A new aircraft tracking technology is nearing readiness for the MH370 search. Significant progress has been achieved in refining a fascinating new technology, Weak Signal Propagation Reporter (WSPR), which is a digital radio communication protocol that is providing new hope that a more precise resting place can be determined for MH370 which disappeared over seven years ago with 239 souls on board.
Various known flight paths have been used to help validate this new technology called GDTAAA (Global Detection and Tracking of Aircraft Anywhere Anytime). But recently a new far more complex blind test of GDTAAA was devised by a former Qantas pilot and uses a flight that operated without passengers, that you will not find in ADS-B archives. Since the former Qantas pilot was the captain of the flight, he has the flight track data and revealed the data after the GDTAAA results had been published for the flight. AirlineRatings.com agreed to be an independent adjudicator of the GDTAAA Blind Test.
A short note explaining how this new technology works is available for download
An article by Geoffrey Thomas on this new technology has been published at AirlineRatings:
Following further blind tests using data from recent flights between 2019 and 2021, a key test of GDTAAA will be to follow the track of an AMSA MH370 SAR flight from Perth to the search area in 2014. The Royal New Zealand Air Force has kindly supplied detailed flight information for their sortie during the MH370 aerial search on 28th March 2014. Many thanks to Brian Anderson for his help in obtaining this data. These tests will take two months to run and evaluate and are planned for August and September 2021.
Mike Glynn has set a new mystery flight test.
Departure was GMT Date-03 Jun 2021.
Time of takeoff 0328Z (Actually leaving the runway.)
Lat/Long of takeoff position (Actually Leaving the Runway runway) Approximate as not shown on ADSB. -13.830/172.00.
Briefing: The aircraft is an Airbus A330-200 which has a max takeoff weight of 233.0 KG. Takeoff weight was not a heavy weight. Fuel at takeoff was 48.0 tonnes of which 4.5 tonnes was added by the Captain. The aircraft was not tracking towards Australia initially after takeoff due to runway direction. The aircraft flew at normal levels and was planned initially at FL380. A further step climb was available during the flight.
Its destination was a main airport Qantas destination airport in the eastern half of Australia. There were no known weather deviations along the track.
I attached a short note to the post above explaining how this new technology works. I tried to explain GDTAAA in layman’s terms and required no prior knowledge of GDTAA or WSPRnet. What I have written is a ridiculously massive over simplification of the reality.
On 12th December 1901 Guglielmo Marconi received a signal on Signal Hill, St. John’s, New Foundland, Canada at 47.5685°N 52.6758°W that was sent from on top of a hill near Poldhu Cove in Cornwall, England at 50.0303°N 5.2639°W over a distance of 3,435 km. A skywave propagation can cover a distance of anywhere between 1,000 km and 4,800 km. I use a nominal hop distance of 4,000 km and the first Marconi transmission across the Atlantic was of that order of magnitude.
More recently radio amateurs have bounced radio waves off the surface of the moon and back to earth (768,800 km round trip). The NASA Voyager 1 probe is now 25 B km from the earth and has just left the solar system and is still sending information back to earth, although it takes over 20 hours to arrive.
Prof. Joe Taylor has received signals from pulsars at around 950 light years from the earth. Amongst many other inventions, he defined the WSPR protocol, which is a means of secure communication in an environment where the signal level is poor and the noise floor level is high. Radio amateurs regularly record transmissions on WSPRnet over distances close to 20,000 km (or half way around the globe).
I have today run an experiment involving WSPR transmissions over distances greater than 40,030 km, in other words more than one global circumnavigation. Radio Amateurs and other scientists appear sometimes imprisoned in their thinking that the short path is a limiting factor. Actually the short path (SP) is not the limiting factor. Neither is the long path (LP) the limiting factor. The limiting factor is the signal to noise ratio (SNR) and our ability to define methods like the WSPR protocol, which are designed to pick out key information from a poor signal just above the noise floor level.
The longest WSPRnet transmission recorded is just under 20,000 km. This is only because WSPRnet never measured further than to the other side of the globe. How do you differentiate between long path and short path propagation, because the time differences involved will be less than 100 ms? WSPRnet only measures every two minutes. Radio waves travel around the globe 7.5 times per second.
HF radio propagation comprises a massively complex, multi dimensional and dynamic set of components that result in multiple paths, multiple changes of direction, as well as signals supplementing and interfering with each other. Skywave propagation includes multiple refractions, diffractions and scattering, each element at a different incidence and different reflection angle, where incidence and reflection angle are likely not equal. Skywave propagation experiences different signal losses at different layers in the atmosphere. Skywave propagation is diurnal, nocturnal, grey line, seasonal and subject to the solar cycle and other solar and geomagnetic disturbances. Skywave propagation changes significantly depending on the solar flux index (MAR2014 SFI = 150.5, FEB2019 SFI = 70.6 and JUN2021 SFI = 79.8). According to Rob it should be at least 80 for GDTAAA and greater than 90 for a blind test! MH370 detection and tracking using WSPRnet is looking good in March 2014.
The WSPR protocol allows a meaningful exchange of data between a transmitter and receiver at very low signal levels in a noisy environment. Every WSPRnet link is dependent on the antennas used at both transmitter and receiver. (Glen Elmore, call sign N6GN, has a repeater to a remote WSPR station at 2,286 m on a mountain top 13 miles from his home in Fort Collins, Colorado, USA. He also hangs an antenna below a quad copter drone and flies a wire (most likely dipole wires) up to 100 m above his garden.)
I hypothesise that it is possible that WSPRnet signals travel further than the current WSPRnet limit of 20,000 km on the other side of the globe, assuming calm ocean for interim points for hops landing on the earth’s surface. Rob reports WSPR LP and SP are both possible. In the mornings in his home in Germany he receives VK3MO in Melbourne, Australia on a regular basis, probably LP, and in the late afternoon again on 14 MHz, then very likely SP. Every propagation or space weather report will tell you that both LP and SP are possible.
The more academic papers on HF radio propagation I read, the more I understand that no one fully understands how HF radio propagation works. There are major differences of view on fundamental issues. For example, the average hop length of around 4,000 km comes from one academic paper. The theoretical maximum figure for a hop of 4,800 km comes from another academic paper. The diagram showing a cut off at 4,100 km from yet another source.
I plan to write another explanation of GDTAAA in academic terms requiring prior knowledge of the physics of HF radio propagation. I am still doing the literature review of the research into the subject. No ray tracing software can come anywhere near representing the reality of what is going on, when some WSPRnet Tx in New Zealand gets picked up at a WSPRnet Rx in Spain. The disturbance of a WSPRnet link in the form of an anomaly in the SNR or frequency drift has been shown to be caused by aircraft (including engine exhaust and wake turbulence) crossing the propagation path. On some aircraft detections using WSPRnet data more than one aircraft might be involved or contribute. Rob has detected aircraft far away and the SNR is additionally enhanced by aircraft in the vicinity of the transmitter and/or the receiver.
GDTAAA currently uses a SNR signal level anomaly of one standard deviation based on an algorithm which includes the distance between transmitter and receiver, the power level and frequency of the transmission and a safe value for the noise floor level. It is clear that anomalies can be detected at much lower levels. GDTAAA is a powerful sensor, with the gain turned down. At the moment, I feel that I am using an elephant gun to shoot flies.
re “At the moment, I feel that I am using an elephant gun to shoot flies, ” it might be that the information gathered about the missing passenger aircraft, in this case MH370, could be amplified by, for example, information about the thermal, acoustic and electromagnetic properties of Rolls-Royce jet engines.
Rolls-Royce have been utilising the current aviation downturn to improve the Trent series, so this data might already be available, or it may be that information gained from the latest tests can be usefully retro-applied to the discontinued Trent-800 engines that powered MH370.
Alternatively, it may be that Rolls-Royce, or an airline that is still operating a Boeing 777-200ER with Trent 800 engines, would be able to perform specific tests to gather precise data for that engine on that model at various speeds, height, thrust , load etc.
You might then be able to use these engine signatures to help refine and consolidate the WSPR tracking.
At the other end of the spectrum, it may be that different radio enthusiasts around the world picked up the emergency communication between MH370 and MH386, and that between them they may be able to help, together with modern acoustic analysis and decryption technology, to “unjumble the mumbling”.
Depending what engines were on MH386, it might be that it too, and the interflight conversation on the emergency channel between MH370 and MH386, can be isoated from the plethora of airline and other communications and interferences in the period just before and just after MH370 stopped registering on civilian radar.
The official report from Malaysia’s Ministry Of Transport cites Shanghai as the destination of MH386, and the Straits Times on March 9th 2014 cites Narita, Japan as the destination of a flight on which the pilot made contact with MH370, so there may have been more than one emergency channel contact made with MH370, by more than one pilot.
WSPR, GDTAAA, and enthusiasts of other radio-television-satellite communications technologies may be helpful in resolving this.
http://mh370.mot.gov.my/MH370SafetyInvestigationReport.pdf (pages 245 and 300)
(By the way, there appears to be a temporal sequencing error between items 20 and 21 on page 245 of the accident report, which could inadvertently cause confusion or
misinterpretation to WSPR and other data analysts)
The search for MH370 using WSPRnet data is starting to generate interest in the radio amateur community. Many thanks to Jean-Marc Garot for pointing out this article in the Southgate Radio Amateur News:
John Willliam VK4JJW has an update on how WSPR, the Weak Signal Propagation Reporting remains a key source of hope in the search for the wreckage of the missing airliner.
WSPR is undergoing some refinements to help in the search for Malaysia Airlines Flight 370, which crashed more than seven years ago in the Indian Ocean while enroute to Beijing. The low-power digital communication protocol, used by radio amateurs to test propagation, is now being employed by aerospace engineer Richard Godfrey in conjunction with a system he developed known as Global Detection and Tracking of Aircraft Anywhere Anytime, or GDTAAA.
There will be some preliminary tests in conjunction with Qantas airliner data before a different blind test is conducted later this year using the Malaysia Airlines data. The goal is to see whether tracking with help from the GDTAAA system can be more successful this time around.
According to an article in AirlineRatings.com, the tests will take place in October and November with an eye toward ultimately finding the exact crash location. Two separate searches for wreckage after the 2014 crash came up empty, although more than 30 pieces of debris washed up in various places.
While the principal dazzle test neglected to follow the airplane absolutely from Johannesburg to Perth because of a computation blunder in planning the WSPRnet connect way all throughout the planet the mistake has now been rectified and a subsequent visually impaired test will begin this week.
@Susan Craddock (aka Sarah Harris)
If you have a scientific contribution to the discussion on MH370, then you are welcome to this blog.