Extracting tracks of individual planes from recorded radar data
 

I would like to understand the methods and mechanisms by which an individual plane's track may be extracted from recorded radar data.

My motivation is trying to find an explanation for the "sharp corner" in MH370's track during its turn back minutes after going dark.

It seems obvious, that a B777 cannot instantaneously change course and create a sharp corner track. The most common answer, I so far received, is "its an artefact". However, no one would give a more detailed explanation of what kind of artefact it may be.

- did the graphic artist mess up?
- was there a mismatch between several radar data sources, creating a sharp corner when grafted together?
- did any automated algorithm "sanitize" the track and eliminate loops, dangling tails, etc.?
- is the corner due to the vertical planar projection of a three dimensional flight path?
- etc.

I would be grateful for answers to the following, and possibly more, questions:

- what radar return data is routinely stored? E.g. Range, bearing, altitude, plane ID, time stamp, etc.
- at what frequency is data stored?
- is the extraction of tracks performed manually or by automated, algorithmic means?
- if automated, are details of the algorithms publicly available? Manufacturers, source code, etc?
- what is stored, if the plane's ID is no longer available?

I'll probably have many more questions, but let's start there.

Cheers
Will

A possible clarification
 

I think the post opening this thread needs some clarification.

The MH370 saga was discussed a lot yet somehow the new published radar track and the mysterious "Sharp Corner" in it got little attention.


The ATSB report contains a Fig. 2 titled "MH370 flight path derived from primary and secondary radar data":

http://www.atsb.gov.au/media/5243942..._18aug2014.pdf

The source of Fig. 2 is the JIT, the official Joint Investigation Team located in Malaysia and comprising of experts from Malaysia, China, US, UK and France.

Fig. 2 is high resolution and you should examine it under at least 400% magnification. A very interesting feature is the very sharp 110 degrees corner above the word "Last" in "Last secondary radar data 1722".

Simple calculations show that such a sharp corner can't be a level flight turn even with bank angle of 67 degrees and higher. It could indicate however a kind of Split-S, an unusual aerobatic maneuver for a Boeing 777. There are other more esoteric explanations possible, e.g. ID duplication.

The thread poster is apparently asking for an ATC technical insight about this sharp corner hoping it'll throw some light on the MH370 mystery.

Some possible notes:

* some of the radar data processing is probably done in real time so concentrating on the extraction of data from recordings may miss it

* under high magnification (e.g. 1600%) it seems the data points density is higher than an ordinary SSR so there may be ADS-B data included too

* considering the ranges involved it seems the relevant data could come from Vietnamese SSR and ADS-B but there may be Malaysian data too

* It's possible other data e.g. ADS-C, ELINT was included too

It's also possible that track smoothing is being seen, a feature of processed radar pictures. It is used to prevent track jitter causing targets to jump around on the display, by predicting where a flight will be rather than where the raw radar return says it is, but it can also mask sharp turns until it decides the manoeuvre is credible.

The acute 110° turn to port could have been a lazy 250° turn to starboard with the intermediate points missing.

A bit like this graphic published by Flight's Air Transport Editor after BA762's cowls fell off, which omitted the figure-of-eight:

https://dn3pm25xmtlyu.cloudfront.net...VGSUJFNRFZBBTA

Thanks Ron for your clarification.

As you suggested off line, I post links to the close up image:

https://www.dropbox.com/s/gx20mul24o...aleV3.pdf?dl=0

The left image is the turn back near IGARI, the right a similarly sharp, but plausible, corner soon after take off.

And the calcs relating the circles to turn radii of different bank angles:

https://www.dropbox.com/s/3hhc8toqwx...Calcs.pdf?dl=0

Also thanks 10W and Dave Reid for your responses.

Re track smoothing, is that an automated processing feature performed by software? If so, is this step performed prior to recording the data or during the extraction of data from the records in the database?

Re Dave's mentioning of "editor", are you referring to a human putting the graphic together manually and while doing so ommitting the second loop of the figure of eight?

Cheers
Will

No, simply that it appeared on D K-M's Twitpic:

http://twitpic.com/csse7a

I doubt he would have done any manipulation of the image, which looks like it has originated from one of the flight-tracking websites.