First Principles Review
This report documents the proceedings and outcomes of the First Principles Review meeting on the search for missing Malaysia Airlines flight MH370 held in Canberra from 2 to 4 November 2016. Participants consisted of experts in data processing, satellite communications, accident investigation, aircraft performance, flight operations, sonar data, acoustic data and oceanography. The purpose of the meeting was to reassess and validate existing evidence and to identify any new analysis that may assist in identifying the location of the missing aircraft.
Throughout the search, the ATSB has issued several reports updating the definition of the search area based on analysis progressively refined, or when new information has come to light. This document complements those reports and provides a summary of the detailed analysis of the satellite data combined with new evidence derived from the modelling of the drift of debris from the aircraft.
The experts attending the meeting considered:
- The results of the search to date.
- Satellite communication metadata and its analysis including methodology, assumptions, limitations, the probability distributions of possible flight paths, and validation results.
- Results from simulations and the aircraft manufacturer’s analysis of aircraft performance.
- The width of the search area based on what is known about the end of the flight.
- Hydro-acoustic analysis potentially relevant to the search.
- Failure analysis of recovered debris.
- Drift analysis of aircraft debris.
For background information, please refer to the previous ATSB publications available online at www.atsb.gov.au/mh370
The updated independent analysis of the satellite data and the drift analysis consistently identified the most likely impact location of MH370 as being close to the 7th arc (within ~25 NM) and bounded by latitudes of approximately 33°S to 36°S.
There is a high degree of confidence that the previously identified underwater area searched to date does not contain the missing aircraft. Given the elimination of this area, the experts identified an area of approximately 25,000 km² as the area with the highest probability of containing the wreckage of the aircraft. The experts concluded that, if this area were to be searched, prospective areas for locating the aircraft wreckage, based on all the analysis to date, would be exhausted.
 The 7th arc is an arc of possible aircraft positions, equidistant from Inmarsat’s Indian Ocean Region satellite, where the accident aircraft made the final series of satellite communications transmissions. It is the key datum in the search for MH370 and its derivation is described in previous ATSB search area definition reports.
The First Principles Review
In May 2014 the Governments of Australia, Malaysia, and People’s Republic of China agreed that the Australian Transport Safety Bureau (ATSB) would coordinate the underwater search for MH370. The primary objective was to assist the ICAO Annex 13 investigation by locating and positively identifying the aircraft wreckage. It was also essential that any searched area assessed as not containing the aircraft could be discounted with a high degree of certainty.
In November 2016, the high priority underwater search area defined in the ATSB’s previous reports was in its final stages of being searched. As such, the ATSB determined that a review of all available information relating to the search area definition was required. Participants at the three-day meeting held from 2-4 November 2016 at the ATSB office in Canberra included experts from each field of expertise where information was available to assist in the search area determination.
Accordingly, there were representatives at the meeting from all of the organisations participating in the Search Strategy Working Group including Australia’s Defence Science and Technology Group (DST Group), Boeing, Thales, Inmarsat, the National Transportation Safety Board of the US, the Air Accidents Investigation Branch of the UK and the Department of Civil Aviation, Malaysia. In addition, there were representatives from the Commonwealth Scientific and Industrial Research Organisation (CSIRO), Geoscience Australia, Curtin University, Malaysia Airlines and the People’s Republic of China.
The aim of the First Principles Review meeting was to consider the results of the underwater search to date, examine the data and analysis associated with the accident flight, and the definition of the search area. The meeting also focussed on new analysis relating to recovered aircraft debris.
The first day of the meeting focused on a comprehensive review of the satellite communications (SATCOM) data, and the analysis methods used to estimate the aircraft’s final location. The meeting included presentations from all organisations with subject matter expertise related to the SATCOM system. The review re-examined the detailed assumptions, calculation methods, and validations that led to the definition of the search area. The analyses had been verified using data from previous flights. The meeting participants determined that techniques used were acceptable and the results for the analysis of the accident flight were valid.
The second day of the meeting focused on the debris that was recovered from the islands to the east of Africa and the African coast from Tanzania to South Africa. CSIRO presented the analysis which examined the rate of movement (drift) of the flaperon (found on La Réunion Island in 2015) and other items, and its correlation to drift models. The results of this analysis identified an area of likely origin of the recovered debris (i.e. the aircraft’s impact location). ATSB experts presented conclusions from the examination of the recovered flap section with respect to the likely configuration of the aircraft at the end of flight. The meeting participants determined that the analysis presented on the debris and subsequent results were valid.
The results of the underwater search were also presented including the search methods, completed coverage, sonar data interpretation, quality assurance and confidence in the results. A presentation on the hydro-acoustic data at the time of the accident allowed meeting participants to determine that hydro-acoustic analysis did not contribute any useful new information to the search.
On the third day, the meeting examined how the results of the search to date, the flight path modelling based on the satellite data, the drift modelling and the debris analysis may assist in the definition of any future search activities and the priorities for such activities.
Throughout the search, data from a communications satellite along with aircraft performance data has been used to attempt to reconstruct the flight path of MH370 from the time of the last radar contact. Information recorded by a satellite ground station at the time of handshakes with MH370 was used to estimate the track of the aircraft. SATCOM information from two unanswered ground to air telephone calls was also available. This information showed that the flight continued for approximately five and a half hours after the last radar data.
The methodologies for the calculation of flight paths were included in the ATSB’s reports MH370 – Definition of Underwater Search Areas report of 26 June 2014 and updated on 18 August 2014, MH370 – Flight Path Analysis Update of 8 October 2014 and MH370 – Definition of Underwater Search Area Update of December 2015.
The calculations identified seven ‘arcs’ which lead to the definition of the search area along the current north-south line of the 7th arc (which is regarded as the primary datum for determining the search area). It was then necessary to define which section of the 7th arc has the highest likelihood of finding the aircraft and the width of the search area to the east and west of the arc; encompassing aircraft performance limits and end of flight scenarios.
The DST Group provided expert analysis of the available SATCOM data relating to MH370. The analysis used models of the Inmarsat SATCOM data, aircraft dynamics, and meteorological data to determine likely flight paths. The DST Group analysis has been published in the book, Bayesian Methods in the Search for MH370. The methodology was subjected to a set of validation experiments to ensure that the set of predicted flight paths aligned with actual flight data for previous flights of the accident aircraft (registered 9M-MRO) and other flights in the air at the same time as the accident. Further burst frequency offset (BFO) analysis performed by DST Group was contained in the ATSB report of 2 November 2016, MH370—Search and debris examination update.
At the First Principles Review meeting, experts discussed their work to refine the analysis and considered the possible analysis techniques, validations and accident result scenarios. Their particular focus was the effect of the sonar search results to date on the probability map for the search area.
Search results and likely aircraft autopilot mode
The DST Group analysis which used a dynamic model of the aircraft, examined five different lateral aircraft autopilot control modes:
1. Constant magnetic heading (CMH).
2. Constant true heading (CTH).
3. Constant magnetic track (CMT).
4. Constant true track (CTT).
5. Lateral navigation (LNAV).
The initial analysis assumed equal likelihood (uniform probability) across the lateral aircraft modes throughout the southern portion of the accident flight. The results of the analysis suggested that based on the satellite data, the aircraft was more likely to be operating at the time in the CTT or LNAV lateral control modes. These lateral control modes generally resulted in the aircraft’s flight path being towards the southern end of the search area. The remaining three lateral control modes generally resulted in the aircraft’s flight path ending north of the CTT and LNAV results, up to 33°S in latitude along the 7th arc.
The southern portion of the search area which mainly encompasses the CTT and LNAV lateral control mode results has been thoroughly searched with underwater assets (deep tow, AUV, and ROV). Based on the results of the underwater search, the probability distribution was updated to reflect the new information of the unsuccessful search which then favoured the other control modes.
Flight crew advice
During the First Principles Review meeting, flight crew with extensive experience on the aircraft type indicated that the aircraft is usually flown in the LNAV or CMH lateral control modes.
This information, combined with the results from the areas already searched to increases the likelihood in the northern section of the probability map in Figure 13.
Width of the search area
The following factors which affect the width of the search were defined previously, in the ATSB publication MH370 – Definition of Underwater Search Area Update of December 2015:
- Tolerance in the calculation of the arcs.
- A series of end of flight simulations conducted in 2014.
- A basic turn analysis.
- A review of previous accidents.
- The maximum range of the aircraft if glided after fuel exhaustion.
Analysis of these factors resulted in three areas of decreasing probability at 20 NM, 40 NM, and 100 NM from the 7th arc.
The latest ATSB report of 2 November 2016, MH370—Search and debris examination update, provided new information relating to the end of flight. This new information included:
- The analysis of the debris recovered from MH370 indicating that the flaps were likely in a retracted position.
- The analysis of the BFO from the final two SATCOM transmissions which indicated that the aircraft was likely to be on an unstable flight path.
- Results from recent simulations showed high rates of descent broadly consistent with the BFO analysis. These simulations indicated that the aircraft was likely to be within 15 NM of the 7th arc.
This information provided significantly more weight to the aircraft being located closer to the arc than the ATSB concluded in 2015—probably within 25 NM, with locations closer to the 7th arc a higher likelihood.
This analysis, and the implications for the width of the search area were presented at the First Principles Review meeting. All participants were in general agreement that the distance required to be searched from the arc could be reduced to 25 NM from the 7th arc with a weighting to the west to account for the arc altitude.
- The reference 7th arc (search datum) is at an altitude of 40,000 ft. It is likely that the aircraft was at a lower altitude (and therefore closer to the sub-satellite point) at the time of the final SATCOM transmissions and so the distance to the west of the reference arc which should be searched is slightly more.
More than 20 items of debris have been recovered and identified as likely to be, almost certainly or definitely originating from MH370. The first of these was the flaperon, found on La Réunion Island on 29 July 2015. Other items have been located along the east and south coast of Africa, the east coast of Madagascar and the Islands of Mauritius and Rodrigues in the Indian Ocean. A complete list of recovered items was published by the Malaysian investigation team and can be found at www.mh370.gov.my/index.php/en/.
An oceanographer from CSIRO was involved in the initial sea surface search for MH370 coordinated by the Australian Maritime Safety authority (AMSA). CSIRO continued to assist in the search for MH370 by completing work predicting landfall areas for floating debris, and analysing the drift of recovered debris.
In April 2016, the ATSB commissioned CSIRO to perform a detailed study of the drift of the recovered debris. The final report of this study has been released concurrently with this report and should be read in conjunction with this section. The CSIRO report and can be found on the ATSB website.
The aim of the CSIRO drift study was to determine the probability of locations along the 7th arc (defined by SATCOM analysis as between 45°S and 22°S) being the origin of the recovered debris.
A forward-tracking numerical simulation was created. Within the simulation, flaperons and other modelled debris were deployed on and around the 7th arc and allowed to drift freely. The simulated paths of the debris were calculated and compared against information from three sets of observations:
1. The extensive aerial and surface search conducted between 18 March and 28 April 2014.
2. The absence of debris from MH370 found anywhere along the coast of Australia.
3. The timing and location of parts from MH370 found on islands in the western Indian Ocean and on the east coast of the African continent.
The following is a summary of the results from the CSIRO drift analysis:
1. From the number and size of items found to date from MH370 there was definitely a surface debris field, so the fact that the sea surface search detected no wreckage argues quite strongly that the site where the aircraft entered the water was not between latitudes 32°S and 25°S along the 7th arc. Those latitudes are also contra-indicated by an absence of aircraft parts being found off Africa earlier than December 2015.
2. Latitudes south of 39°S are quite strongly contra-indicated by the arrival times of the flaperon and other debris reaching Africa, and the fact that those items were many while findings anywhere on the Australian coastline were nil.
3. The absence of debris being found on African shores in 2014 suggests that the impact site was likely not north of 25°S.
4. Assuming that the crash was within ~25 NM of the 7th arc (ATSB 2016), this leaves only points between 36°S and 32°S that are less than ~25NM from the arc but outside the area that was searched in late 2014 and early 2015.
5. There is a region within the 36-32°S segment of the 7th arc, near 35°S that is most consistent with all of the following lines of evidence, taken together including:
a. The absence of detections during the 2014 surface search.
b. The absence of findings on the WA coastline.
c. The July 2015 arrival time of the flaperon at La Reunion Island.
d. The December 2015 and onwards (only) arrival times of other debris in the western Indian Ocean.
Residual Probability Map
DST Group have used the Bayesian approach to incorporate multiple sources of analysis into a single probability map, or probability density function (PDF), of the location of the aircraft.
The main PDF is generated from the analysis of the satellite data and the search area width analysis. The residual PDFs show the effects of the other analyses (i.e. search results and drift analysis) on the probabilities.
Figure 11 shows the result of the SATCOM PDF updated with the search results. The areas searched were removed from the PDF. The residual probability is located in areas yet to be searched, with two clear areas of interest: north of the current search area and south of the current search area.
Source: Google Earth / DST Group
Figure 12 shows the SATCOM PDF with the CSIRO drift study of the flaperon arrival at La Reunion Island likelihood overlaid. The alignment of the high likelihood point of origin of the flaperon from the drift analysis (green) and the peak of the northern SATCOM analysis PDF can be clearly seen (red circle).
Source: Google Earth / DST Group
Figure 13 shows the search probability distribution based on the SATCOM data and CSIRO drift analysis. The section south of the indicative search area (pink box in figure 12) becomes much less favourable once the drift results are incorporated.
Source: Google Earth / DST Group
1. There is a very high level of confidence in the search results to date in the current indicative search area. The high resolution sonar coverage is very high and the data has been subjected to a very thorough analysis. The area has been searched to a level of confidence >95% without identifying the aircraft debris field.
2. The analysis of the last two SATCOM transmissions, the likely positon of the aircraft’s flaps at impact and results from recent end of flight simulations has allowed a revision of the distance required to be searched away from the 7th arc. 99% of the DST Group analysis results lie within 25 NM to the east and west of the arc.
3. A residual probability map based on the comprehensive satellite communication data analysis and updated with the latest search results and the CSIRO drift analysis identified a remaining area of high probability between latitudes 32.5°S and 36°S along the 7th arc.
4. The participants of the First Principles Review were in agreement on the need to search an additional area representing approximately 25,000 km² (the orange bordered area in Figure 14). Based on the analysis to date, completion of this area would exhaust all prospective areas for the presence of MH370.
Click image to enlarge