MH370 Debris Drift Analysis by Brock McEwen

What does the fact that a piece of MH370 washed up on Reunion Island tell us about the plane’s likely impact point? For the last few months independent researcher Brock McEwen has been hounding the ocean drift modelers of the world in an effort to shed light on that question. After carefully analyzing the results of nine drift models–CSIRO, UWA, GEMS, Deltares, GEOMAR, ICMAT, Adrift, IPRC, and OM–Brock, a trained statistician, concludes that taken together they suggest:

  • Without implausibly strong wind effects debris could not have reached Reunion Island from the current search area.
  • Before debris could have reached Reunion Island, other pieces should have washed up in Western Australia and on other shorelines in the Indian Ocean.

These points undermine the claim put forward in the most recent ATSB report that “the location of the recovered debris is consistent with drift modelling predictions of objects starting in the areas identified as possibly containing MH370.”

The entire report can be found here.

I hope that officials investigating the crash, whether they be in Australia, Malaysia, France, or elsewhere, take heed of Brock’s significant contribution. The evidence has long been mounting that the authorities are looking for the plane in the wrong place.

32 thoughts on “MH370 Debris Drift Analysis by Brock McEwen”

  1. Brocks work is bang on.

    The counter SIO POI data is irrefutable.

    A day of reckoning is approaching.

    The day the SIO MH37O search aborts.

    Until such day…

    The SIO POI delusional fantasy prevails.

    After such day…

    The world will listen to men like Brock

    Andre Milne
    Unicorn Aerospace

  2. Hi Jeff, Wouldn’t it be interesting if someone included the segment of flight surface washed up at Kudahuvadhoo in their drift analysis?
    With two such widely-spaced pieces of debris you could presumably narrow the likely search area down markedly.
    Cheers, Pat

  3. Sorry, I meant the flight surface segment washed ashore at Vabbinfaru not Kudahuvadhoo.

  4. Hi Jeff, If you look at the Vabbinfaru debris you’ll see that it is made of a nomex honeycomb with carbon/graphite sheets, a common type of construction on the B777.
    Can you come up with plausible reason for dismissing it or is it just your belief?
    Cheers, Pat

  5. Thanks, good effort Jeff on promoting this(I already thanked Brock for his great analysis) however I’m afraid officials will reject anything that doesn’t comply with their silly theory.

    The problem is they have cornered themselves from the beginining restricting the search to only one area around the southern 7th arc instead of leaving possibility for others. Admitting they were wrong isn’t going to be easy for them.

  6. Thanks to all who have offered feedback and encouragement – appreciated.

    Off topic (sort of):

    I’ve also updated my “end of flight” model to replace “3:40” with “2:00” as the manufacturer’s suggested reboot time (I repeat my question: why was the wrong time used to limit the search zone for 14 months?). Note that the ripple effects of this change include

    1) the time during which flight must run on one engine
    2) assumed coordinate of 2nd engine fuel exhaustion

    If the (airlandsean flight sim-calibrated) one-engine dynamics are maintained, the model hits a conundrum: straight and level flight – whilst decelerating at the indicated -0.315 knots/sec – leaves the plane generally unable/unlikely to bear left soon/severely enough to actually reach Arc7.

    My first inclination, quite frankly, is not to bother incorporating this new reboot time, as it I attach to it the same veracity I attached (rightly, it turns out) to the GEMS drift study a year ago.

    But I am happy to humour all those who still think the Oz search is authentic, and refine my model to maximize insight. My options:

    A) reduce altitude while on one engine: when the difference in time between left and right engine flameout was only 4:50, airlandseaman’s report that speed dropped as above without altitude change was sustainable; but if this period is extended a further 100 seconds, speed drops below what my untrained eye deems an implausibly low speed. To me, the most natural remedy is to allow altitude to drop, as the slowing plane simply sinks. This improves Arc7 attainability by bringing it closer (effect of lower altitude), albeit slightly.

    B) have plane slow down les dramatically: -0.315 sounded like a pretty stable result of the flight sim trials, but I’ve reached out to the IG to see if they have either a better average number, or an opinion on how this deceleration rate might change as the one-engine leg enters its 6th and 7th minutes

    C) increase initial speed: here again, I’d thought the IG was telling us 14 months ago that >460 KGS was pushing the endurance envelope, but am happy to hear all thoughts.

    D) add a more precipitous drop AFTER engine 2 flameout; I think this is Dr. Ulich’s preferred response; like A), this brings Arc7 closer. I’d assume horizontal speed shed due to increased bearing below horizon is replenished by increased overall speed due to effects of gravity, but welcome all thoughts.

    Thanks in advance for all feedback.

  7. Re: D) above: “precipitous drop” was “in altitude” – which may not have been clear from context.

  8. Brock,

    What does your end-of-flight model predict if you assume terminus of the CTS-model instead of the AP?

  9. After 15 minutes of flight with one engine inoperative the airplane probably has arrived at its single engine ceiling of 29500 ft at the Flap0 Maneuver Speed of 206 kt IAS or perhaps a few knots less.

  10. @VictorI
    Indeed, the surfboard explanation is the one you’ll find mentioned on social media but it doesn’t hold water. The Vabbinfaru segment is made from nomex, you can tell by the fine red streamers on the fractured face which is the phenolic resin used as a coating. The only company in the world that makes a surfboard anything like it is Varial that you refer to. Except if you look at their design closely you’ll see that the honeycomb doesn’t reach to the edge and this area is filled with foam. This is necessary because otherwise the board wouldn’t have enough impact resistance. In the Vabbinfaru debris, the honeycomb reaches right to the leading edge. My assessment of the Vabbinfaru debris is that it is a segment of the outboard flap of a B777, in particular, the outboard end that protrudes past the hanger.
    Cheers, Pat

  11. @Pat: not endorsing, but neither am I rejecting, your theory. Just exploring:

    Have you found imagery matching the partial lettering on the Vabbinfaru object to lettering found on an outboard flap?

    I suppose the same challenge should go out to the “surfer dudes”: any imagery matching the lettering to a surfboard logo?

  12. @Gysbreght: thanks – very helpful.

    1) The internets tell me 206 KIAS at FL(295) & 0.02 OAS correction = 325.480 KTAS – is this best practice conversion?

    2) Would you model roughly LINEAR deceleration throughout this stage?

    Thanks again.

  13. @Brock, Hi, wrt the imagery, there are surprisingly few images of the underside of a B777 at sufficient resolution/quality to do this.

    The writing on the underside of the Vabbinfaru object clearly finishes in “IC” and there is enough of the previous letter to indicate “TIC”, maybe “FIC”. My take on this is that it was a warning to ground staff to beware of static. The underside of composite flight surfaces doesn’t normally incorporate the metal gauze found in the top layer for lightning strike protection and hence is prone to building up electrical charge.

    If you look at one of the pictures of the object you’ll see what looks like a splintered black stick. This is the carbon fibre reinforcement that runs along the thin trailing edge of the flight surface. Nothing like it in a surfboard.


  14. @Brock:

    1. 206 kIAS, FL295, IAS equates to 325.3 kTAS. Not sure what ‘OAS’ means.
    2. If the Autopilot/autothrust was engaged in a typical cruise mode, i.e. AP=Alt Hold and AT= Speed Hold, then after engine failure the airplane would first decelerate at constant altitude to maneuver speed, then descend at constant IAS. The rate of deceleration at constant altitude would depend on the altitude.

  15. @Pat Janssen: I am in @Brock’s camp relative to the debris found. Despite my suspicions, I have no idea what the object is, and would hope that a positive identification could be made. Certainly the authorities at this point know exactly what it is. Relative to the lettering, the partial letter looks like it is curved on top and not the letter T. Also, the font looks more like a logo than the “block” lettering you would expect for a “STATIC” warning.

  16. @Victor, @Pat: is this the lettering in question? I just grabbed what I thought we were talking about from a Google image search – I make no claim of authenticity.

    If so: @Victor, if I ever need someone to design a sleek, sporty logo for me, I may have to look elsewhere…

    Also not sure what you mean by “curved on top”.

    Profuse apologies if I am peddling some doctored version inadvertently.

    But I’m still still on the fence: three or four surfboard logos end in either FIC or TIC. After that, I suppose it’s on to windsurf, knee, & paddle boards. Imagine it’s an easy cross-reference against the logo & core designs of each.

  17. @Brock, @Victor

    The best quality pictures are on the facebook page of Mohamed Wafir. You’ll need to scroll down to August 6th:

    Alternatively, just Google “Vabbinfaru MH370” and select images but some of the images have been cropped or downsampled.

    The vast majority of surfboards and the like just use foam inside. The top model Varial is the only one I’ve been able to find with a nomex honeycomb, not surprising considering the cost.

    Cheers, Pat

  18. Pat Janssen – As someone who has owned many surfboards it would be a pretty unusual material to use in one. They can’t be made of cavities as they end up full of water eventually. In a limited way on a plank(Mal) to cut down some weight would make sense. But as you say it needs a good layer of the more traditional polystyrene/polyurethane foam surrounding it. Not common at all.

  19. I was asked to update my stochastic “end-of-flight” model to re-assess the probability wreckage from an unpiloted spiral to impact lies beyond the area already searched, in light of

    1) increased search area since last May, and
    2) reduced reboot time (per Dec. 3 ATSB report)

    As expected, both serve to further decrease – below the 1% reported in May – the likelihood of finding debris yet further afield.

    Thanks go out to Gysbreght for the suggested altitude process (above), and to Alexander Rabinowitz for model vetting he provided in August.

    I won’t quote precise numbers, because I’m still trying to pry expert advice out of the IG, but I’ll share what, under current parameterization, is the “money shot”:

    I have not plotted the Go Phoenix area, but it is almost as wide as the NE zone I do portray – if the model is well-calibrated, the Go area even further NE is likewise searched out.

    Please don’t anyone misunderstand me: if wreckage is found where they are currently searching, I will celebrate the possibility of closure for next of kin. But I would, at that point, feel compelled to ramp UP my efforts to disseminate my research, because of what it says about such a discovery’s a priori probability.

  20. Sorry – title in image should read:

    “Distribution of DISTANCE from impact point to 7th arc (FL0), in nmi”

  21. I think Brock has done some sound work in establishing the shonkiness of the current search area. However, at the end of the day, drift analysis based on a single object found so long after the event has its limitations. Isn’t it lucky then that we actually have 3 objects from MH370. Here’s a wee analysis of my own:

    The 1st object was the fire suppression bottle washed up at Baarah a couple of weeks after the loss. As it was empty, it caught a lot of windage. The south-eastern trade winds moved it at a good clip with the sea surface currents nudging it a bit to the east resulting in a roughly south to north resultant path. Thus working backwards the plane came down somewhere along the southern portion of the Laccadive ridge.

    The Vabbinfaru flight surface fragment and the flaperon are a bit trickier. At the time the plane came down, the sea currents in the vicinity of the Laccadive ridge were still running from west to east carrying the debris field with them although the field would having been getting enlarged as it went. The flaperon ended up in the southern extremity of the field. After a month or two the current direction reversed, now being east to west for the rest of the year. The southern part of the field including the flaperon was caught up with the southern equatorial current heading towards La Reunion.

    The northern part of the field including the Vabbinfaru object was driven back towards the Laccadive ridge. To get the timing right, the object passed through one of the channels in the ridge to the western side. It wasn’t until the currents changed back to going from west to east that it was driven back to the ridge and eventually finding land at Vabbinfaru.

    Cheers, Pat

  22. @Pat
    Do you have any hard evidence, that the fire bottle and the assumed flight surface fragment are from a 777?
    Officially these parts have never been confirmed by official authorities as belonging to MH370 or even to a Boeing 777.

  23. @RetiredF4

    He doesn’t have any of course and I can’t understand fascination with those objects but to each his own.

  24. In Figure 3 on page 5 of Brock’s PDF report is a CSIRO chart showing drift lines of actual drift buoys over 15 months preceding.

    Buoys from the 7th Arc seabed search area are depicted in red, whilst buoys outside that seabed search area are depicted in white.

    It is noteworthy that where satellites in March 2014 spotted 122 floating objects bracketed by a 24m long object to the east and a 22.5m long object to the west are in an area where white drift lines did reach Reunion.

    As I understand it Brock refutes the possibility of debris from south of the seabed search area, because he argues those debris would have washed ashore on the coast of West Australia.

    That assumption is wrong and Brock’s study does not appear to consider the Leeuwin counter current which travels south from the Timor Sea pushing any such debris offshore.

    The Reunion Flaperon was presumably not characteristic of other debris from MH370. Due to its closed cell honeycomb sandwich the Flaperon had what Mike Exner calculated as 770lb of positive buoyancy. This means it was likely more subject to wind currents than sea currents and different rates of drift. I do not see where Brock has considered this.

    Why would the flaperon be the only closed cell honeycomb sandwich component to reach Reunion?

    Possibly because others remained attached to their respective wings and sank with those wings when they finally did sink. The Flaperon may be almost unique in having detached from the right wing several weeks after impact due to wave action.

    Brock is a mathematician yet he does not seem to have modeled these other possibilities. When one omits to consider all possibilities in a mathematical model the errors grow exponentially.

    Likewise electrical failure of both Right and Left Main AC relays means MH370 lost AIMS navigational processing.

    MH370 could not have navigated from Penang to VAMPI then turned slightly to intercept MEKAR before 18:25 without AIMS. No Main AC Relays = No AIMS = no navigation. If electrical failure invalidates the Straits detour then MH370 had the range to fly further south where debris were spotted.

    Jeff Wise speaks of how the satellite data must have been spoofed. Electrical failure would spoof the ACARS data by altering BTO transmission bias.

  25. I agree with Brock that the search area should be further north, but in my analysis further west also, to keep any debris from getting to the NW corner of Australia. I would also like to see prevailing winds analysis of the suspected area(s, just in case this was a devious event…..the devious pilot would want to enter the water at the slowest possible airspeed, under control, which might explain a deflected flaperon being ripped off.

  26. I absolutely believe that the Mh 370 came down on or close to Madagascar ,This is what I conveyed to the authorities days after the incident ,The finding of recent pieces of the plane locally should determine a search of the waters around there ,My interest is solely for the relatives closure,Ihave conveyed my disappointment tothe ATSB that they won’t admit a mistake and change tack ,Truth is more important than face- saving ,

  27. @ Jef
    I have just watched the BA38 777 crash video at Heathrow airport by National Geographic. At 15:26 you can see the words FUEL STICK are painted next to the Fuel tank covers of the “dip sticks ” .
    Could explain where the letters TIC are from on the debris found perhaps ?

  28. Now the ATSB is finally considering that a live person possibly turned the aircraft (into the wind) to attempt a successful water landing (think Hudson River) which is why they aren’t coming up with results from their theoretical straight glide/crash scenario. MH 370 is within a 15 mile radius of the last (partial) ping, and the heading will be in the direction of the prevailing wind on that day.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.