Cambridge arrival having a *very* good look
Thread Starter
Cambridge arrival having a *very* good look
https://www.flightradar24.com/data/f...a8235#22f7400f
This morning's BA8235 seemed to want four flyovers before landing. Normal? It was a bit over-cast so not sure how much they could see.
This morning's BA8235 seemed to want four flyovers before landing. Normal? It was a bit over-cast so not sure how much they could see.
One could say that they certainly needed some practice flying the holding pattern 
But unless they were afterburner equipped (given the speed spikes by moments), one can conclude the FR24 speed & trajectory was somewhat unreliable. So these chaps knew perfectly well what they were doing.
And for whatever reason they went into the holding, it was not weather related. Good visibility, a touch of wind and a bit of not too low clouds. And they could certainly see more than enough For the METAR lovers:
25/11/19 08:50 EGSC 250850Z 15008KT 9000 SCT018 BKN030 10/09 Q1003
25/11/19 08:20 EGSC 250820Z 14007KT 9000 SCT018 BKN030 09/08 Q1003
But unless they were afterburner equipped (given the speed spikes by moments), one can conclude the FR24 speed & trajectory was somewhat unreliable. So these chaps knew perfectly well what they were doing.And for whatever reason they went into the holding, it was not weather related. Good visibility, a touch of wind and a bit of not too low clouds. And they could certainly see more than enough
For the METAR lovers:25/11/19 08:50 EGSC 250850Z 15008KT 9000 SCT018 BKN030 10/09 Q1003
25/11/19 08:20 EGSC 250820Z 14007KT 9000 SCT018 BKN030 09/08 Q1003
If you look at the FR24 flight paths, it didn't even land at Cambridge airport and it took off for Goteborg from a different part of Cambridge, also not the airport.
Paxing All Over The World
FR 24 does not smooth the turns but only the points at which the signal changes sufficiently to indicate a turn. FR24 - and other sites - and NOT radar!
In my experience, FR24 plots are typically accurate points on an aircraft's track, several seconds (or more) apart, but the trajectory between successive points is undefined. It's when sites (including FR24 itself) simply join-the-dots that those impossible tracks and instantaneous 90° turns start to appear on plots.
A good example being yesterday when a flight approaching LHR , I think from Istanbul, separated into two as it turned south off the easterly arrival track north of London and joined itself back into one again as it became established for Clearly not what actually happened and great as FR24 is it does have these moments of madness,
Avoid imitations
Join Date: Nov 2000
Location: Wandering the FIR and cyberspace often at highly unsociable times
Posts: 14,573
Received 422 Likes
on
222 Posts
I agree that FR24 can be wildly inaccurate. I recently checked what I had been up to at work and it reckoned I'd been flying at well over 300 kts.
I can say without any hesitation that I've certainly never flown that fast in a helicopter.
I can say without any hesitation that I've certainly never flown that fast in a helicopter.
Presumably not in the case of your aircraft ?
Mr Reid, I am sure you know the answer, but I was looking at FR24 this afternoon and there were some interesting LHR movements consisting of Piper Cherokees and a Cessna . Similar ones from/to LGW, is this because they are visible to LHR or LGW radar and therefor FR24 shows them as headed there since clearly they were not going to actually use either airport .
Tks PB
Tks PB
Mr Reid, I am sure you know the answer, but I was looking at FR24 this afternoon and there were some interesting LHR movements consisting of Piper Cherokees and a Cessna . Similar ones from/to LGW, is this because they are visible to LHR or LGW radar and therefor FR24 shows them as headed there since clearly they were not going to actually use either airport .
If you look at the replay for 12:20, Sonaca S200 G-LKAM can be seen on approach to Blackbushe at the end of a 40 minute sortie, with the route shown as N/A to LHR !
I have no idea why FR24 shows an obviously incorrect destination - I was under the impression that it usually deduced the origin and destination from the callsign/flight number, but it looks like there might be more to it than that.
Please people, stop talking about the tracking accuracy of FR24 as such. This is totally meaningless, unless you add in the discussion that the a/c concerned has ADSB-OUT (meaning pretty precise position, speed data, etc. is available + mode C altitude) or that no adbs was available (only mode C) and FR24 groundstations employed an MLAT (multi-lateration) technique to get a rough position indication (pretty impressive techniques & calculations used, nonetheless the result gives a rough indication at best) and a derived groundspeed and track, calculated with an algorithm that lacks good smoothing logic and certainly can use some improvement.
Depending on ground-receiver coverage and terrain, MLAT stops working below +/- 1500ft AGL at best (multiple receivers need to within line-of-sight). ADSB reception is also 'line-of-sight' but one receiver is sufficient. So with at least one FR24 receiver pretty close to an airport, many/most ground movements can be tracked.
Perfectly good ADSB data can still be screwed up by a 'misbehaving' FR24 groundstations. I once asked the FR24 guys to block out a specific groundstation, as it somehow was providing valid ADSB data with an incorrect, offset timestamp, undermining the reliability of data provided by other stations in the area.
And in both cases the 'data sampling' rate determines the end-result presented on your screen. In real-time trajectories, you see the position data 'updates' as a slightly brighter dot, so you have immediately an idea about the sampling rate. The lower the 'sampling rate', the more meaningless becomes the straight line connecting two adjacent dots and the more important becomes the human interpretation of the result (and with MLAT, human interpretation and common sense is always needed). With data 'replay', no dots are visible on the trajectory to easily asses the sampling rate. Only during the actual replay, the a/c symbol 'hopping around' will give some indication of the position points available. And whatever data available, the 'presentation' software of the FR24 website, has its limitations; when following a flight in close-up, real time, sometimes refreshing your browser page will improve the track drawn on your screen.
Regarding the other 'flightplan' data provided, the smaller the airport and the 'less scheduled' the flight, the more guesstimating becomes apparent.
Finally, FR24 data has no (direct) relation whatsoever with whatever radar coverage in the area (I guess, indirectly the interrogation rate by SSR radars and TCAS, will influence somewhat the broadcast rate of the transponders involved and maybe, very mayby, the end result in FR24).
Regarding this MAN-CBG flight, the tracking was MLAT based (with all its limitations).
Depending on ground-receiver coverage and terrain, MLAT stops working below +/- 1500ft AGL at best (multiple receivers need to within line-of-sight). ADSB reception is also 'line-of-sight' but one receiver is sufficient. So with at least one FR24 receiver pretty close to an airport, many/most ground movements can be tracked.
Perfectly good ADSB data can still be screwed up by a 'misbehaving' FR24 groundstations. I once asked the FR24 guys to block out a specific groundstation, as it somehow was providing valid ADSB data with an incorrect, offset timestamp, undermining the reliability of data provided by other stations in the area.
And in both cases the 'data sampling' rate determines the end-result presented on your screen. In real-time trajectories, you see the position data 'updates' as a slightly brighter dot, so you have immediately an idea about the sampling rate. The lower the 'sampling rate', the more meaningless becomes the straight line connecting two adjacent dots and the more important becomes the human interpretation of the result (and with MLAT, human interpretation and common sense is always needed). With data 'replay', no dots are visible on the trajectory to easily asses the sampling rate. Only during the actual replay, the a/c symbol 'hopping around' will give some indication of the position points available. And whatever data available, the 'presentation' software of the FR24 website, has its limitations; when following a flight in close-up, real time, sometimes refreshing your browser page will improve the track drawn on your screen.
Regarding the other 'flightplan' data provided, the smaller the airport and the 'less scheduled' the flight, the more guesstimating becomes apparent.
Finally, FR24 data has no (direct) relation whatsoever with whatever radar coverage in the area (I guess, indirectly the interrogation rate by SSR radars and TCAS, will influence somewhat the broadcast rate of the transponders involved and maybe, very mayby, the end result in FR24).
Regarding this MAN-CBG flight, the tracking was MLAT based (with all its limitations).
Please people, stop talking about the tracking accuracy of FR24 as such. This is totally meaningless, unless you add in the discussion that the a/c concerned has ADSB-OUT (meaning pretty precise position, speed data, etc. is available + mode C altitude) or that no adbs was available (only mode C) and FR24 groundstations employed an MLAT (multi-lateration) technique to get a rough position indication (pretty impressive techniques & calculations used, nonetheless the result gives a rough indication at best) and a derived groundspeed and track, calculated with an algorithm that lacks good smoothing logic and certainly can use some improvement.
Incidentally, I can probably claim to be among the first to use amateur MLAT (many years before FR24 and its counterparts) when I recruited a small group of fellow early SBS-1 users and captured raw data from the receivers including the 20 MHz clock timestamps, to derive the locations of a number of non-ADS-B aircraft. Having proved that the concept worked, I kind of lost interest in what was just an academic exercise ...
First time a question of mine has ever stumped you Dave. DIBOs post is very interesting and it underlines the fact that FR24 is very interesting and quite an achievement in terms of complexity but for all sorts of reasons has quite a few illogicalities and oddities.
