstantial evidence from the derived flight
path, the home simulator track, MODIS anomalies, the recovered aircraft parts, and links to possible
implosions all provide support for the hypothesis and landing location.
This completes what we can extract, and parsimoniously interpret from existing studies.
If a new search is commenced, I recommend the suggestion made by Dr Peter Last that the final
search, whether successful or not, also involve a survey of this unique biogeographic deep-water
environment and its diverse sea life, which were once part of Gondwana and the Kerguelen Plateau;
and to this day contain relic species of the past5. Following the expedition, a commemorative
publication is recommended that pays tribute to the victims and their loved ones, the many-
disappointed involved with the searches, and the unique biogeography of this region.
The final parameters for MH370 are as follows:
Latitude: 33.0oS
Longitude: Longitude of Penang
Water depth: ~5960 m
Deep-Water Sound Generation Time: 1:11.7 UTC
Landing Time (at average 420 NM/hr): 0:52.7 UTC
Time Delay to Sound: 19 minutes
Faster or slower average aircraft speeds will change the landing time, but the Sound Generation
Time remains the same.
Version 2.0. Update (7/2/21): We use the scenario to derive a detailed flight path that not only rules
out locations such as the Batavia Seamount, but also provides a more precise final location that
involves Perth, Adelaide and the 95.5oE longitude. Note that the arcs used previously and in this
version were later updated.
Version 3.0. Update (2/3/21): Updated Flight Path using corrected arcs and Defence Jindalee Radar
Network Range. Reconciling that the Pilot’s home simulator track precisely crosses the final location.
Further analysis of seafloor sound propagation and aircraft damage to confirm glided landing and
implosion(s) at depth. 2nd March 2021.
Version 4.0. Update (8/12/21): Our analyses reconcile and resolve sounds interpreted by Usama
Kadri (Kadri 2019), using his acoustic gravity wave theory, as coming from our postulated location,
rather than the two divergent and irreconcilable routes he proposed. Nonetheless, Kadri’s insightful
5 https://pubmed.ncbi.nlm.nih.gov/27988737/ : A new deepwater legskate, Sinobatis kotlyari n. sp.
(Rajiformes, Anacanthobatidae) from the southeastern Indian Ocean on Broken Ridge.
analysis supports our proposal that the sound heard at Diego Garcia was sea-floor sound, and that
the sound heard at Cape Leeuwin and Perth Canyon were waterborne. This is a remarkable and
unexpected demonstration that the postulated location can reconcile what experts thought were
divergent or irrelevant observations.
MH370 Features
The final map illustrating the main features involved in interpreting the MH370 evidence is shown in
Figure 1.
Figure 1 Final map showing the main features from the proposed location of MH370. The white curves are the 7 arcs. The
postulated location is at the Penang longitude (green vertical line) near 33oS, between the yellow and red lines which go
from the 7th arc across the Broken Ridge to the Perth Canyon. The general drifter path is shown in cyan taken from the most
promising drifter simulations; note how the path follows closely the curvatures of the Broken Ridge system (shown above
the cyan curve by the red curve (channel) and the brown curve (ridge)) up to the end of the deep-water “break” region
where it curves across towards Africa. Localities of note are: Perth Canyon (RCS station), Cape Leeuwin (HA01 station),
Diego Garcia (HA08s station), Maldives, Madagascar, Gulden Draak Rise and the Batavia Seamount. The top image shows
the overall scene and the lower is a zoomed in image.
The elements of what we want to demonstrate are noted in the Long Summary so I will only present
the salient features of the analyses. The strategy for the analyses is the same as that for the location
analyses, namely:
Analysis Strategy
1. Keep focussed on the main outcome of the analyses which is to firm up or refute the
location postulated for MH370. Don’t get distracted by interesting details at the expense of
losing sight of the focus. This is a very easy trap to fall into with the hydro-acoustics,
especially where and what caused sound generation at depth. We leave this for others to
resolve and focus only on plausible scenarios.
2. Construct scenarios for the location and the evidence. Initially I went with the scenario that
the sound was generated at the location and was water-borne within the sound channel in
the ocean which is at about 1000 m depth in this region6.
3. Test the scenarios against the evidence. This involved testing triangulations and to
determine any issues with the data or with the scenario.
4. Update scenario and data interpretations, as necessary. This was the most difficult part as
there were inconsistencies with the water-borne theory. I next went with a scenario of
deep-water injection of sound and propagation along the seafloor. Whilst this could explain
the Diego Garcia signal, the Perth Canyon and Cape Leeuwin signals could only be explained
by a water-borne source from the Batavia Seamount.
5. These observations led to the third hybrid scenario where sound was injected at depth and
proceeded to Diego Garcia and also north along the seafloor before being radiated from the
Batavia Seamount as a weak water-borne signal that was monitored at the Perth Canyon
and Cape Leeuwin. I’m not able to explain why it preferentially radiated from the Batavia
Seamount rather than the shallower but wider Gulden Draak Rise. I can only guess that the
morphology (narrower exit) and depth must have played some part? I leave this for others
to resolve. Update Version 3.0: We present bathymetry details and an elevation profile to
support the notion of sound emission from the Batavia Seamount.
6. These analyses carried me as far as I needed as the location and scenario were both not only
confirmed by the very cryptic evidence, but details emerged of the time of sound generation
and when the aircraft may have landed. If the location and interpretations are right, all
evidence must point to the location. It would have been easy to abandon the evidence or
dismiss it as noise or some other unexplained source, as some have done, but if you are right
all evidence must intersect at the right location. We used the above adaptive strategy to
improve our understanding and analyses. This feedback adaptation is a fundamental process
in evolution, rather than the slower wasteful mutation process7 of trying this and that and
expecting to get lucky.
In the next sections I will summarise the results of these scenarios and analyses.
6 Alec Duncan (pers. comm. 2021)
7 Don’t wait for the virus to mutate because it will adapt at much faster rates as we have witnessed!
Water-Borne Scenario
To facilitate the water borne analysis, I define what I call the Chorus Ratio which will give us an
understanding of which locations viably explain where sounds originate from.
The equation for when sound is heard is as follows:
= + + +
Where, is the time when the sound is heard,
is the time when the aircraft turns and starts it descent from the 7th arc,
is the elapsed time for the aircraft to go from the start to a landing location,
is the elapsed time delay from landing to sound generation,
is the elapsed time for the sound to travel from generation to the signal
station.
For each test location, construct equations for two signal stations (multiple stations are also
possible but let’s keep this simple for now) and subtract them so that the elapsed-time terms for
plane travel and sound-generation delay drop off (since they are the same for both signal stations).
We can then define the Chorus Ratio, so-called because this value is equal to 1.0 when you have
found the right landing location where the two signal generation times are in tune:
= (
) / (
-
))
Where, definitions as the same as the previous equation and the superscripts refer to signal
stations (1 and 2). Note that the only variable relevant to the plane is the travel times of sound, and
there is no dependence on aircraft average speed or the time delay before sound generation. We
can scale this up to multiple stations using all plausible pairings and multiplying them together. This
will lead to N * (N – 1) / 2 pairings. Finding the right location will allow us to determine the time of
sound generation.
Let’s deal with the Perth Canyon and Diego Garcia pairing for now which is shown in Figure 2.
Figure 2 Latitude and longitude arc along which the Chorus Ratio is equal to 1.0 for signals heard at the Perth Canyon and
Diego Garcia. The line “Diego+200s” and “Diego+300s” show the arcs if the sound was heard at Diego Garcia after 200 and
300 seconds respectively from the actual time it was heard (at 1:58 UTC, so the 300 seconds curve would be for a 2:03 UTC
signal). The green dashed lines mark the Penang-longitude location along the Broken Ridge.
At a latitude of 33oS, the identified location is within the 7th arc search area, and its bearing at Diego
Garcia is not consistent. Along the path from Perth Canyon to the Maldives, the location is south
west of the Batavia Seamount (Figure 3), which incidentally is also at the longitude of Penang! Yes,
Batavia also matches the time and bearing to Cape Leeuwin! The inevitable BUT is that it reaches
Diego Garcia at the wrong time (6.5 minutes earlier) and at the wrong bearing. Besides, the aircraft
goes from the 6th arc to the 7th arc and then back to the 6th arc and taking quite a long time doing so
(about 45 minutes).
An in-tune Chorus Ratio at our postulated location can only be possible if the Diego Garcia signal is
received at about 2:03 UTC, 300 seconds after the 1:58 UTC signal. With the record during this
period under embargo we are unable to proceed further with this. Let’s suppose there is a signal
with the right timing and bearing. This would then be at odds with the timing at Cape Leeuwin (the
delay between the Perth Canyon and Cape Leeuwin signals would be too short to explain).
The mismatches lead us to abandon the water-borne scenario, but we take note that a water-borne
signal from the Batavia Seamount, which is at the right bearing, gives a time difference between the
Perth Canyon and Cape Leeuwin signals of 59.2 seconds compared to an observed 63 seconds—the
uncertainty is within the noisy estimate of when the sound started (or when it reached its peak). It is
an excellent match of bearing and time difference, under the circumstances of the noisy signal and
environment, but the signal is weak particularly at Cape Leeuwin. Let’s store this for future reference
for now as it is our best lead on plausible water-borne signals arriving at the Perth Canyon and Cape
Leeuwin.
Figure 3 Location of the Gulden Draak Rise and the Batavia Seamount. Source: ATSB:
https://www.atsb.gov.au/media/5773565/operational-search-for-mh370_final_3oct2017.pdf
Deep-water Sound Injection Scenario
There are a myriad of airtight vessels, and systems like the batteries, which are all capable of
catastrophic destruction by pressures at depth. For example, there are 14 large heavy-duty tires
inflated to about 200 psi8 (equivalent to the pressure in the ocean at about 140 m depth). What
depth would these implode at, and what sort of sound would be generated by 14 of these going off?
8 https://en.wikipedia.org/wiki/Aircraft_tire
Then there are the high-pressure spheres used for rapid deployment of the slides in emergency
situations which are inflated to 3,000 pounds per square inch (200 standard atmospheres)9. This
pressure is equivalent to 2000 m in the ocean and failure of the pressure vessels would be expected
to occur at much deeper depths. Our work would never end if we were to try and identify where
sound would be generated and whether or not it would have a strong directional component –
which would depend on precisely how the plane is oriented in the water column as it sinks. All we
can do is assume that deep-water sound injection is possible and then assess the consequences of
this assumption against observations.
First, let’s assume this is possible and the seafloor sound is heard at all signal stations. Immediately
we know from the time difference at Perth Canyon and Cape Leeuwin that this is not possible. This
essentially rejects the second scenario, so we move onto the third which is the hybrid scenario.
The path to Diego Garcia: Let’s assume there is deep-water injection along Broken Ridge. The
escarpment at Broken Ridge is below 2000 m and at its base there is a deep-water channel filled
with sediments coming off the escarpment. Let’s assume seafloor sound proceeds along Broken
Ridge, scoots off the western end to Madagascar then reaches Diego Garcia. What speed of sound is
necessary for the sound to be heard at Diego Garcia at 1:58 UTC? We can’t resolve this unless we
have a start time for the sound, or we link up a start time with a backtrack of when sounds were
heard at the Perth Canyon and Cape Leeuwin. If memory serves us right, those signals are water-
borne off the Batavia Seamount, and we can plausibly assume that the sound travels via the seafloor
to Batavia Seamount before it is radiated as weak water-borne signals to the Perth Canyon and Cape
Leeuwin stations (see later update for a reference to this possibility). This would also explain the
weak signals observed, particularly at Cape Leeuwin.
Let’s look at the statistics of this scenario (refer to Figure 1 for the various paths):
ID
Variable
Value
P1
Path length from western end of Broken Ridge to Diego Garcia via
Madagasca (seafloor sound)
3881 NM
P2
Western end of Broken Ridge to Penang longitude location (seafloor
sound)
569 NM
P3
Penang longitude to Batavia Seamount (seafloor sound)
436 NM
P4
Batavia Seamount to Perth Canyon (water-borne sound)
859 NM
P5
Batavia Seamount to Cape Leeuwin (water-borne sound)
907 NM
The eastern path is from the Penang longitude location directly north (along the Penang longitude
remember – hence the coincidence of Batavia Seamount also with the Penang longitude) via the
seafloor to the Batavia Seamount and then water-borne radiation and travel to the Perth Canyon
and Cape Leeuwin (and possibly as a very weak water-borne signal to Diego Garcia?). We have
already seen that the water-borne path leads to the right bearing and time difference. This leads to
sound generation times at the Penang Location of 1:11.57 UTC and 1:11.68 UTC for Perth Canyon
and Cape Leeuwin respectively.
We can now use these start times to calculate the speed of sound required to reach Diego Garcia at
precisely 1:58 UTC. This speeds turns out to be 1.6 NM/s compared to the water-borne speed of
9 https://en.wikipedia.org/wiki/Evacuation_slide
sound of 0.81 NM/s – the seafloor speed is almost double as expected10. Of course, there is some
uncertainty about the precise path and speeds but given the rapid speed of sound along the seafloor
the agreement in timing and bearing is very remarkable. This scenario is compatible with the timing
and bearings from all 3 stations and we adopt the results as being highly plausible and in very close
agreement with the landing location at the Penang longitude. The scenario also explains the reason
for the mystery seismic signal monitored at Diego Garcia, which we now know was from MH370.
Finally, we plot the Chorus Ratio for this scenario and find that sounds from the postulated location
are in tune at Diego Garcia, Perth Canyon and Cape Leeuwin for a water-borne sound speed of 1460
m/s, and double that for the seafloor speed (Figure 4).
Figure 4 Contours of the Chorus Ratio for the seafloor scenario showing that signals from the postulated location (marked
by the intersecting green lines) are in tune for a sound speed of 1460 m/s in water.
Comparison with Usama Kadri’s Analyses
Two divergent possible routes for MH370 were derived by Kadri (2019), shown in Figure 5, from
detailed acoustic gravity wave analyses of potential sounds recorded at hydrophone stations at
10 https://www.researchgate.net/publication/330711872_Effect_of_sea-
bottom_elasticity_on_the_propagation_of_acoustic-gravity_waves_from_impacting_objects
Longitude (E)
Latitude (S)
-2
-1
0
1
2
3
4
5
94 96 98 100 102 104
-35 -30 -25 -20
Diego Garcia (Station HA08s) at 01:58 UTC and at Cape Leeuwin (Station HA01) at 01:34 UTC on the
8th March 2014. Kadri (2019) analysed possible sound source locations corresponding to
transmission routes combining water and sea-floor paths according to the “acoustic gravity wave”
model. This gave a range of possible source locations depending on the mix of seafloor and
waterborne components of the sound paths. As shown in Figure 5, the matchup with our theory is
possible if the sound heard at Diego Garcia originates from the Madagascar end (longest potential
path away from detector) as sea-floor sound, whilst the sound heard at Cape Leeuwin originates
from Batavia Seamount (shortest possible path to detector) as water-borne sound. This
interpretation remarkably reconciles and pinpoints both sounds emanating from one source,
compared to the two divergent incompatible routes and crash sites along extended uncertain paths
which confused Kadri despite his insightful interpretations (which we now know are plausible for a
different reason) that both signals were potential MH370 candidate sounds. This is yet another
surprising, unexpected reconciliation of seeming divergent observations that baffled expert
investigators, which adds significant credibility to the MH370 location identified by our theory.
Figure 5 Map from Kadri (2019) showing analyses of potential sounds recorded at hydrophone stations at Diego Garcia
(Station HA08s) and Cape Leeuwin (Station HA01) on 7th and 8th March 2014 between 23:00 and 04:00 UTC. Two potential
divergent flight paths (in orange) are shown by Kadri corresponding to assessments of the sounds recorded at two stations:
Route I for the sound heard at Cape Leeuwin at 01:34 UTC, and Route II for the sound heard at Diego Garcia at 01:58 UTC.
The white-grey polygons represent possible sound source locations corresponding to transmission routes combining water
and sea-floor paths according to Kadri’s “acoustic gravity wave” model. We have marked red circles around the most likely
source locations according to our theory as follows: the sound heard at Diego Garcia originates from the Madagascar end
as sea-floor sound; whilst the sound heard at Cape Leeuwin originates from Batavia Seamount as water-borne sound.
Other elements of the map as described by Kadri are as follows: “Purple: bearing of signals recorded on HA01 that could be
associated with MH370…. Cyan: bearing of signals recorded on HA08s that could be associated with MH370, based on Table 3 - bearings
170° and 234° could be related. White: satellite data of the last ‘handshake’ with MH370, known as the 7th arc. Red: bearings of military
action that were recorded intermittently on HA08s between 23:00–04:00 UTC. Orange: two possible MH370 routes; only route I is in
agreement with the 7th arc. Attribution: Data SIO, NOAA, U.S. Navy, NGA, GEBCO; ©2018 Basarsoft; US Dept of State Geographer; ©2018
Google.”
Landing Parameters
The selected scenario allows us to calculate the landing and sinking parameters for MH370.
The time to land at the Penang longitude location is shown in Figure 5. At a speed of 420 NM/hour, it
takes nearly 34 minutes to reach the landing site. These times are relatively precise as we have had
to juggle a very sensitive balancing of extremely fast sound speeds in confirming the Penang
longitude location and in deriving the sound generation time.
Figure 6 Time for aircraft to land as a function of the average speed (in NM/hour) from the 7th arc to the landing location at
the Penang longitude. The y-axis is the actual time, and the time to travel can be calculated by subtracting the time at the
7th arc of 0:19 UTC.
The precise sound generation time also allows us to calculate the average descent speed in the
ocean of the plane by assuming a depth at which sound is generated, which allows for floating time
at the surface plus sinking. For an assumed depth of 5000 m the sinking speeds are as shown in
Figure 6. Shallow assumed depths will require slower speeds.
0
10
20
30
40
50
60
70
480 460 440 420 400 380 360 340 320 300
Time mins past 0:00 UTC
Aircraft Speed (NM/hour)
Landing Time
Figure 7 Minimum descent speed of the aircraft to a depth of 5000 m, where for illustration sound is assumed to be
generated, as a function of the average aircraft speed from the 7th arc to the landing location at the Penang longitude.
Finally, the delay from landing to sound generation is shown in Figure 7. For an average speed of 420
NM/hour, the delay is about 19 minutes. The shortest delay is just over 5 minutes at a slow average
speed of 300 NM/hour which is just over the descent speed. At the cruise speed it is about 23
minutes.
Figure 8 The time delay from landing to sound generation as a function of the average aircraft speed from the 7th arc to the
landing location at the Penang longitude.
0
2
4
6
8
10
12
14
16
480 460 440 420 400 380 360 340 320 300
Descent Speed (m/s)
Aircraft Speed (NM/hour)
Speed to 5000m
0
5
10
15
20
25
480 460 440 420 400 380 360 340 320 300
Time Delay (mins)
Aircraft Speed (NM/hour)
Delay to Sound
Updated Flight Path:
Version 1, 7/2/2021: Flight Path from 3rd Arc using standard JORN
Version 2, 2/3/2021: Update Flight Path using Defence JORN
In the previous version of this report (detailed above), the highly plausible emission of sound from
Batavia Seamount, its longitude in line with Penang, and human connotations with the mutiny
aboard the Batavia11, raised some doubts in my mind as to whether we should have dismissed it as a
plausible landing site.
The scenario where this might be possible is one where the aircraft spends a long time (about 45
minutes) surveying the landing site and the prevailing wind waves before taking a long slow glide to
land. The lack of sound at Diego Garcia could be explained as being too weak to be detected and/or
discriminated from the background noise. It might be plausible and pragmatically, the eastern half of
Batavia Seamount has already been surveyed so it’s not a great effort to have the western half
surveyed, which should be done anyway to complete the survey of this unique micro continent12.
This concession sat uneasily with me as the scenario I proposed was a very meticulous one that
fitted the evidence well. What I didn’t investigate with the scenario was the flight path leading up to
the 7th arc, and whether that path (according to the scenario) might shed some light as to whether
the Batavia Seamount was a plausible site or not.
The scenario is predicated on the plane being undiscoverable at its final location and during its path
to that location. Hence the secrecy with communications being turned off, or not responded to. A
key element of this secrecy in relation to the path, was to avoid detection by JORN (Jindalee Over-
the-horizon Network)13; in particular from the station at Laverton.
In the first version of this update, we used the standard JORN range which had a southern boundary
that was a latitudinal boundary at the latitude of Laverton. In the second version, the southern
boundary is revised to be compatible with a Defence publication which shows the boundary
extending south-west to north-east, with the south-west tip of the boundary intersecting the 3000
km range circle at about 30oS, and a western longitude limit of 92oE (Figure 8). In the second update,
we also revise some issues with the arcs in the first update. This led to a simpler flight path that
accorded with published technical analyses of the range of potential flight paths shown in Figure 9.
11 https://en.wikipedia.org/wiki/Batavia_(1628_ship)
12 https://eos.org/science-updates/geological-insights-from-malaysia-airlines-flight-mh370-search
13 https://en.wikipedia.org/wiki/Jindalee_Operational_Radar_Network
Figure 9 Coverage of the Jindalee-Over-the-Horizon Radar Network (JORN) and its stations. Note that the range from
Laverton which is used in this report is inclined north-east to south-west with a southern boundary tip in the west at
approximately 30oS. Source: https://adbr.com.au/over-the-horizon/
Figure 10 Range of plausible flight paths up to the 7th arc based on timing of satellite communication signals from MH370.
Source:
https://en.wikipedia.org/wiki/Timeline_of_Malaysia_Airlines_Flight_370#/media/File:Reunion_debris_compared_to_MH3
70_flight_paths_and_underwater_search_area.png
Finally, we revisited the simulated tracks on the Pilot’s home simulator. This is shown in Figure 10
taken directly from a report by the Australian Transport Safety Bureau. The remarkable aspect of the
simulator track is that it crosses precisely over the postulated final location, in yet another
circumstantial confirmation of the final location. Early investigations assumed that the track
suggested a final location in the Southern Ocean when that location was most likely a waypoint that
would allow crossing over our postulated location.
Figure 11 Path reconstructed from data found on Pilot’s home simulator, together with the 7th arc. Note that the simulator
track crosses precisely over the postulated final location – shown by the red dot. Source image from:
https://www.atsb.gov.au/media/5773565/operational-search-for-mh370_final_3oct2017.pdf
Apart from the timing of the arcs, we examined plausible paths that were consistent with the
assumptions:
1. Skirting around the edge of the 3000 km Defence JORN range from Laverton.
2. Navigation waypoints kept as simple as possible.
3. Assume the aircraft cruises at an average of 440 NM/hour and that transit distances
between the arcs are compatible with timings. This speed was derived after-the-fact in
trying to fit the flight path to be compatible with the timings of the arcs and in keeping with
the plausible flight path range in Figure 9; but initial speeds from the 1st to the 3rd arc may
have been higher.
4. Once past the southern limit of the JORN range, commence the easterly path and descent to
the final location (in the general direction of Perth).
Using Google Earth Pro, the derived path that was closely consistent with the assumptions is shown
in Figure 11.
Figure 12 MH370 flight path derived from the scenario requiring secrecy in skirting around the JORN range (shown as the
purple circle of 3000 km radius centred at Laverton, and by the purple line for the southern boundary), and a path to the
final location marked by the red circle with centre near 33oS and the longitude of Penang (vertical green line). A zoomed in
view is shown below the overall view. From about the 3rd arc, the plane heads southerly along the 92oE longitude till it
passes the JORN southern boundary tip at about 30oS, between the 5th and 6th arcs. Thereafter, the path changes to either
the Adelaide waypoint, or south-easterly to a waypoint at 33oS and 95oE. At this point, which occurs just before the 7th arc,
the aircraft changes to an easterly path on its final descent path towards the longitude of Penang. The orange line is the
path found on the pilot’s home simulator which has a southern waypoint at about 46oS and 105oE (Figure 10). As shown in
the zoomed in view, this path passes directly over the final location. We suspect that this path was an early simulation to
investigate waypoints, flight duration and/or direct paths which pass over the final location.
The final derived path shows the following features:
Path
Latitude
(deg S)
Longitude
(deg E)
A general south westerly path from the 1st arc till the 3rd arc.
At about the 3rd arc, the path changes to a persistent southerly along
the 92oE longitude, which is outside of the JORN range, till beyond the
5
th
arc.
92
At about 30oS and 92oE is the south-western limit of the JORN range. At
the latitude of 31oS, and just before the 6th arc, the course changes to
south-easterly and proceeds outside the JORN range.
31
92
Two paths are possible to the final destination: 1) a path towards
Adelaide stopping at the Penang longitude; 2) a south-east path to 33oS
and 95oE followed by an easterly path along 33oS to the Penang
longitude. The latter path is 20 NM longer and requires a course change
just before the 7th arc. Either path ends in the final destination at 33oS
and the Penang longitude.
33
Penang
At an average speed of 440 NM/hour, the derived paths met all constraints imposed by the timings
and distances whilst ensuring secrecy (which may not have been needed as JORN may not have been
operational at that time, nor effective even it was operational14). The addition of Adelaide provided
an alternate route to the same final location, but this may not have been used as the slightly longer
path based on integer latitudes and longitudes was simpler and appears to account for changes
made before the 7th arc handshake and the change to a final eastward path.
Figure 11 also displays the early flight path found on the Pilot’s home simulator showing a straight
path (in orange) from northwest of Sumatra ending in the Southern Ocean at about 46oS, 105oE.
Given that this path passes directly over the final location, we suspect this was an early attempt at
investigating flight duration, and/or finding waypoints to the final location via a direct path. A
second flight path found on the computer (see image at:
https://nymag.com/intelligencer/2016/07/mh370-pilot-flew-suicide-route-on-home-simulator.html)
resembles more closely the path we derived and may indicate a change of plans to avoid detection
by the JORN from Laverton. Without recognition of the role of JORN and the final location, these
paths do not make sense and may have led the initial searches astray; in contrast, these paths
provide additional circumstantial confirmation of the planning around our proposed final location.
To summarise, the final location has not changed from where the 33oS latitude is intersected by the
longitude of Penang. The path analysis and match ups with the timing of the arcs further confirm this
location, as does the flight paths found on the Pilot’s home simulator.
Returning now to the original question on the Batavia Seamount, we can conclusively state that this
scenario is not compatible with that final location as it is well within the JORN search range—hence
that location would be a major error inconsistent with the meticulous planning.
To summarise the final scenario, that explains all evidence and potentially also the mysterious
seismic signal at Diego Garcia, is as follows:
1. The path of MH370 was planned to be as simple and direct as possible whilst keeping
outside the JORN 3000 km range from Laverton.
2. The aircraft travels persistently southward along the 92oE longitude from about the 3rd arc
till just after the 5th arc. This persistent path led investigators astray in thinking that the
aircraft was in autopilot mode and destined to travel that path till it ran out of fuel.
14 https://en.wikipedia.org/wiki/Jindalee_Operational_Radar_Network
3. After passing the JORN southern boundary at about 30oS, the path changes to south-
eastward from 31oS, heading towards 33oS and 95oE, or else to Adelaide.
4. The aircraft reaches 33oS and 95oE just before the 7th arc, thereafter the final descent path is
easterly along the 33oS latitude.
5. The final location is where this path is intersected by the longitude of Penang.
6. The destination is an ultra-deep location, nearing 6000 m in water depth where the plane is
meant to be undiscoverable, and has personal significance connected to Penang.
7. After the course change to the final path, a handshake is initiated monitored at the 7th arc,
presumably in preparation for a glide landing – and possibly involving the Aux Power Unit.
The aircraft has enough fuel to reach its destination and starts its descent, or else (Plan B) it
is out of fuel and prepares to glide to a pre-mature landing; the latter is unlikely but needs to
be included to cover alternate interpretations of technical events at the 7th arc (handshake,
BFO analyses of rapid descent).
8. Approximately 34 minutes passes before the plane makes as smooth a landing as possible at
the precise location. The right flaperon may have been subjected to large lateral forces
breaking off its hinges15 during the landing, or subsequent deep-water events including
implosions, crushing of the central and wing fuel tanks, and crashing against the seafloor.
9. After 19 minutes or so, deep-water sounds are injected into the seafloor at precisely 1:11.6
UTC.
10. The northern component of this seafloor signal emits as a water-borne signal from the
Batavia Seamount which arrives at the Perth Canyon station at 1:33:44 UTC and the Cape
Leeuwin station at 1:34:50 UTC.
11. The western component proceeds along the seafloor channel at the base of the Broken
Ridge escarpment. It shoots out the western end of the ridge and reflects off Madagascar to
the Diego Garcia station. The signal at 1:58 UTC matched the arrival of the MH370 signal as a
seafloor signal, travelling roughly at double the water-borne speed.
12. The sound generation time was precisely determined by requiring all stations to match the
time of arrival and the bearing of arrival.
13. Update 8th December 2021: Our interpretation reconciled the analyses of the
hydroacoustic sounds by Kadri (2019), and clarified how the sounds propagated in the
water column and seafloor. We also showed that the divergent inconsistent MH370 routes
proposed by Kadri could in fact be reconciled as sound generated at depth from our
postulated location.
14. The time of arrival allowed us to estimate the time of landing and other parameters of the
MH370 final phase which all seemed plausible and reasonable.
Finally, the survey tracks of all bathymetric surveys conducted during the MH370 search (Figure 12)
shows that the tracks just miss the Penang longitude location despite the huge area surveyed
(710,000 km2).
15 http://jeffwise.net/2017/06/19/how-did-mh370s-flaperon-come-off/comment-page-1/
Figure 13 Tracks of the bathymetric surveys conducted during the search for MH370. Note that the Penang longitude
location marked by the white cross was not traversed. Source: https://www.atsb.gov.au/media/5773565/operational-
search-for-mh370_final_3oct2017.pdf
Update Version 3: Bathymetry Around Final Location
We present here the bathymetric information around the final location site, with a view to
understanding the circumstances of why the sound heard at the Perth Canyon and Cape Leeuwin
came from the Batavia Seamount, and why the sound heard at Diego Garcia may have been from a
seafloor signal that progressed along the Broken Ridge and reflected off Madagascar. And, to
understand in greater detail the environment surrounding the final location.
We describe the bathymetry around the region through a series of depth bands:
1. Ultra-Deep: From 4000 m to 7000 m
2. Deep: From 2000 m to 4000 m, which is below the Deep Sound Channel (next layer)
3. Deep Sound Channel: From 500 m to 2000 m, is the predominant layer within which sound
propagates as a water-borne signal
4. Shallow: Less than 500 m – we do not examine this layer
Ultra-Deep Layer: 4000 m to 7000 m
The bathymetry greater than 4000 m and up to 7000 m depth shown in Figure 13 has a number of
broad features:
1. To the north of Broken Ridge is a shallower triangular plate surrounded further north and
east by deeper basins.
2. The mini-continent on which Batavia Seamount is located (at about the latitude of Shark Bay
in Western Australia), is seen extending out from this triangle, with the Batavia Seamount
almost at the apex before it deepens again.
3. To the west of Broken Ridge runs the Ninety East Ridge system16.
4. The plate continues further west before it deepens into a basin that is just to the east of
Madagascar.
5. The path from Madagascar to Diego Garcia passes a northern basin slightly deeper than
4000 m before it shallows again towards Diego Garcia.
Figure 14 General bathymetry depths from 4000 m (reds) to 7000 m (deep blue). Zoomed in view shown in Figure 14.
The zoomed in view of the bathymetry from 4000 m to 7000 m (Figure 14) shows a deep hole at
about 33oS and the Penang longitude (100.27oE). This is the approximate location proposed in our
theory as the resting place for MH370. The hole is almost 6000 m deep and isolated by being
surrounded by bathymetry shallower than 4000 m. South west of this location are even deeper holes
that extend down to almost 7000 m (Figure 14).
To summarise the key features of this layer, the path of sound to Batavia Seamount is possible
especially as it is at the apex of the plate north of Broken Ridge, potentially implying that the
16 https://en.wikipedia.org/wiki/Ninety_East_Ridge
seafloor sound may be reflected off the triangular sides to be concentrated to the apex where
Batavia Seamount is located.
The postulated final location is in an isolated deep hole, but deeper holes do exit so the intention
was not solely based on depth of the holes. The additional considerations were possibly: flight
endurance, ensuring that the location was south of the JORN range and in line with an integer
latitude (33oS) but also intersecting the Penang longitude.
Figure 15 Zoomed in view of Figure 13 showing black dots aligned at the intersection of longitudes from 99oE to 101oE, and
latitudes from -34oS to -32oS. The approximate location of the Penang longitude is in the deep hole slightly east of the
middle dot.
Deep Layer: 2000 m to 4000 m
The depth layer from 2000 to 4000 m (Figure 15) shows the following features:
1. The Broken Ridge appears as an east-west escarpment rising from 4000 m and extending for
some 1400 km.
2. The Gulden Draak Rise and Batavia Seamount also rise from 4000 m as isolated structures.
The main conclusion to be drawn from this depth layer is that the seafloor sound must ideally
propagate in the layer deeper than 4000 m, both along the Broken Ridge escarpment and to the
Batavia Seamount. While seafloor sound could be generated in a shallower layer up to 2000 m, it
must proceed deeper before it can escape from the structures in this layer. We discuss this in
relation to Batavia Seamount in the next section.
Figure 16 General bathymetry depths from 2000 m (reds) to 4000 m (deep blue).
Sound Channel Layer: 500 m to 2000 m
The sound channel is the depth layer in the ocean within which the speed of sound is minimal17.
Sound emitted within the 500 m to 2000 m can be expected to propagate within the sound channel
as a water-borne signal. The bathymetry image of this layer (Figure 16, Figure 17) shows that both
the Gulden Draak Rise and Batavia Seamount are embedded in this layer as isolated features.
Coupling of seafloor to water-borne sound emission has been shown to be possible18, and this is the
main mechanism invoked in our explanation of the sound heard at the Perth Canyon and Cape
Leeuwin if indeed the sound was generated as a seafloor sound from MH370. This mechanism is in
concert with the focussing mechanism associated with the Batavia Seamount being located at the
apex of the triangular plate from the Broken Ridge. The shallowest site of the Batavia Seamount is
located at its eastern end, which is within reach of the sound channel.
17 See for example: https://370location.org/2017/05/a-global-sofar-channel-wave-speed-map/
18 See review section on “Models of the T-phase Arrival” in Dziak et al. (2011):
https://www.researchgate.net/publication/262260364_Hydroacoustic_Monitoring_of_Oceanic_Spreading_Ce
nters_Past_Present_and_Future
Figure 17 General bathymetry depths from 500 m (reds) to 2000 m (deep blue). This is the approximate layer within which
sound is expected to propagate as a water-borne signal in the ocean. Note that both the Gulden Draak Rise and Batavia
Seamount are embedded in this layer as isolated features.
Figure 18 Zoomed in view of Figure 16.
Penang Longitude Elevation Profile and Batavia Seamount Sound Emission
In this section, we explore the mechanism responsible for the emission of sound from the Batavia
Seamount and implications of that mechanism for how sound at the postulated location was
injected into the seafloor.
The possible mechanisms for sound emission from the seafloor are summarised in the report by
Ralph Stephen (http://msg.whoi.edu/CTBTO_Key_Lecture_RAS.pdf) and shown in Figure 18. The
mechanism we envisage for Batavia Seamount corresponds to the Bathymetric Radiator (Figure 18
(d)) but its application to Batavia Seamount involves a different mechanism by which sound is
injected into the seafloor and how it travels to, and up, the seamount.
Figure 19 Mechanisms for exciting water-borne sound waves from the seafloor19. (d) depicts the mechanism that we
postulate is responsible for the sound emitted from the Batavia Seamount. Here, sound from a deep source is focussed to
the top of the seamount where it is emitted into the sound channel as a water-borne signal.
As background to the seafloor propagation of sound from the deep hole at the final location, we
show in Figure 19 the elevation profile of the seafloor along the Penang longitude (north to the left
and south to the right). According to the profile, the seafloor slopes deeper towards the north till it
reaches a channel at the northern base of the Batavia Seamount. The other point to note is that
sound which occurs below about 4500 m depth is focussed towards the deepest point of the hole
through reflection (Figure 20), so that the deepest point can be expected to experience an enhanced
(louder) sound. This deepest point then provides the sound source that is injected into the seafloor
at about 6000 m depth.
19 Williams, C. M., Stephen, R. A., and Smith, D. K. (2006). " Hydroacoustic events located at
the intersection of the Atlantis and Kane Transform Faults with the Mid-Atlantic Ridge,"
Geochem. Geophys. Geosys. 7, doi:10.1029/2005GC001127.
Figure 20 Elevation profile (bottom graph) along the Penang longitude (green line on map) from south to north, right to left
respectively. The postulated deep hole location is marked by a vertical line in the profile together with a depth marking
(5947 m). From right to left (proceeding north), the elevation profile deepens till it reaches the Batavia Seamount which has
a channel at its northern edge.
Figure 21 Mechanism by which sound is injected into the seafloor and travels to the top of the Batavia Seamount to be
scattered into the sound channel as a water-borne sound. Sound at the final location is focussed to the deepest point of the
final location where it is injected into the seafloor. From there it travels to the Batavia Seamount and is reflected and
focussed to the top of the seamount where is then emits as a weak water-borne sound that is heard at the Perth Canyon
and Cape Leeuwin.
As shown in Figure 20, the seafloor sound is reflected off the northern part of the Batavia Seamount
which contains the deep channel. This reflection focusses the sound to the top of the seamount
where it emits as a Bathymetric Radiator. Note also that additional focussing may have occurred
from the triangular plate structure, upon which Batavia Seamount was located at its apex. The
emitted sound is then heard as a weak sound at the Perth Canyon and Cape Leeuwin at the correct
bearing and times.
We postulate that the sound injection to the deepest point at the final location also provides the
focussing and injection of sound into the seafloor which may have been responsible for the seafloor
sound which travels to Diego Garcia via Broken Ridge and Madagascar.
Damage to Aircraft and Debris
The damage to the aircraft deduced from recovered debris is shown in Figure 21. Several
observations come to mind:
1. The damage to the engine and wing parts located close to the fuselage are similar to those
found in a Boeing 777 that crash landed on its fuselage in 2008 (Figure 22).
2. More damage occurs on the right-hand side wing and engine, which possibly indicates that
the right side of the aircraft contacted the water surface first. This could be due to the
aircraft tipping over to the right, or more likely that the right side contacted the crest of a
wave first. Conditions at the site were therefore likely to be rough at the time of the landing.
3. Recovered interior parts from the front and rear indicate that the fuselage was breached at
some stage. This is inconsistent with a similar landing on the Hudson River where the
fuselage remained intact. The other possibilities are an implosion at depth or crashing
against the seafloor. With the front pilot cabin closed and the aircraft sinking rapidly in the
ocean, there is a likelihood of air being trapped in the aft fuselage, in addition to the locked
front cabin. At some point, and perhaps separately, these two fuselage areas could implode
at shallow depths (no more than 100’s meters) and eject interior parts from the aircraft.
Then, there are the high-pressure vessels which could implode at much greater depths – and
hence, more likely to be the source of the seafloor sound, as well as any material which may
have been carried to the surface with the imploded air bubble cloud. This material in turn
may have been responsible for the observed MODIS anomalies—described in the next
section.
Figure 22 Items recovered from MH370 and their approximate locations, whilst some general parts may have come from a
different location than shown. Source: https://www.atsb.gov.au/media/5773565/operational-search-for-
mh370_final_3oct2017.pdf
Further support for the possibility of a glided landing is provided by the lack of severe damage
expected of wing parts in a catastrophic high-speed dive impact. This was noted in several news
articles and comments by experts. For example, here is an excerpt taken from:
https://www.news.com.au/travel/travel-updates/incidents/malasian-news-agency-speculates-mh370-floated-for-awhile-
after-coming-down-in-indian-ocean/news-story/8342d7542e5e63f61ebcd5932032f18a
Aircraft ‘floated for a while’
Malaysia’s national news agency cites the lack of a debris field after MH370’s disappearance and the
lack of crushing to the flaperon recovered late last month are evidence the aircraft was largely intact.
“I believe that when the aircraft went out of fuel, it glided downwards and landed on the water with
a soft impact ... that’s why I believe the plane is still largely intact,” Zaaim Redha, who was involved
in analysis of the missing airliners satellite-sourced flight data immediately after the incident in
March last year.
“It (the flaperon) was only slightly damaged and was just encrusted with barnacles. Its appearance
indicates that it was not violently torn off from the aircraft’s main body ... it does seem that it got
detached pretty nicely at its edges,” he said.
Figure 23 Damage to the undercarriage area, the engines and flaps near the fuselage from a crash landing in 2008. Source:
http://www.cargolaw.com/2008nightmare_b-777.html
To conclude, the damage to MH370 could plausibly, and most likely, have come from a glided
landing, followed by implosion(s) at depth that dislodged internal parts (and possibly other external
coverings) from the front and rear of the aircraft, and deep implosions that were the source of the
seafloor sound.
MODIS Anomaly
Out of curiosity, I investigated anomalies in images from the MODIS satellite data
("NASA/OCEANDATA/MODIS-Aqua/L3SMI"). This is a quality-controlled level 3 product of ocean
biology parameters that includes chlorophyll_a, which is a measure of phytoplankton biomass in the
oceans. We used the band “chlor_a” at a spatial resolution of 500 meters. Two types of changes to
measured chlorophyl are envisaged: dissolved substances affecting water colour and productivity,
and debris causing surface reflection. Clearly, any debris will be much less than the pixel scale (500
m), but the highly reflective nature of the debris, plus any effects from dissolved materials or slicks,
will be either classified as cloud especially during the winter season when the relative anomalies will
be large, or else very high chlorophyll during the peak season. So, any anomaly will be at the pixel
scale, or multiple pixels if between pixels (or the reflection is non-unidirectional and specular). These
will either get masked out as cloud or a high chlorophyll value. Alternately, a debris/material patch
which is the likely scenario will subsequently spread beyond the pixel scale to affect neighbouring
pixels. Sequential connections of anomalies will be associated with drift of the debris/materials and
the spatial scale will be associated with the drift velocity and timing between successive clear
monitoring of the debris/materials sequential locations. Our expectation is that anomalies will be
compelled by the Broken Ridge to flow east, and north, up till at least 103oE longitude before
starting to take on a more northerly component, and then curving north west towards Africa.
David Griffin’s best simulation20 shows that despite a general overall trend, local drift paths are
highly convoluted and appear to bifurcate at the location of the deep hole (Figure 23) leaving an
area largely devoid of drifters. We postulate that this feature of the deep hole is yet another
manifestation of geostrophy at play where the rapid deepening as currents try to cross the hole
results in geostrophic forces that act across the current. Forces will be relatively stronger for more
rapid deepening as for example trying to cross the hole along its main (horizontal) axes, but stronger
currents and density stratification may oppose geostrophy. However, once within the deep hole, the
opposite may be true (for MH370 debris for example) as circulation around the perimeter of the
hole will entrap and possibly connect any debris and materials till they ultimately escape with a
change of currents, or through random interactions. The simulated drifters are a good
demonstration of drifters from the outside trying to cross the hole, but they may not be
representative of debris located within the deep hole region getting out.
Figure 24 Cropped image of Dr David Griffin’s best simulation for drifters arriving at La Reunion Island. Red arrows are the
respective latitude (horizontal) and longitude (vertical) of the deep hole at the Penang location. Note the relative paucity of
drifter tracks that cross over this deep hole.
To assess anomalies, we accessed individual satellite images and processed them using Google Earth
Engine Code (GEE)21. We setup 3 rectangular areas (2 degrees by 2 degrees) with the Penang
location rectangle area in the middle, one area to the west and one to the east. These areas were
rectangles denoted as (llLon,llLat,urLon,urLat), where llLon is lower-left Longitude, llLat is lower-left
Latitude, urLon is upper-right Longitude and urLat is upper-right Latitude:
West: (97,-33.5,99,-31.5)
Mid (Penang Location area): (99,-33.5,101,-31.5)
East: (101,-33.5,103,-31.5)
20 https://ecos.csiro.au/mh370/
21 https://code.earthengine.google.com/
Our expectation is that anomalies are most likely in the Mid and East areas (on the assumption that
any debris will strongly reflect radiation which will be detected as high/anomalous chlorophyl), with
the West area setup more as a control location with respect to identifying anomalies. A range of
dates were used to assess the anomalies and we started off with a broad assessment by computing
the maximum chlorophyl values within the test areas for individual images across the years: 2013,
2014 and 2015 as shown in Figure 23.
Figure 25 Time-series of maximum chlorophyll within the 3 areas: West, Mid and East for 2013 (upper plot), 2014 (middle
plot) and 2015 (lower plot). Note the extreme anomaly monitored in the East area towards the end of April 2014 – which is
the subject of this analysis.
An extreme outlier observed in the East area on the 29th April 2014 warranted further examination.
We first mapped chlorophyll_a over a broad region to provide context for the anomaly as shown in
Figure 24. The high chlorophyll band in the Southern Ocean is highly structured and contains high
chlorophyll finer structures in the west that gradually progress north east with a more mixed
structure that stops progressing northward at the end and south of the Broken Ridge structure.
There is no evidence of high anomalies progressing into the Penang location area (shown as the
transparent blue rectangle).
Figure 26 Broad distribution of the maximum chlorophyll_a mosaiced over March-April 2014 inclusive using the maximum
value observed at each pixel location. Note the extensive band of high chlorophyll structures in the Southern Ocean which
heads north east and becomes less structured and weaker by the time it passes by the Penang location area highlighted by
the blue rectangular area (Mid). Blue pixels are areas affected by cloud over the period of the mosaic.
A zoomed in view (Figure 25) shows the anomalous spot located 183 km to the northwest of the
Penang location site, and this occurs on the 29th April 2014 as shown by the time-series in Figure 26.
Assuming this spot is from an MH370 debris as it seems likely to be, this gives an average drift speed
of 3.52 km/day (straight line distance from the landing location), which is slower than the average
speed of over 12 km/day required to reach La Reunion Island by the 29th July 2015. The curvature of
the chlorophyll patterns around this site, and David Griffin’s simulated tracks, suggest that
recirculation drift processes may be responsible for the slower average path speeds as the debris
would be drifting around before reaching a location. The zoomed in image also shows several
isolated but lesser intensity anomalies in the Mid area, to the north of the landing location. These
observations taken together reinforce this location as the site from which MH370 debris may have
commenced its drift.
Figure 27 Zoomed in and enhanced image of Figure 24. The high anomaly is to the east and towards the north of the
Penang location area, seen as high intensity single pixel spot. A range of lesser anomalies and pixels marked as cloud are
seen in the Penang location area.
The lack of additional anomalies during April 2014 seems to be due to chlorophyl values being low
and the area was affected by clouds (Figure 26), so the observation of the high anomaly was
fortuitous.
Figure 28 Time-series of maximum chlorophyll values across the 3 areas for April 2014, where Site refers to the Mid area in
the main text.
In an effort to track this anomaly, a further test area was setup to the east of the East area, and the
time-series of maximum chlorophyl was plotted for May and June 2014 (Figure 27) which showed a
very high anomaly in the eastern most area on the 19th June 2014. We assume that this anomaly is
the same one monitored in April 2014.
Figure 29 Time-series of maximum chlorophyll values across 2 areas for April 2014, where “Site” refers to the East area and
“East” refers to an area east of the East area referred to in the main text. A very high anomaly appears on the 19th June
2014 in the eastern most area.
The map of the anomaly shows two anomalies responsible for the high anomaly observed in the
time-series on the 19th June 2014 (Figure 28). The computed straight line drift speed from the April
2014 anomaly was 4.2 km/hr which indicates a slight speed increase. The location of the northern
spot at -31.8oS and 103.9oE, and its slight northward path of the April location indicates that it is
reaching the turning point where it will drift further north and then later curve north west towards
Africa.
Figure 30 Location of two very high chlorophyl anomalies monitored on the 19 June 2014 to the east of the East area
(transparent green area). The connected nature of the anomalies indicates possible location of the northern anomaly at the
intersection of (or reflection spreading across) 3 adjoining pixels and likewise for two adjoining pixels in the southern
anomaly. We assume that the northern anomaly is the same debris that was observed in April 2014.
Unfortunately, we are unable to track the debris any further but given its location and timing other
satellite images can now be examined to further investigate these anomalies and to track their drift.
One possible scenario that may explain why the anomaly could not be further tracked is that the first
anomaly was a concentrated patch that subsequently split into two and spread further. Idealized
diffusion of circular patches suggests that the area over which debris/materials spreads goes like the
square of time. With approximately equal time differences between the first and second anomalies,
debris area will be expected to quadruple at the second sighting. We cannot apply the diffusion
analogy to the anomaly strength (decrease to one-quarter) however as it may not involve dilution
(of substance/debris concentration). The actual observation of spread from one to 5 pixels, and into
two patches suggests that the actual dispersion is far more complicated than this ideal model—if
indeed the initial debris was a patch splitting into two. However, we now have a possible strong
new source of direct evidence to investigate using higher-resolution satellite imagery. This would
be our best new direct lead, if indeed correct.
Figure 31 Location of the two anomalous high MODIS chlorophyl spots (April Bright Spot and June Bright Spot) in relation to
the final location for MH370 (red circle) and the expected general drift path from the drifter simulations (cyan curve).
Overall, the timing, location, drift speeds and directions (Figure 29) are consistent with these
anomalies being due to MH370 and appears to provide further confirmatory evidence of its final
location.
Final Message:
This ends my attempt to unravel the MH370 mystery by evaluating all plausible evidence against our
postulated location. Many thanks to the folks at Google Inc. as Google Earth Pro has been a constant
and faithful companion. The impetus for continuing the research has always been the realisation
that if the location was right, or approximately right, the mystery would unravel. I’ve gone through a
systematic iterative process with the scenarios and evidence which unearthed and connected what
were disparate/abandoned, and in the end valuable, pieces of possible evidence. As always, one can
never be sure but the scenario I have put forward remarkably integrates and intersects all viable
available evidence to the final location, which to date no other theory has succeeded in doing.
As stated in the Long Summary, I would expect any search to also be followed by a biogeographic
survey of this iconic ancient deep-water region.
Best wishes to all in the final search.
References
Kadri, U. (2019). "Effect of sea-bottom elasticity on the propagation of acoustic-gravity waves from
impacting objects." Sci Rep 9(1): 912.
