Questions and Answers

In preparation of my 2024 keynote presentations and webinars, I started an e-mail address dedicated to answering the questions on the MH370 case: MH370questions@gmail.com

This sections includes the questions and my answers and it is continuously updated as new questions and answers arrive. In preparation of the webinar, we requested questions from participants. The questions are listed in the chronologic order of the registration messages, but some questions were grouped together if they have a common answer.

Q: What factors do you believe contributed most to the mysterious disappearance of Malaysian Airlines Flight 370?

The crucial factor in my opinion is the will of an individual (possibly with accomplices) to carefully plan and execute this diabolical hijacking and mass murder. The aviation system is unprotected in case a person with criminal intentions takes control over the cockpit.

Q: Which is the explanation of this groping at the planetary scale, which continues?

I have a general explanation at the society level, and a specific explanation for MH370.

We live in a new world where any individual can make his or her voice heard over the social media. This proliferates all sorts of theories and ideas, obscuring or even obliterating real and relevant scientific research. A similar havoc happened in the Covid pandemic. When everybody is an expert, nobody bothers to search for the authentic experts, and nobody really values their opinions. The opinions of the experts are not really interesting, they might be boring and difficult to understand. The most clicks are attracted by conspiracy theories and weird ideas. The values of our digital society are established by the number of clicks.

Specifically in the MH370 case, the huge pressure of the public opinion pushed the decision makers in the search operations to act first and think afterwards. Lots of opportunities to search in plausible places were sacrificed to the rush to search everywhere else. I remind you that MH370 is one of the most horrible crimes in the history of aviation, carefully planned to hide the evidence. I believe it is one of the most difficult problems that mankind is trying to solve.

Q: In the case of the pilot not reading back the frequency, shouldn’t the ATCO firmly request a read back?

and

Q: Wasn’t this lapse/violation on the part of the pilot (not reading the frequency back) one important clue of suspicious behaviour that the ATC should have taken into consideration on the spot? Can this ever be regarded as acceptable coming from a very experienced pilot?

Readback – hearback is the backbone of the ATC safety. However, the frequency of the next sector is not on the ICAO Doc 4444 PANS-ATM mandatory readback list:

“The flight crew shall read back to the air traffic controller safety-related parts of ATC clearances and instructions which are transmitted by voice. The following items shall always be read back:

• ATC route clearances;
• clearances and instructions to enter, land on, take off from, hold short of, cross, taxi and backtrack on any runway;
• runway-in-use, altimeter settings, SSR codes, level instructions, heading and speed instructions and transition levels.”

Q: Is there any chance of finding this flight ever?

Nobody can rule out a chance. A discussion of probability is another matter. The probability is very low, and the analogy with the AFR 447 case in 2009 may be useful to understand why: in 2009 the impact point was approximately known because the tail was found afloat after 48 hours. The Atlantic Ocean’s bottom shape is known. In spite of all these, it took 2 years to find the wreckage of the aircraft on the bottom at 5000 m depth.

Q: After this disappearance and the investigation, are organisations better prepared to deal with a similar situation?

and

Q: Could a similar disappearance happen tomorrow?  If not, what are the critical changes which have made a repeat unlikely?

I believe that such a disappearance was not possible in that part of the world which values aeronautical safety and security the most: US, Canada, and Europe. Here, an aircraft which loses radio and radar contact with ATC is immediately intercepted by military jets. This was the rule before MH370 and it is the rule now. In this so-called collective West, organisations learned their lessons, reinforcing the importance of air police, and maximising the measures to prevent the access in the cockpit of individuals with bad intentions. I am not sure about the rest of the world. To me, what happened in Malaysia was baffling, having an unidentified ariplane flying undisturbed for one hour in the national airspace. I do not know if the Malaysian authorities learned their lesson.

Better preparation also requires new technology. We already have the Global Satellite Tracking System operational. For the moment, it is based on ADS/B, so it uses the Latitude and Longitude as reported by the aircraft transponder. I hope this will evolve to satellite multilateration, to back up the dependent surveillance by an independent surveillance technology. The newest communication technologies combine communications with navigation, so that will help. In terms of black boxes technology, more steps need to be made, like the ejectable floatable data cartridge instead of a black box which sinks with the aircraft in case of water impact.

Q: The theory of a terrorist attack is void? Are there any proof about the terrorists and what organisation are they from (or country)?

The mass murder crime can be considered a terrorist attack in a sense, because it spread terror on the whole planet. But the motivation is not evident. In other terrorist attacks, the motivation is clear.

There were two Iranian passengers who managed to embark using stolen passports, but the investigation concluded that they were not terrorists, just illegal migrants. A terrorist attack is  claimed in most cases, like for example the one we recently saw in Russia. Terrorist attacks serve a purpose of a terrorist organisation. The lack of any claim over MH370 in 10 years leaves this theory without support.

Malaysia is a Muslim country, so hardly in the target of jihadi terrorist organisations.

On the other hand, several months later we had the case MH17, and that was a terrorist attack by Russian forces from the separatist region of Ukraine. It was hardly a coincidence the choice of another B777 of Malaysian Airlines. The Netflix documentary on MH370 presents this coincidence as evidence that the Russians were involved in both cases. This logic is bad: while it is hard to deny that the choice of the MH17 as a victim flight was intentional, to maximise the mystery and the impact on the public, it has no retrospective consequence on the previous choice of the MH370.

It is also true that MH370 happened at the same time with the annexation of Crimea by Russia, diverting the public attention from it. There is no proof of the Russian involvement in MH370. The captain is certainly involved in the hijacking and as far as I know the investigations did not reveal any connection with a terrorist organisation or with a foreign state secret service.

Q: What advancements or changes in tracking technology have been implemented since the MH370 disappearance to prevent similar accidents?

ICAO mandated NAV Canada and Aireon to establish a Global Satellite Flight Tracking service, covering the globe, including the vast oceanic areas without ground antennas. The satellites get ADS/B messages from aircraft in flight and locate them. However, in the MH370 case, the ADS/B transponder was deactivated, so the new system would be useless.

In order to prevent similar acts (I would not call accident a premeditated and carefully planned crime), the Aireon network of satellites should move two steps further:

(1) Start to do satellite multilateration to confirm the ADS/B position as reported by the aircraft. That enhancement is in progress from what I understand from NAV Canada sources. This will address mainly the GNSS jamming and spoofing problems which impact the ADS/B, but it still leaves a case such as MH370 undetected.

(2) Extend satellite multilateration on more frequencies, capturing and using any radio transmission from the aircraft. For this step, a new generation of satellite may be needed.

There is also another idea called the Autonomous Distress Tracking (ADT) initiative, which could not have prevented the disappearance of MH370 either, because MH370 was never in distress, on the contrary. Both these ideas (Global ATS Surveillance and ADT) come from the perspective of aviation safety, which is about preventing accidents. MH370 is no accident. It is clearly a well planned hijacking case by someone who was inside the cockpit. All the actions of this perpetrator were in line with hiding the presence of the aircraft and the wreckage in a place no one would ever find it. Such unlawful interference can be stopped by security mitigation measures, not by safety mitigation. In the case of ADT, an emergency radio locator will start to transmit “the location of an accident site within a 6 NM radius. It uses on board systems to broadcast aircraft position” [8]. The problem is that radio does not get through water and this only helps when the aircraft impacts the ground. We already have a system in place for these situations, the ELT sending distress radio messages to COSPAS/SARSAT satellites on 406 MHz.

Q: How has the disappearance of MH370 impacted aviation safety and regulations?

As a unlawful interference, the MH370 case is not a safety case, but an aviation security case. In that respect, it was reinforced the obligation of the military aviation to ensure air policing for aircraft flying without transponders and radio communications. Also, the importance of thorough passport security checks was reinforced. The presence in the cockpit of persons with evil intentions was reiterated, and the cockpit security door problem was once again questioned – it may prevent people to enter the cockpit and fight the perpetrator inside, such as in the GWI 2985 case. Most security measures are not disclosed for obvious reasons, so this answer is just an educated guess.

Aviation safety learned some lessons though. Firstly ICAO mandated NAV Canada and Aireon to establish a Flight Satellite Tracking service, covering the globe, including the vast oceanic areas without ground antennas. The satellites get ADS/B messages from aircraft in flight and locate them. However, this measure in a safety one, since in the MH370 case, the ADS/B transponder was deactivated.

Other consequential safety measure is the Underwater Locator Beacon (ULB) requirement of battery life operation increased 3 times (from 30 to 90 days) as from 1 July 2014. This would not have helped though, because it took 7 months to arrive with acoustic sensors to the plausible impact area.

More consequential measures have been taken in aeronautical security: emphasis on mental checks and security checks on pilots, international data cooperation and data sharing (that is a two-blade sword in a world with state terrorist actors), passenger screening and security enhancements at airports. In the Annex I also provided the answer of CharGPT 4 to this question.

Q: What could we do to find the airplane?

In 2018 the search operations were terminated. Their total cost was €120M (AUD 198 mil). Most of it was wasted by searching in unplausible places. In 2017 ATSB published a heat map [B3] which coincides with our area published in 2015 [A3]. However, the actual search operations did not really followed that heat map. In conclusion, what we could do is to find the money to restart the search.

Q: With today’s technology (more data, more powerful computers as compared to 10 years ago, the emergence of AI) can we find more clues or can we do better calculations?

The data on the MH370 case is not increasing in time, with the exception of the debris being discovered. More powerful computers would not matter because our research [A1], [A3] took weeks of computing, and now we could hope to get the same results in days, but they are the same results. Maybe, it is worth repeating the calculations with accurate primary radar tracks that the Royal Malaysian Air Force did not provide at the time, neither to us, nor to the INMARSAT team. We only had a map and a description of the vertical flight profile. To our research, the vertical profile of the flight matters because our method is based on BTO of INMARSAT signals validated by the flight simulation to calculate where the fuel was exhausted. The vertical profile could change the quantity of fuel on board after the manoeuvre. For the INMARSAT team’s method based on BFO validated by BTO, this vertical profile does not matter.

In my opinion, AI is not really useful directly because the MH370 case is unique in many respects in the aviation history. For the AI we would need massive data to train the neural network.

Q: Do we need to try to calculate the trajectory based on different assumptions?

In our research published in July 2015 [A3] we started from making all imaginable assumptions, and narrowed down the range of scenarios based on what the INMARSAT signals were telling us. Thus, a lot of assumptions were discarded, including that one of multiple turns of the aircraft during those 6 hours of flight. This assumption has been recently resurfaced by the WSPR theorists, who claim that the weak signals indicated to them these turns. However, by looking at the BFO diagram one could easily discard once again such a hypothesis. So my answer is yes, we should try based on different assumptions if someone can provide them, but as far as I know, all the imaginable assumptions have already been considered, even the weirdest ones.

Q: Can we calculate the probability of such an event to happen again?

There is a very low probability, because the surprise factor now has disappeared. After 9/11 there was not another similar attempt. Also I rely on the lesson learned that any civil aircraft without an active transponder and in the absence of radio communications will be promptly intercepted by the military jets.

Q: Can this accident be attributed to human error. If so, what would be the details that would lead in this direction?

This is no accident, it is a meticulously prepared intentional act. The intention of the perpetrator was most probably to hide the evidence of the crime in a remote, uncirculated place. We can talk about human error when it comes to this act short of the perfect crime. The perpetrator disabled the ACARS, but failed to disable the downstream INMARSAT SATCOM tranceiver, which remained in contact with the satellite for 6 more hours. Also I consider human error of the Malaysian ATC by delaying the ALERFA and DETRESFA phases and also the lack of the military response to the presence of an unidentified aircraft in their airspace. I can provide lots of evidence of human error in the way the search was conducted, by mostly ignoring accredited scientific results.

Q: Did the accident investigation team analyse satellite imagery data to try to identify this a/c and its flight prior to the impact with the sea? B777 has a large fuselage footprint with its length&wingspan of roughly 65 meters. Were there any imagery/geo-observation satellites covering any area crossed by the trajectory of the MH370 flight?

It seems that as many as they are on various orbits, the satellites cover just a very small proportion of the globe at any given moment, and the satellite pictures are not taken continuously, some missions take pictures when there is an objective in the region they overfly. Military satellites change orbits when they need to cover a certain area of interest. However, satellite imagery provides a “static” image of the whole world, and this is assembled from tiny fragments gathered in time. We are very far from an online global image of the world. Another point is that the ALERFA and DETRESFA phases were triggered many hours late, and the search initially was in the South China Sea. I remember that some French satellites were put to work intensively to  gather images of the area, and they found many objects afloat. As for the Southern Indian Ocean, that area came under scrutiny as late as October 2014, 7 months later. I presume the archive images were analysed, but not by us. Why? Firstly because our research was funded by ourselves, and the access to many satellite images is based on a fee. Secondly, I remember doing some research with a student with the identification of aircraft in satellite images and the contrails were grey and hard to differentiate from the background. The body of the aircraft itself was white and more visible, but very small, very hard to identify. Perhaps a neural network trained to discover pixel-sized aircraft in satellite images on a sea background could be employed. I expect nobody erased the archives with satellite images in the 7-8 March 2014 interval, but this endeavour is going to be expensive.

About the orbital mechanics of the satellites which may have taken pictures of MH370, we did not consider it because our focus was to find the most probable trajectories. Unless you have a 4D trajectory of the aircraft, it is not possible to intersect it with the 4D trajectory of a satellite. It took us one year to have certainty on a number of 38 possible trajectories. In the meantime, we found that other entities searched the satellite pictures intensively, so we expected them to use our results to refine their work, because they seemed to have better access to the photos.

Also it is very important to mention that approximately 80% of the flight was protected by the darkness of the Moonless night, when satellites do not take pictures. If our calculated trajectories are right, the sunrise happened just one hour and a half before the crash, so just the satellite pictures in that time frame would be relevant.

Q: In the absence of accurate position data prior to the impact, I am wondering whether data from the sonars of the ships located in the estimated probable locations would have been useful to detect and possibly triangulate the location of the impact of the aircraft with the sea. I am not sure if sea vessels are equipped with a black box that would record sonar data. If yes, probably the data has been overwritten before the investigators could have even accessed it.

No, normal vessels do not possess underwater acoustic monitoring capability, but your idea is a good one, and a 2022 paper already promotes it [*]. During the search, specially MBES equipped ships were deployed to receive the acoustic signals from the ULB. Unfortunately they were deployed West of Australia, very far from the place where the research converged [1-4], and by the time the search area was moved to the right place (South of Broken Ridge), that was 7 months after the crash. The ULB batteries were supposed to function for 30 days, 45 days the most.

[*] Lowes G J et. al, Passive Acoustic Detection of Vessel Activity by Low-Energy Wireless Sensors, J. Mar. Sci. Eng. 2022, 10(2), 248; https://doi.org/10.3390/jmse10020248

Q: Were the ACARS reports meaningful in establishing the probable areas?

No, because the ACARS was disabled in the same time interval (1721Z-1737Z) as the ADS/B, Mode S SSR XPDR, and VHF radio communications were disabled. Normally a B777 pilot has no instructions on how to disable ACARS, but apparently in this case the circuit breaker was cut off. At 1708Z the last ACARS report was received by Rolls Royce on the functioning of the engines. The next report was expected at 1737Z (30 minutes later) but never came.

Q: Why passengers cell phones continued to operate after flight’s disappearance?

Probably because the flight returned to overfly Malaysia in its 6 hours flight to the South Indian Ocean, after the disappearance of the ATC radars in the South China Sea. Probably the passengers themselves were already dead due to hypoxia.

Q: Was enough the overall action taken by ATC since it was stated that the aircraft disappeared? Are any failures considered?

One failure is the major delays in triggering the ALERFA and DETESFA phases. In 2018, the head of the Malaysian Civil Aviation quit on the grounds of failures in the ATC response. This failure is partially explained by the fact that after the IGARI point, the Malaysian ATC believed that the flight was in the Vietnamese airspace, and not their concern. The Vietnamese ATC did their best to establish contact by relay aircraft and by using emergency frequencies. However, the two ATC providers can coordinate within minutes to decide what to do about an aircraft that disappeared from radar and did not return ATC frequency calls. The use of emergency frequencies ruled out any mistake in switching to the Vietnamese frequency over IGARI.

The other failure was that an unidentified aircraft was not intercepted by the military for one hour of flight in the Malaysian airspace. Perhaps this failure is linked to the previous one. It is not clear what were the rules of engagement of the Malaysian military in air police services at the time, and if they required ATC actions or not.

Q: What are the challenges encountered with respect to the SAR operations? Why couldn’t the entire wreckage be found, even if a lot of countries were involved?

The number of countries does not matter if the search place is wrong. ATSB started to look in the right places (according to our research) as late as October 2014, and never searched some parts of the area we indicated [A3]. A second possible reason for the lack of results is the possibility that the place of wreckage was searched but the wreckage escaped the acuity of the acoustic mapping software. For instance, if the fuselage parts fell in a crevasse or in a fracture valley on the bottom, or if they were covered by alluvions or marine life forms, they can hardly be identified. The Indian Ocean in those areas was uncharted, so there was impossible apply the method which was used in the case of AFR 447 in the Atlantic Ocean. It is easy to spot a large foreign object on the bottom if you can compare the shape of the bottom with the known shape from the database. If you do not know how the bottom should look like, it is left to chance to spot the object, as something “unusual”, since a lot of the ocean floor is unusual itself.

Q: How did the ATC manage the safety of the traffic that could be endangered by the fact that no data was available regarding the actual position of the MH370 flight?

Your question is right in the sense that an aircraft flying without an active transponder is deprived of TCAS as well as of secondary radar service, so it may collide with other flights. That is yet another argument to emphasize the urgency of air policing services, to immediately scramble military jets and intercept the transponder-less NORDO aircraft. The Malaysian ATC seemed not aware of the danger at the time, and their management of the situation was severely criticised.

Q: What were the findings of the examination of the recovered debris from MH370, and how did they contribute to the investigation?

The debris were found on the shores of the Indian Ocean, starting with after 16 months since the crash, mostly on the African coast, where the oceanic currents would normally carry such objects (Fig. 1).

What the debris found did for this investigation: they confirmed the following hypotheses:

• The South corridor (as opposed to the North corridor, the INMARSAT signals ambiguity);
• That the aircraft crashed at the end of the 6 hours of unknown flight;
• That the crash site was in the South-Eastern side of the Indian Ocean, but not too close to Australia, because the majority of the debris were not found in Australia West Coast (officially none was found in Australia, but there are unconfirmed claims of such findings).

Fig. 1 – Debris carried by oceanic currents from the presumed crash location normally land ashore in Madagascar and Reunion Island [B5]

Q: What were the capabilities and limitations of the radar systems used to track MH370’s movements?

The civilian ATC in both Malaysia and Vietnam relies on SSR (Secondary Surveillance Radar), like in most other parts of the world, but since MH370 had its transponders switched off over the hand-over point of IGARI, the target became invisible on these radars. That is a major limitation of the civil aviation radars, needing the target to cooperate in order to be seen, but there seems to be no return to the days when civilian radars where primary and secondary. Their performance does not come into play. Apart from the SSR tracks of the MH370 until the point IGARI, we also have the ADS/B signals which contain the latitude, longitude, and the height of the flight until IGARI (Fig. 2). The positioning data in this case is sourced by the onboard navigation systems. The errors of these positioning is small (normally within ±0.04 NM horizontally and ±25 ft vertically).

The military radars are primary, so they can track targets just based on the radio returns of the metal body of the aircraft. MH370 was seen on the Royal Malaysian Air Force primary radar after it disappeared from the secondary, and their primary radar did a circular scan and also a vertical scan to measure the angular altitude of the target. This is how we know about the climb to the dynamic ceiling, then descent down to a low flight level, and climb back. The RMAF refused to disclose data on their radars and denied to acknowledge the identity of this target. It is impossible for me to know the exact capabilities and limitations of a military facility, but what I can say is that we checked the moment in time when the allegedly unknown target was lost by the primary radar and it coincides to the place where the aircraft exited the expected coverage range due to the curvature of the Earth, assuming the radar somewhere on the Malaysian coast.

Fig. 2 – The first part of the MH370 flight on route to IGARI; this was recorded by FlightRadar 24 based on the ADS/B positions reported by the aircraft itself

Q: There was a documentary recently which implied that a self motivated sleuth has been using perturbations in radio wave propagation to ‘locate’ the passage of a vehicle (in this case MH370) – are you able to talk to that process please?The 

The WSPR (weak signals) theory by Godfrey is the latest method attempted [B9], [B10], [A20], [A21], based on the logs of the HF 10 MHz signals recorded by the Radio Ham community.

The MH370 disappearance in 2014 was followed by a number of attempts to find the trajectory of the missing airplane using the INMARSAT communications with the onboard SATCOM transceiver [A2][A3][A5]. Some relevant papers were published in the Journal of Navigation [A2][A3]. The last of these publications was the official ATSB, which confirmed the findings of the previous sources by publishing a so-called heat map (high probability crash area) [B4]. We do not know the methodology used in [B4] because this was no scientific paper with the obligation to disclose it, but we assume that ATSB employed a scientific team of experts to do their own calculations. The results of these calculations accurately matched the results in [A1][A3], so we may conclude that there are at least these three teams who did independent work with the published INMARSAT signals, and the final results matched.

There was a claim in the media [C9] that the published INMARSAT signals were not reliable, even with speculative accusations that INMARSAT tried a cover-up. I disagree based on the following:

(i) If INMARSAT wanted to reduce the probability of the MH370 wreckage ever being found, why would they release anything publicly? Their best action would have been to refrain from publishing the raw satellite signals as early as March 2014. INMARSAT is a satellite communications company with no obligation to track airplanes. Before their data was published, the search area was in the South China Sea, around the last known point on radar (IGARI). This search area at the time was definitely wrong, and proof of that came after one year and four months when the first debris was found in the Indian Ocean basin. The South China Sea does not belong to this basin. The evidence provided by INMARSAT was the reason for ending the useless search in the South China Sea and start to look at the Indian Ocean.

(ii) If INMARSAT data were altered or skewed, how come they are consistent with three separate methods [1-4]? Suppose that the method in [A2] published by the INMARSAT team (Ashton et. al.) was not objective, if INMARSAT had any interest in tampering with the data. However, our own method based on BTO and validated by the onboard fuel calculations would never concur with the INMARSAT team’s method based on BFO. Our own method was invented by us and communicated in August 2014 to ATSB. Genuine data and false data can be discriminated by consistency. INMARSAT raw data as published in March 2014 were consistent with both the Ashton et al. method [A2] and our own method [A1], [A3] invented later, in March-April.

Our conclusion is the INMARSAT radio signals were genuine and reliable. As compared to the radio weak signals, these are the strong UHF signals which originate undoubtedly onboard of MH370, they follow a line of sight propagation path, and they were captured by the INMARSAT IOS satellite, which has a defined position in time.

The expectations are high on this weak signal theory (WSPR). The radio ham community HF signals (mostly on 10 MHz) recorded in 2014 are the only new source of data on Malaysian 370 not explored yet until 2021. However, Godfrey who is the leader of this group, initially came up with a very short trajectory short of the 7th Arc. That was inconsistent with the strong INMARSAT signals. We criticised it. The weak signal theory is a multi-static statistical theory. The fact that the reflections came from Malaysian 370 and not from any other object is disputable. In the South Indian Ocean, the waves are sizeable, of the order of the HF wavelength, so just the assumption that the water surface is flat is inaccurate. How could this theory contradict the strong UHF radio signals coming surely from this aircraft? What you see here is the new version solution proposed by Godfrey on his site. It is changed to be consistent with the INMARSAT signals, ends on the 7th Arc and flies on the same heading every time a ping occurs, to be consistent with the BFO diagram (Fig. 3). It is simply impossible to fly like that because some pings were initiated by the satellite (the SATCOM calls for instance) and these moments were unexpected in the cockpit to have the time to turn such an airplane on an exact new heading. I do not appreciate results which evolve in time and adapt to criticism. The weak signal database was not made publicly available.

One may track aircraft which fly in the proximity of the HF radio ham WSPR network nodes. Airplanes are the only radio-reflective objects in a large range of angular altitudes. However, this advantage goes away when the same technology is used at a global scale distance.

We dispute the claim that WSPR can be used to track MH370 but we are ready to do our own research if the raw WSPR data are made publicly available.

In the MH370 case there are no antennas in the proximity of the flight. The radio ham community from Australia and Africa were too far away, and the Indian Ocean together with Antarctica is a large region of the world lacking any radio ham activity. The WSPR is a multi-static technology based on statistics and requires statistically significant data. How close can these data originate from the presumed MH370 trajectory? The final point of the flight as calculated by our team [A3] is 1,500 NM away from the Australian coast and 2,700 NM from the African coast.

The radio weak signals technology is based mainly on HF frequencies, with a mixture of sky wave and line of sight propagation, and MF frequencies, which also suffer from ground wave propagation. HF is the only part of the radio spectrum which has never been used for radio location, and the reason for that is exactly the uncertainty of the propagation path due to sky wave and the multipath effect, which is the cause of the specific phenomenon called fading. Thus, to attempt radio location on 10 MHz is a very difficult bet. The strong point of the HF radio spectrum is radio communications and not radio location. The WSPR project itself is not a radio location project. Its main purpose is to enhance low power radio communication range using multipath to its advantage, by speculatively using any objects for reflections, including the Moon, meteorites, and airplanes.

When the first document was published by Godfrey in 2019, there was a claim that innovative weak signals method indicated that MH370 made many turns before ditching, thus contradicting all existing references [A1][A2][A3][A5]. He did not even use those references. In fact, the recently published documents [B9][B10] are not scientific papers, they are reports with no validation from independent researchers. Between 2019 and 2024 there was enough time to write and to publish a scientific paper if the authors wanted. The difference between a scientific paper and any other type of document is that the calculations in a scientific paper should be reproductible in principle by any educated reader, so the scientific paper discloses all the relevant data and assumptions in order to allow for other teams to replicate the results. A scientific paper is an exercise of transparency and honesty (at least it should be). A report such as [B9][B10] has no such obligation, and we believe that these documents are not disclosing many details of how they determined that MH370 did those turns. How to push science forward if an author keeps secrets for himself and the published method is not usable by others?

Before the analysis of the contents of the weak signals theory results [B9][B10], there is an important point to make:

The weak signals theory is new and uncertain, based on multipath HF radio signals, which involve many uncertainties, and many reflections in any object that could be present in the area. How could the weak signals results contradict the strong signals in the case? The strong signals methods [A1][A2][A3][A5] were also new at the time, but more teams independently reached at the same results using different methods, and this decreased the uncertainty of the methods. The propagation of the strong UHF signals is thoroughly studied and certain, being for example at the core of the Global Navigation Satellite Systems accurate positioning and timing.

The authors of [B9][B10] claim that WSPR is validated to track aircraft which fly in proximity of the network nodes, and this is perhaps correct. The reason why this statement is perhaps correct is that in a certain proximity, the number of aircraft is limited and also because other objects which cause reflections (ships, buildings, etc.) have an angular offset with respect to the antennas incompatible with airplanes. An airplane flying in proximity is visible under a typical angular altitude of 10-40°, and reflections of it do not mix with other reflections. Airplanes are the only radio reflective objects in a large range of angular altitudes. However, this advantage goes away when the same technology is used at a global scale distance. In the MH370 there was no possibility to involve this proved proximity WSPR technology to locate the flight simply because there were no antennas in the proximity of the flight. The radio ham community from Australia and Africa were too far away, and the Indian Ocean together with Antarctica is a large region of the world lacking any radio ham activity. The WSPR is a multistatic technology based on statistics, and require statistically significant data. How close can these data originate from the presumed MH370 trajectory? [A3]. The final point of the flight as calculated by our team [A3] is 1,500 NM away from the Australian coast and 2,700 NM from the African coast.

Fig. 3 illustrates the Burst Frequency Offsets (BFO) of the 7 pings between the INMARSAT IOS satellite and MH370 [A2]. Changes in the BFO indicate a turn of the aircraft as a consequence of the change of the relative velocity between the aircraft and the satellite, which cause a different Doppler shift. If we assume a major turn followed by a flight in a straight line (orthodrome in fact), the Doppler frequency shift would look like the red line or like the green line, depending on the major direction of flight: to North or to South. This is how, as early as March 2014, experts realised that the missing flight flew for 6 hours to the Southern Indian Ocean and not over India to Kazakhstan. The BFOs of the 7 pings (the blue line) were consistent with the green line, so undoubtedly the direction of flight was South. But this image also is a proof that MH370 did not fly around in circles and in fact did not do any further turn during the 6 hours. This contradicts the Geofrey’s team results with flying round in circles (Fig. 4). In [A3] we also considered the multiple turns hypothesis, and we explained better there why we had to discard it on factual grounds.

Fig. 3 – The BFO analysis [A2] clearly demonstrates that after the large turn in the Andaman Sea (the vertical blue line), the rest of the flight followed a straight line (with reservations on the use of the word straight for such a long flight)

Fig. 4 – The author of the WSPR theory presents on mh360search.com this multiple turn trajectory

The weak signal theory is a multi-static statistical theory. The fact that the reflections came from Malaysian 370 and not from any other object is disputable. In the South Indian Ocean, the waves are sizeable, of the order of the HF wavelength, so just the assumption that the water surface is flat is inaccurate. How could this theory contradict the strong UHF radio signals coming surely from this aircraft? What you see in Fig. 4 is the new version solution proposed by Godfrey on his site. It is changed to be consistent with the INMARSAT signals, ends on the 7th Arc and flies on the same heading every time a ping occurs, to be consistent with the BFO diagram (Fig. 3). It is simply impossible to fly like that because some pings were initiated by the satellite (the SATCOM calls for instance) and these moments were unexpected in the cockpit to have the time to turn such an airplane on an exact new heading. I do not appreciate results which evolve in time and adapt to criticism. The weak signal database was not made publicly available.

In conclusion, I consider the claims that WSPR is a reliable tool to locate the MH370 are grossly exaggerated in [B9][B10] and I rely on (i) the scientific arguments related to the uncertainty of the radio propagation trajectory paths of HF signals, and (ii) the results of these weak signals which are contradicted by the strong INMARSAT signals.

Q: I have seen a lot about MH370 including that it was a suicide flight by the pilot. Is there evidence to support this please?

We provide circumstantial evidence that it was an intentional act managed from the inside of the aircraft, not necessarily suicidal since there is a way to jump from a B777 using a parachute. That would require an accomplice to wait in a boat at an agreed meeting point LAT/LONG. That opportunity existed exactly in the middle of the Malacca strait, where according the military primary radar, the aircraft made a temporary descent, possibly to 10,000 ft, before climbing back.

Q: Why with so many hot tips as to what happened  has there been no conclusion?

The common conclusion among the experts is that it was an intentional act meticulously prepared and conducted from the inside of the aircraft. If by “conclusion” you mean finding the wreckage on the bottom of the Indian Ocean 5000 m deep, that is more challenging than finding the AFR 447 wreckage for example. The depth was the same, but that ocean was at least charted and the approximate impact place was known (the vertical stabilizer of the airplane was found floating in a matter of days). If this looks easy as compared to MH370, please remember that it took two years of searching to find the wreckage.

Q: What part did aero navigation play in this conundrum?

For our method [A1], [A3], the flight simulation was very important to establish the probable moment in time when the airplane ran out of fuel, and this moment should coincide with the 7th ping, which was only half a ping, it was never completed. The almost certain cause for that was the failure of the aircraft generators after the fuel starvation. The RAT turbine does not feed the SATCOM transceiver (if anybody asks, we belive it should). We considered the weather at the time and at the place of the flight in these simulations, most importantly the wind direction and velocity. For instance when we removed the weather data from our simulations, the square sum of the trajectory errors jumped from 20 km to 200 km. Even for the researchers who considered a constant ground speed of the aircraft (ignoring the wind), the velocity of the aircraft was a major input to calculating the relative velocity to the INMARSAT satellite, which is the cause of the Doppler effect and the BFO.

Also the method we used in [A3] is an illustrative case of a flight trajectory optimisation, using multimodal optimisation, based on a full search algorithm. That is why I included this case in the Chapter 11 of my Air Navigation book. The general method works backwards and can be used to optimise any flight trajectory  before the flight.

Q: The INMARSAT tracking process and the margins of errors associated with the calculations

There are many sources of errors in the radio signals, but also in the flight simulation itself. The sources of the errors in the radio signal itself are the following:

• INMARSAT satellite transceiver errors and the ground INMARSAT station transceiver errors; these are frequency and latency errors estimated and partially corrected by the INMARSAT team [A2]; frequency errors affect BFO and latency errors affect BTO;
• Aircraft SATCOM transceiver frequency and latency errors, which are assumed within the expected limits; however, here there is a special factor which might have an impact: the temperature inside the depressurised aircraft was probably −50° and that might have impacted the frequency and the latency errors; in a further attempt to calculate the trajectory, this is a factor which needs to be taken into account;
• Every signal from the aircraft to the satellite and to the ground and back passed four times through the ionosphere and this added to the latency of the BTO; [A1], [A2], [A3] include attempts to compensate these errors, but the ionosphere is variable in time and space;
• Refraction in the atmosphere affected all the transmissions four times as well, but not that much because of the position of the satellite in the sky, which was between 40°-70° of angular height.

We tried to avoid all the following errors in the flight simulation itself in what we published [A1] and [A3], but other researchers did not give consideration to these errors:

• The wind influence over the flight (most other teams presumed the ground speed of the aircraft constant);
• The temperature influence on the true airspeed;
• The pressure influence on the flight level;
• The influence of the automation mode of the flight (autopilot maintaining heading, or track, or guided by the FMS)

Q: Have the radio signal intercepts of the WISPER (?) really helped to give a more informed position for the final resting place?

The short answer is no. A longer answer will be presented in a separate paper addressing this subject in more detail. An even longer answer is expected 6 months from now, from a team of researchers of the University of Liverpool, who are expected to do a thorough scientific validation of the MH370 WSPR documents. Other researchers would try if the scientific data behind the MH370 WSPR were made publicly available, but it seems that just the Liverpool team was supplied with the undisclosed details.

Q: How would you, could you, deploy UUVs with sonars to search a larger area than a manned vessel could search?

We are no underwater search experts, but we can use engineering knowledge to say this:

(i) UUVs deployment still requires manned vessels as home bases, because of the UUV autonomy and range.

(ii) UUV with deep submersing capability are capable of a better sonar picture, since they float closer to the bottom. Surface vessels may drop sonar heads underwater though, but still it would be too risky to drop them too deeply (because the length of the steel ropes cannot exceed the minimum depth and the shape of the ropes is parabolic, with a horizontal component which grows with the speed of the vessel).

(iii) The problem of the identification of the B777 wreckage on the bottom of the sea is challenging because:

(iii-1) The bottom of the Southern Indian Ocean is not accurately charted, and the identification of cylindrical parts such as parts of the fuselage for example, would require a comparison between the empty bottom and the bottom with such a cylindrical formation.

(iii-2) In 10 years, the shape of the B777 parts may have been reshaped by the environmental factors, so the search should not assume just the initial shapes of the parts. The wing is probably not recognizable anymore because the most probable position is horizontal and its shape may be masked by the alluvial deposits, underwater life forms, etc.

(iii-3) The parts with evident shape such as parts of the fuselage may find their resting place in crevasses on the bottom, so their presence might be masked, if the bottom is not flat.

What seems to us as very relevant is the quality and the sophistication of the debris detection software, and that is more important than the carrier of the sonar (surface vessel with underwater sonar heads or UUVs).

Q: Although debris has been found in the western Indian Ocean, the drift rate from the area of search seems very high. Are the results consistent?

The drift in one year and four months can easily cover such distances. The Pacific Ocean cases demonstrate even larger distances. Yes, the results seem consistent to us. We expected to see debris in one year after the crash, and we were wrong just by four months. It is not only the oceanic currents which cause the drift, but it is also the surface wind which pushes the object by aerodynamic drag, and also the tides and the waves. Accurate modelling is hard, because part of the object is immersed following the current, and part of it is out of the water, being pushed by the wind with a force equal to the aerodynamic drag, but with a reaction from the hydrodynamic drag.

Q: In determining the search area for MH370, the Northern and Southern Corridors are always cited. Is this determined by a single range/altitude from the Indian Ocean Region satellite and with what certainty?

Yes, and the reason for that is that the only functional radio transceiver onboard MH370 was the INMARSAT SATCOM transceiver with 7 radio exchanges (pings) over the 6 hours of unknown flight. The certainty of the INMARSAT radio signals is high since this network has been extensively used for aviation data link, with a lot of experience, including their involvement in the AFR 447 case in 2009. Actually BTO was introduced in the aftermath of AFR 447 to enable more accurate locations using the communication signals. The accuracy of the INMARSAT signal parameters (BFO and BTO) is best described by Ashton et. al in [2], with some factors causing errors (such as the satellite orbital imperfections, and the thermal condition of the satellite electronics). Even if we assume lower accuracy margins for these signals, our calculations are not sensitive because BFO comes in Hz and BTO in microseconds, and the size of the errors which could sensibly affect our results would need one more order of magnitude. We did consider these error factors ourselves in [A1], but more accurately in [A3], because we benefited from the extensive coverage of the error factors in [A2].

As for the discrimination between the Northern and the Southern corridors, our first paper [A1] found one order of magnitude more probability in the Southern direction. That was confirmed to us by the BFO diagram in [A2]. The Northern corridor is now totally excluded by the evidence which came later in the form of the debris found on various Indian Ocean beaches.

Q: Is the debris collected on the Indian Ocean islands actually from MH370?

The short answer is yes, some of the parts have been clearly identified as components of that airframe. The others cannot be directly confirmed, but the pattern of the locations is consistent with the expectations. Also, Boeing 777 debris are not exactly a common object to find, simply because these parts are rather expensive, and there is only one Boeing 777 (MH370) which crashed in the Indian Ocean basin since the 1990s and up to now. Thus the probability that they come from another B777 is non existent. Another clue is that most fractures of the found parts are consistent with a high speed aircraft impacting the water.

Many of the debris were found in places where debris were expected, and that explains why some people became MH370 debris searchers and managed to find parts on some beaches where they looked on purpose. If you follow a map with the oceanic currents, you get a fair idea where to search.

Another indirect evidence is the barnacles deposited on some of the debris found (Fig. 5). Their age is consistent with the time passed since the crash until they were found. There is even a marine biologist who attempted to calculate the location of the crash based on the study of the barnacles.

The conspiracy theories on the subject claim that the debris were fake, or planted on purpose, etc. Although we should not disregard any possibility, we concluded that these theories have no merit. Too many details are convergent with the hypothesis of authentic debris, and not a single detail is seriously questioning it. Also, even if some debris were planted to fake the Indian Ocean basin as the location of the crash, why are the authentic debris completely missing in other areas?

Fig. 5 – This image taken from a video, shows the flaperon, the first piece of debris found on July 29, 2015, in Saint-Andre, Reunion. Barnacle shells attached to the frame are clearly visible. (VOA News, Reuters)

Q: The aircraft must have left a contrail at some point in the flight south, were there any met or other observation satellites that could have seen it? A single observation would be enough to ‘fix’ the trajectory. In the unlikely event no one has ever looked for it, does the data still exist?

This is a question that has bothered us as well. It seems that as many as they are on various orbits, the satellites cover just a very small proportion of the globe at any given moment, and the satellite pictures are not taken continuously, some missions take pictures when there is an objective in the region they overfly. Military satellites change orbits when they need to cover a certain area of interest. However, satellite imagery provides a “static” image of the whole world, and this is assembled from tiny fragments gathered in time. We are very far from an online global image of the world. Another point is that the ALERFA and DETRESFA phases were triggered many hours late, and for one whole month the search area was in the South China Sea around IGARI. I remember that some French satellites were put to work intensively to  gather images of the area, and they found many objects afloat. As for the Southern Indian Ocean, that area came under scrutiny as late as October 2014, 7 months later. I presume the archive images were analysed. Yes, the data is still out there, but most of it is not publicly available.

Secondly, I would say that contrails are clearly visible on the sky background, but less so on the earth background. I remember doing some research with identification of aircraft in satellite images and the contrails were grey and hard to differentiate from the background. The body of the aircraft itself was white and more visible, but very small.

We do not know about the military capabilities, but we presume they use resources when they have an objective and do not waste resources on random search. Anyway, there is much secrecy around the military. For example, the Malaysian primary radar data were not disclosed for weeks after the disappearance, and later all the information was withdrawn and all requests made to them to confirm or to explain were left unanswered. Luckily, we saved everything at the time, but nobody is backing it up.

Q: The Australian military famously operate Over-the-Horizon radar. Did they really not pick up anything?

As for the Australian radar, the trajectories calculated in our Journal of Navigation paper are passing at 1,135 NM from the closest point on the Australian coastline, and just avoids the Jindalee Operational Radar Network (JORN) Radar 2 Western coverage corner by a small margin (Fig. 6).

Q: What if any are Low Frequency Radio transmissions playing into the investigation?

The most credible calculations are based on UHF radio transmissions. There was the Australian military radar using lower frequencies, but the trajectories calculated in our Journal of Navigation paper are passing at 1,135 NM from the closest point on the Australian coastline, and just avoids the Jindalee Operational Radar Network (JORN) Radar 2 Western coverage corner by a small margin (Fig. 6).

Fig. 6 – The Jindalee Operational Radar Network (JORN) coverage, an over-the-horizon radar (OHR) network operated by Royal Australian Air Force [*]

[*] Wikipedia – https://en.wikipedia.org/wiki/Jindalee_Operational_Radar_Network

Q: How many vessels were involved in search for this aircraft and who was the main coordinator of this sea search?

In the first phase of the search (South China Sea and then West of Malaysia) 8-15 March 2014, there were 28 aircraft from the People’s Republic of China (2), Japan (5), Malaysia (10), Singapore (4), Thailand (1), United States (2) and Vietnam (4). Also 34 vessels from the People’s Republic of China (7) Malaysia (19), Singapore (3), United States (3) and Vietnam (2) [B3]. BBC news at the time, reported the number of vessels as 40. The SAR operations were coordinated by the Kuala Lumpur Aeronautical Rescue Coordination Centre (KL ARCC).

In the same time interval Royal Malaysian Air Force conducted their own SAR operations involving 36 aircraft and 35 vessels from Australia, Bangladesh, the People’s Republic of China, India, Indonesia, Malaysia, New Zealand, Republic of Korea, Singapore, Thailand, United Arab Emirates and the United States [B3].

Because the question is on vessels, we skip the aerial search and satellite search operations from this list. These are presented in [B3].

Between 28 March – 3 April 2014 AMSA/ATSB coordinated a search West of Exemouth, Australia, but mainly from the air, employing aircraft from Australia, New Zealand, United States, Japan, the People’s Republic of China, Republic of Korea and Malaysia and vessels from the Royal Australian Navy and the People’s Republic of China. Between 3–28 April 2014 the search turned to another phase, involving RAN vessels and from Royal Malaysian Navy, Royal Navy, United States Navy, and the People’s Republic of China. The number of the vessels is not mentioned [B3].

Between 2-10 April 2014, 3 specially equipped vessels were deployed to the search area around the 7 arc. Also “underwater search using an autonomous underwater vehicle (AUV) commenced on 14 April 2014 in the Zenith Plateau area based on the analysis of the acoustic detections from the TPL on board Ocean Shield. Thirty AUV missions conducted at depths from 3,800 m to ~5,000 m were completed” [4]. All this search was still North of the Broken Ridge, which we believe not to be consistent with the INMARSAT data combined with the flight performance of the aircraft and the weather at the time, as calculated in [A3].

The initial search in the higher priority area (around Broken Ridge, but to the north of it) started on 14 June 2014 with two vessels: Fugro Equator (ATSB contract survey vessel, multibeam echo sounder (MBES)) and Zhu Kezhen (the People’s Republic of China hydrographic survey vessel, MBES). Their mission was to do mapping of seafloor.

The search missions are comprehensively illustrated in Fig. 7 [B2]

Fig. 7 – The search areas in the Indian Ocean and their chronology [B2] – Pleter et. al results are not mentioned here, the so called October 2014 hashed area refer to Ashton et. al results [A2]

Q: Is it feasible to implement a comprehensive aircraft tracking system that would facilitate swift identification of the location of an aircraft in case of such an emergency?

The Aireon Global Satellite Tracking system has been introduced, but it is not effective when the ADS/B transponder is cut off from the cockpit, like in the case of MH370. The existing ELT/ULB are ok but the battery life of the ULB would need to be increased even more than the 90 days (in the case of MH370 it took 7 months to start the search in the plausible area).

There is also the proposition of ADT but this also fails in a similar case to MH370, because radio transmissions are impossible if the antenna is sunk in water.

A definitive solution would be the ejectable floatable data stick, which we proposed and tried to patent in 2014 but failed to do so because this is a safety device. However, we are happy that the idea gathered some support, so it appears in the ChatGPT list of post-MH370 tentative measures (see Impact and Lessons Learned).

Q: Why were there conflicting reports and confusion about the disappearance at the time?

The place and the time of the disappearance were meticulously planned. At the hand-over point, IGARI the Malaysian ATC believed that the flight is no longer their concern, whereas the Vietnamese ATC did not establish any radio contact. Diversions in the plan of a perpetrator have the very purpose of generating havoc. Initially, nobody expected this to be a voluntary act, so the picture was very confusing. In the moment the voluntary act hypothesis was later considered, things started to come into focus.

Another factor of the confusion was that the Malaysian millitary were first reluctant to release the primary radar tracks of the aircraft and in fact they have never recognised that the target was indeed MH370, or any another airplane.

Q: Are airlines flying across bodies of water fitted with devices similar to EPIRBS? (Emergency position indicating Radio beacon ) like on ships these would have a hydrostatic release on entering the water. These devices would be outside the plane possibly on the tail section  and could not be interfered with by people inside the plane.

Yes, they are called ULB and transmit acoustic signals for 90 days. In the MH370 case the requirement for the battery life was just 30 days. Unfortunately, the MH370 search in the plausible area started not sooner than 7 months later, in October 2014, so 90 days wouldn’t have helped.

Q: What was learned from the routine signals transmitted from the engines to the engine manufacturers?

Until the IGARI point, the engine reports were transmitted regularly and indicated no engine problems. After the transponders and the transceivers were disabled or switched off, the next expected engine report never came, although the SATCOM transceiver was on and logged on the INMARSAT satellite in the Indian Ocean. This is evidence that the ACARS unit was disabled, presumably from the circuit breaker (pilots are not instructed on how to do it). After this moment, for six more hours the INMARSAT satellite got just 7 pings from the aircraft SATCOM, and these are the main proof that the aircraft continued to fly and for how long.

Q: What was learned from the simulation which the pilot had run on his home simulator?

The simulator was very advanced and accurate, albeit not too expensive (so not in the professional category). The Malaysian forensics managed to recover a 7 points entered manually (LAT/LONG). Points 6 and 7 point to a final destination in the Southern Indian Ocean. Their position is not consistent with the INMARSAT signals, so we assume they were used as the LAT / LONG fix as to navigate to in LNAV mode of the FMS. The aircraft ran out of fuel before reaching them. Points 1 to 4 follow the track through the Malacca Strait and point 5 is in the Andaman Sea. These recovered waypoints were published in the official accident report Fig. 8 [B4]. The points were stored on the hard disk one month before the flight.

Fig. 8 – Deleted waypoints recovered from the MH370 captain’s home flight simulator [B4]

Q: Please explain how Inmarsat were able to derive ranges from satellite signals from MH 370.

The radio connection between the aircraft SATCOM transceiver and the INMARSAT ground station via the INMARSAT satellite (used as a repeater – relay with known delays) is timed, and a BTO (Burst Timing Offset) is logged for every transmission. This BTO started to be recorded in the log in 2009, after the AFR 447 accident. So, in principle, you know that the aircraft is on a sphere with the centre in the satellite and the radius is given by that measured range (a correction is made for passing four times through the ionosphere). If you intersect that sphere with the Earth ellipsoid you get those 7 quasi-circle arcs for the 7 pings. Of course, there is the flight level uncertainty, which has an influence over the results, but in our calculations, the results are not significantly influenced by the 45,000 ft uncertainty in the total vertical range.

Q: After running out of fuel, would that type of aircraft enter an uncontrolled descent causing a mid-air break-up, and is the damage seen on the parts recovered consistent with this?

No. The aircraft glides down on a 7° to 9° slope and impacts the water in a negative pitch attitude. This is what most B777 flight simulators can replicate. The challenge is with estimating the turn. Due to the fact that the left engine becomes fuel starved before the right engine, there is a turn to the left which will continue over this descent. Because there was no way to predict the time between the engine out events in that particular case, we needed to consider all possible turn angles and we assumed the area in Fig. 9.

Fig. 9 – B772 powerless glide trajectories with uncertainty over the left turn rate due to the fact that the left engine shuts down first and there is an unpredictable delay between the two engine out events; the coordinates are latitude and longitude, but the values are generic, just indicating the range of the resulted area, not a particular position of the crash; the red star is the start of the glide at the moment the aircraft becomes powerless and without electric power except for the RAT, and this moment coincides with the 7th ping (half-ping) [A1]

Q: Given the sea floor profile, was any consideration given to design, build and operate a large fleet on Autonomous Surface Vehicles (1000s of) to search larger area of the sea?

We are not underwater search experts. Our contribution regards the flight itself to find the most probable area to do the search. This area is large, so your idea is excellent! Unfortunately, there is a shortage of funds and there is also a problem with the technique to sense the aircraft parts on an uncharted sea floor.

Q: Has anyone thought of dropping some GPS tracked rafts into the sea in possible search areas? Then the rafts which wash up in Madagascar where wreckage items were found, should indicate the best place for new search!

The drift of floating objects at sea for a long time is affected by dispersion and dispersion is affected by entropy, so you cannot replicate the exact route of any given object, the route is subject to many so-called bifurcations. If you release at sea a group of separate objects, they will land at different locations, as distant as thousands of miles away from each other. This experiment has been done already.

It is not only the oceanic currents which cause the drift, but it is also the surface wind which pushes the object by aerodynamic drag, and also the tides and the waves. Accurate modelling is hard, because part of the object is immersed following the current, and part of it is out of the water, being pushed by the wind with a force equal to the aerodynamic drag, but with a reaction from the hydrodynamic drag. So, for an accurate experiment, we need to know exactly how every piece of debris looks like and how much is it submerged in water. This is also influenced by the barnacle shells which are attached to it. Arbitrarily shaped rafts may end somewhere else.