38 thoughts on “Background Information on the Pinging of MH370 by Inmarsat-3F1”

1. Hi Duncan,

   First, great work. It’s been fascinating reading and watching the discussion, as a voyeur .

   Had you done any work to cross-check the derived LOS-Speeds as posted here in other threads (derived from the BFO reverse-engineered data). I can find the link here again if you are unsure of the one i’m referring.

   For known times where locations/heading/velocities are all known from the ADS-B data available, I’m unable to reconcile those LOS Speed values to what i calculate at those times and data points (specifically the first 4 times/rates) .

   I will refer to my values as Range-Rate and I will continue to use LOS Speed to refer to your posted derived values as that is how you have described them and to also disambiguate any references to one or the other sets of values.

   Time (UTC) LOS (km/sec) Range-Rate (km/sec)
   16:30 -0.0003410 0.001034 – plane stationary
   16:43 0.0074430 -0.025838 – heading 327, climbing
   16:55 0.0149360 0.069943 – heading 25, climbing
   17:07 0.0110220 0.085786 – heading 25, cruise alt/vel

   16:30 MH370 is stationary at the gate, 3F1’s sub-satelite point is moving north, and slightly west, while 3F1 is moving further away from the earth vertically. My calculation show a positive range-rate, counter to the LOS Speed.

   16:43 is more nuanced. At the :28 second mark A right turn has already begun from heading 328 (radially towards 3F1) to heading 25 (radially away from 3F1). By :41 seconds the turn has progress far enough that the radial component has crossed from towards 3F1 to away from 3F1.
   Maximum negative range-rate would be between :05 and :28 when the right turn starts, approximately -0.047041
   The zero point would be at approximately the :40 point
   Maximum positive range-rate would be at the end of the minute, approximately 0.020787
   The LOS Speed is not conclusively counter to my calculated range-rate here, but as i think there’s an implicit assumption there that the ping times are start of the minute (for lack of anything better or definitive), and the :00 value doesn’t compare well.

   The other two are boring by comparison, first still climbing to cruise altitude, the other at cruise altitude and speed, but both at a constant heading 25. But equally, these don’t compare well to my calculated range-rate. And the 16:55 LOS Speed is inexplicably higher despite have same heading away from the satellite and slower forward velocity by about 40kts (albeit at a lower altitude and still climbing to cruise 35000 feet, that i’m not sure would have a significant impact)

   If i’m correct in this I don’t know if there are any implications on any assumptions either explicit or hidden that may be present in the method by which the derived LOS Speeds were obtained out of the reverse-engineered BFO data, or even the BFO data itself either as-released or altogether.

   I admit freely I have not yet dug into the works on those reverse-engineering efforts by you and your collaborators. I don’t intend to mean any of this to be interpreted as disparaging the great works that any of you have all done here, it is all great effort, and greatly appreciated. I apologize if that is how this has come across.

2. I’m having trouble reconciling the Inmarsat BFO Analysis chart (450 knots) with what we know of the aircraft movements from the ATC transcripts. Partly this is because I think it is incorrect to connect the ping points with straight lines [as I think Michael Exner points out on his charts]. There is information that we know that must indicate variations from a straight line between some of these early pings.
   (i) At 16.30 the aircraft would have been taxiing to hold B ready to enter runway 32R
   (ii) At 16.40.38 the aircraft is ready in position on 32R and is cleared for takeoff, so at exactly that point the aircraft will be stationary
   (iii) When established in the climb the track will still be 320deg, and then at 16.42.38 the aircraft is given a clearance to track direct to IGARI. The aircraft will then turn right on to a heading of approximately 024 deg, continuing to climb to FL 350. I presume the course change ought to result in a discontinuity in the BFO line at that time.
   (iv) I think we know for certain that shortly after the sign-off at 17.19 the aircraft changed direction again, this time onto a heading of about 244 deg for Penang, or 263 deg for VAMPI. At that time then the Doppler ought to reverse.
   (v) So referring now to the Inmarsat chart again, there is a big gap in our understanding of exactly what the BFO, and hence the Doppler shifts might have been between 17.00 and say 18.28, when the Inmarsat chart suggests a possible turn. It surely cannot be a straight line.
   (vi) Hence I also wonder what might have been occurring between 18.30 and 19.30 on the Inmarsat chart, and why the “predicted” tracks show no significant discontinuities out to about 18.30.

   1. Thanks Brian. You have given a good example of why I have slammed that Inmarsat chart for the drawing in of straight lines, which are misleading.

      I repeat, just so those from outside science appreciate the dumbness of that mistake: it’s the sort of thing one would fail a first-year physics undergraduate write-up for including, so as to ensure the student understands it’s a daft thing to do, so don’t do it again.

      Elsewhere I wrote that it’s like seeing someone in a cafe having a breakfast coffee, and in the evening having an after-work beer, and then assuming that he must have been there all day.

3. A few comments from the jungle:

   Speculation about the frequency stability of the S/C LO elsewhere is unnecessary. I’m pretty sure it is a TCXO or OCTCXO (certainly not a simple crystal oscillator) with excellent LTS and STS, but it does not matter. The frequency is effectively calibrated from the LES using pilot carriers and a Rubidium standard, or a GPS locked standard, so don’t waste your time trying to reverse engineer the s/c. Just take Inmarsat at their word, and assume they are not idiots. When they say the BFO is the Expected-Observed, just go with that. Moreover, I am quite sure the AES correction is known in real time at the ground, and taken into account when the “expected value” is calculated. Otherwise, they would not have been able to perform any analysis.

   The C bad correction is all that Duncan needs. My spreadsheet has the provisions to separate the L band Doppler into S/C and A/C components, but the TOTAL L band Doppler is all Duncan needs. The S/C induced L band Doppler ESTIMATE is known to depend on very rough (poor) estimate of the aircraft position. In fact, I fully understand why the S/C component is underestimated. It is due to the fact that I used the airport RRs for the 4-12 transmissions. It is just a placeholder until we have real numbers, or estimates from Duncan’s Model. The first three are based on hard (ADS-B) data, and last 9 are based on a simple assumption to test the model, not solve for the exact s/c component. As I have explained to Duncan, as soon as he has good estimates for the L band RR to the 9 unknown LOPs (not the aircraft), then we will have fairly accurate separate components of the L band Doppler. All that said, I repeat, the TOTAL L band Doppler is a solid value, unaffected by S/C LO uncertainty, and Duncan can use this to constrain the solution. It does not need to be separated to find MH370.

   1. Thanks Mike.

      Comments there will (I hope) satisfy various questioners.

      Having written that, all assumptions made (by me and all others) could and should be questioned (in the fullness of time). At this stage it’s a matter of concentrating on the things about which we have lesser confidence.

      Duncan

4. Sorry, Karl, but you are out with the pixies.

   The elevation angle is measured from the ping time delays, which render a satellite-aircraft range in km, and then knowing the instantaneous geometry (the size of the Earth, the satellite altitude) the range renders a ping ring which has a certain elevation angle above the horizon where the aircraft happens to have been at the time of the ping. This was the whole basis of the initial analysis! (Vide the Inmarsat graph issued by the Malaysian Government on March 15th).

   Go take a lie down my friend.

5. Based on your analysis and my experience as an international airline captain, something seems obvious. We should be looking for the aircraft at 30S 90E and then other positions like 40S 90E and etc. Pilots typically fly to simple coordinates when ocean flying and I think this should be no different. I think we should also consider that since no debris field has been found the aircraft probably sank intact which means it was ditched. And since it appears it was ditched intact, we must consider that whoever did this survived the ditching and then drowned later or was picked up by a boat. Sounds incredible but simple logic, don’t you think?

6. Mike’s decomposition is inconsistent with Inmarsat’s reconstruction of southern tracks.

   In Mike’s decomposition, MH370 has a higher Doppler at 19:40 UTC than in all points up to and including 17:07 UTC. We know that, at 17:07, MH370 was east-northeast of the satellite and moving north-northeast (25 degree heading.).Inmarsat’s southern tracks put MH370 near the equator (east of the satellite) heading south both at 400 and 450 kt. It is impossible to have a higher LOS speed at 19:40 than at 17:07, even accounting for satellite motion.

   Any valid decomposition must have the 19:40 point near zero and the trend line through 19:40 and 20:40 crossing zero some time shortly before 19:40.

   1. It is possible that Inmarsat’s reconstruction of southern tracks is wrong. However, I doubt it. First of all, Inmarsat actually knows how BFOs should be decomposed, but Mike, as far as I can tell, tries to reverse engineer it.

      But consider also this. If MH370 was on the southern track, and it had a higher LOS speed at 19:40 than at 17:07, and since we know that it was going at the heading 25 at 17:07 (and it was much further away from the equator at 17:07), at 19:40 it had to be going in the direction that was at least 25 degrees off due south (that is, heading 155 or less).

      If you draw a route that starts at last known coordinates and goes at, say, heading 150, it puts the final resting place of the aircraft *inside* Australia at all speeds above ~340 kts. Which is, needless to say, pretty hard to believe.

      I’m sure that you can run the numbers in a full simulation and you’ll come to the same conclusion.

      I’m all up for Inmarsat-bashing, but, unless there’s a bulletproof justification behind Mike’s decomposition (with references to specs & such), I’m inclined to think that he got the decomposition wrong, or maybe that the BFO chart we’re all trying to interpret already has the C band Doppler subtracted.

7. You state “But the autopilot (I am told by aviation people including some very experienced B777 pilots) would have been set, one expects, to follow: (i) A constant speed and altitude (we know neither for sure); and (ii) On a direction that has been set as being either a great circle route to or towards some waypoint (e.g. the South Pole) or following a specified magnetic bearing (which varies with latitude and longitude). The first of those (great circle) you can think of as being a straight line, except that the Earth is not flat; a great circle is the shortest distance between two points. The second would mean that the aircraft would keep changing its heading slightly, from minute to minute.”

   Should the two options be either a great circle or a rhumb line?

   http://en.wikipedia.org/wiki/Great_circle

   http://en.wikipedia.org/wiki/Rhumb_line

   If the plane was following a rhumb line, it would maintain a constant bearing relative to meridian lines. But that would not make any allowance for crosswinds. An extreme example of point can be shown by watching crosswind landings on YouTube. Where the plane is pointing and where it is going are not necessarily the same.

   The other question is which party initiated the pings? Some are apparently initiated by the plane, while the last few, evenly spaced (at 60 minute intervals) are seemingly from the satellite.

   Those three pings in a 4 minute interval at 1105, 1107 & 1109 seem to be plane initiated. The following handshake is the first of the evenly spaced 1180, 1240, 1300 & 1360 handshakes.

   The plane seems to initiate handshakes when major scheduling events take place (gear up, reaching cruise altitude). So the three pings likely indicate some other predictable events normally reported (e.g. beginning descent for landing). Somebody ought to ask Inmarsat to explain who initiated those pings and did they convey any flight plan checkpoints?

   1. Bruce Thompson wrote: “(ii) On a direction that has been set as being either a great circle route . . .” I believe this in only partly true. An autopilot has a gyro as its primary keep-the-wings-level input. Added to this is whatever guidance is coming in from sensors: GPS, land-based navaids (VOR), inertial guidance, or a heading bug on the directional gyro (“fly heading 260 degrees”). On a broken, fire-ravaged aircraft, one can easily believe the various communications and navigation electronics fail over a short time. Once the autopilot’s guidance input becomes compromised, it’s flight path is likely to be anything but a straight line.

      1. Duncan,
         I don’t run a PC that I can run STK (all linux & mobile now) so haven’t been able to shadow your work using the same app – I wasn’t aware that it couldn’t model simple headings. Using HDG Sel as an input to the AFDS is a common practice to quickly change course in advance of programming a waypoint or route intercept on the FMC (or simply flying off to nowhere….).
         Interested to know what the difference in computed values might be for gc vs hdg, say, for the ‘pings’ at 20:40 & 21:40?

   2. From Inmarsat’s public explanation:

      “[I]f the ground station does not hear from an aircraft for an hour it will transmit a ‘log on/log off’ message – a ‘ping’ – and the aircraft automatically returns a short message indicating that it is still logged on, a process described as a ‘handshake’.

      The ground station log recorded six complete handshakes after ACARS, the aircraft’s operational communications system, stopped sending messages.”

      From: http://www.inmarsat.com/news/malaysian-government-publishes-mh370-details-uk-aaib/

      So the first four points on the graph (slide 3) would be signals sent from the aircraft to the satellite transmitting ACARS messages. The six other points represent handshakes, as explained by the quote.

      Regarding the OP, the PDF is still accessible from the link in this comment. At the bottom of the right column under “Related” is a link to “Inmarsat Differential Doppler Study (PDF)”. That link is the 4 slide presentation shown.

8. Hi,

   I’m puzzled ….

   My understanding of Mike Exners charts is that the BFO in the inmarsat graph actually represent the Total Doppler. From there Mike subtracts D3 (C-Band Doppler) and D2 (S/C induced L-Band Doppler) and gets to D1. The red circle in the inmarsat slide suggests somehow that only D2 is plotted in the inmarsat graph while actually the Total Doppler appears to be plotted.

   The D2 contribution (S/C induced L-Band Doppler) is very small i.e. in the range of 5Hz. With my primitive vector algebra and the sat position/velocity data Duncan provided, I calculated LOS speed (S/C) – (position of A/C) :

   a) 0.0010 km/s at 16:30 at the ramp in KUL (=> 5Hz L-Band doppler)
   b) -0.0087 km/s at 0:11 at an estimated final position -32° 92° somewhere close to the Australian search area west of Perth.

   b) also is a good match for the LOS velocity S/C – Perth at 0:11 which I calculate as -0.0084 km/s. That LOS velocity matches the C-Band doppler of 100Hz Mike puts for that time in his graph.

   Maybe Duncan and Mike can double check the absolute LOS velocity of the S/C in the direction of the aircraft’s estimated position for times after 19:30 UTC (3:30 local). According to my calculations the S/C induced L-Band doppler should rise to ~40Hz and not to only ~5Hz as per Mike’s graph. That is backed up by the consideration that the C-Band doppler rises to 100Hz (=>8.4 m/s LOS velocity). Because at the end of the flight aircraft position is close to Perth, the LOS velocity (S/C )- (position of A/C) should be similar to LOS velocity S/C – Perth .

   The next mystery: According to Mike’s conclusions the asymmetries between north and south track can only be introduced by the S/C motion and position. Everything else is independent of the aircraft’s position. At the times when the S/C induced L-Band Doppler is smallest in Mike’s graph, the asymmetries between predicted north and south track are biggest. How can that be?

   Regards
   Ole

   P.S.:
   My way to calculate the LOS velocities of the S/C is:
   – Convert lat,long,alt to ECEF
   – get LOS vector by subtracting pos vector from sat_pos vector in ECEF coordinates
   – normalize LOS vector
   – calculate inner product of normalized LOS_vector * sat_velocity vector.

   1. Thanks Ole:

      (1) “at 0:11 at an estimated final position -32° 92° somewhere close to the Australian search area west of Perth.” – We do not KNOW that; we do not KNOW that it took a southerly course.
      (2) ” Because at the end of the flight aircraft position is close to Perth, the LOS velocity (S/C )- (position of A/C) should be similar to LOS velocity S/C – Perth .” – see above.
      (3) Quoting myself: “Please, I am simply unable to explain anything further about these plots. If you don’t understand them, apologies. ”

      Haelsingar,

      Duncan

9. Hello,

   please take into consideration this image of the flight path up to 2:08 Malaysian Time, it was shown to the Chinese relatives.

   http://www.themalaymailonline.com/uploads/gallery/2014-03/mh370_china_screen_press_briefing_20140322.JPG

   1. Stefan: Thanks for the message.

      My answer: No, because I do not KNOW this path to be correct. I am working only on what I know, or think I know, is true.

      Duncan

10. Hi, interesting site although I don’t understand it!! One thing I don’t understand, shouldn’t your flightpaths follow the published radar tracks?
    http://ogleearth.com/2014/03/mh370-updated-route-final-radar-plot/

    1. Answer = fundamentally yes, and that is what I have been addressing in the past 24 hours or so.

       Thanks,

       Duncan

11. Duncan,

    It occurs to me that one could derive the ping ranging from the red track and then check to see if the sub satellite issue has been covered by Inmarsat by trying to reproduce the yellow track with and without satellite motion. Might be too degenerate, or the image not detailed enough but perhaps it is worth a try.

    1. Hi Chris.

       In fact a back-engineering of the ping rings from that fuzzy Google Earth image (have those people no shame?) has been done in a quite beautiful study which I recommend all to look at:
       http://www.reddit.com/r/MH370/comments/21xw2a/mh370_reverse_engineered_ping_data/
       http://www.reddit.com/r/MH370/comments/21jla4/mh370_flight_waypoints_timing_and_speed/

       Right now I am putting the ping ring information (for different times) into my STK scenario so as then to search for possible paths that fit.

       Cheers
       Duncan

       1. You can recover quite precise information from such images. It is not hard to pin down the centers of those tracks and also fit the coordinate grid to great precision owing to its mathematical nature. That is some pretty nice work going on there in that link.

          Here is a question that might be revealing: in the last three pings, there are three ranges from the satellite and three LOS velocities relative to the satellite, so six independent data points. If we assume a great circle course, is there a unique solution in each hemisphere for speed, heading and position that can be derived? If so, that may give rather specific information on where to look. The red track seems to indicate that changing speed required avoiding a great circle course, which might be an odd way to fly.

          If thee pings is enough, then trying again with the three prior to those and checking to see if the dots can be connected between the third and fourth pings back at constant speed might be a test of the constant speed assumption in the South, though working with the sixth ping back may be problematic.

       2. Chris: Well spotted.

          Basic answer is yes, and that’s what I am now trying to do. But first I had to get the initial (pre-18:22 UTC) part of the flight path correct (see other answers). All takes time.

          Thanks
          Duncan

       3. Glad you were able to use the data, and great work here. Your satellite data was crucial to the analysis of course. If I get time I will check that the yellow track BFO calculation matches their published estimate.

       4. GlobusMax you are a star.

          Great work, great write-ups. Exemplary.

          Many thanks,

          Duncan

    2. Also: Inmarsat *did* use the satellite motion in this later study (March 23). It was their earlier attempt leading to the map issued by the Malaysian Government on March 15 that assumed a truly ‘geostationary’ satellite and started me on this whole chase… 😎

12. Thank you for the above. I really don’t understand all the technical know-how \or know-what about the loss of the air craft. But can I ask one Q.

    If this plane was on a mission to no where, with no guidance, since it seems it was transmitting nothing, so assume it couldn’t also receive anything.

    Why would it be assumed << probably not the right word that is kept to a 'air corridor' why not just fly off randomly to anywhere?

    1. Thanks for that, Lea. Honest questions (like your own) are good.

       Actually, the aircraft could transmit one thing (or perhaps I should say, *did* transmit one thing). And that was responses to pings, or handshakes. Computers do just this over the internet: it’s like a brief signal that asks “Are you there?” and if an automated response comes back (from the aircraft, here), then it means it is still “there” in that it is functioning and ‘conscious’ in an electronic sense.

       All the analyses you will find here is about the pings. The point is that the ping returns (to the satellite, thence the satellite ground station) can tell us two things about the aircraft at the times of the pings: (a) The distance between the satellite and the aircraft, from the time delay in the ping (due to the travel time of the radio signal); and (b) The relative speed of the satellite and the aircraft, from the Doppler shift in the radio frequency; look that up on Wikipedia (Doppler effect). In the case of *sound* (not radio waves) you hear it often, for example the change in the pitch of a siren as an ambulance whizzes by.

       That covers your second sentence. Now the third one.

       An assumption being made is that the aircraft flew on under the control of an autopilot with no ongoing human control, for whatever reason. That may have been a default of the B777, it may have been an autopilot setting entered by the crew, we don’t know. But the autopilot (I am told by aviation people including some very experienced B777 pilots) would have been set, one expects, to follow: (i) A constant speed and altitude (we know neither for sure); and (ii) On a direction that has been set as being either a great circle route to or towards some waypoint (e.g. the South Pole) or following a specified magnetic bearing (which varies with latitude and longitude). The first of those (great circle) you can think of as being a straight line, except that the Earth is not flat; a great circle is the shortest distance between two points. The second would mean that the aircraft would keep changing its heading slightly, from minute to minute.

       My understanding is that during long flights the autopilot is flying the plane for most of the time. With two pilots (captain plus co-pilot) one would be resting/eating/whatever whilst the other was in the seat in front of the controls but not with his/her hands and feet doing anything. He/she would just be checking progress.

       Whatever route the aircraft *did* take, the intent in the analysis here is to determine that route based on the information available from the pings, and the known locations of the aircraft earlier in the flight.

       Hope that answers your questions,

       Cheers,
       Duncan

       1. Many thanks. I had gathered the bit about the pings (a bit like cookies on the computer) and that this was how it was tracked to have taken the Southern Corridor (I believe that is what is is called).

          So the bit about the Autopilot makes complete sense. But (sorry always a but) thought that if all systems were down\disabled – would indicate surely amiss with the aircraft.

          Then it begs the Q – no need to answer 😉 Why did the autopilot continue to work, i.e. part of the computer system, but all else apparently failed.

          Again, many thanks.

       2. Re: autopilots and magnetic settings, there is a HDG REF button on the console below the navigation display. If it is set to NORM, the 777 will fly the last magnetic heading of the last waypoint when it passes it; but the pilot can set the switch to TRUE, in which case it will follow a true heading/course after passing the last waypoint.

          I suspect that a constant magnetic heading/ constant speed course might be inconsistent with the ping data, but I haven’t been able to prove this yet. Additionally, the 450 knot Inmarsat appears to follow a constant true heading of approximately 187 for its last, long leg. This could be possible if HDG REF was set to TRUE.

          Meanwhile, slower courses gradually drift to the left. It can be shown by inspection that the 400 knot Inmarsat path is NOT a constant magnetic heading course, nor is a great circle path, nor is a constant true heading course. In other words, it not a normal course. It is a course that looks like it could be flying inertially, under the influence of crosswinds and some Coriolis drift. This would require a pilot flying blind, just trying to keep it wings level with no other navigational cues.

          IOW, only fairly high speed scenarios, such as the 450 knot path are (a) what a normal flight path looks like; and be consistent with the ping data.

          Which leads me to a question: The Inmarsat paths show the 400 knot path making a wide swing around Sumatra–which is consistent with Indonesia’s insistence that it did not cross its airspace. Which favors the 400 knot path. Thus the question is could one have a 450 knot path that makes the wide swing, and is consistent with the ping data, yet produce a normal looking course?

13. Duncan, I have just posted my long ranty *alternative* take on the ‘D1 D2 D3’ diagram at TMF Associates.

14. Many thanks.
    As I understand Total Doppler received at LES is just sum of those Dopplers at every step:
    C band Doppler (3615MHz) + S/C induced Doppler (1643MHz) + Total L band Doppler + Net AER Doppler
    The last is the one of our interest.

    And here is one thing to point… when you calculate Doppler shift for part of recorded track with 900Mhz of initial frequency it fits almost perfectly presented BFOs of Net AER Doppler.
    At least with Hamster3null’s spreadsheet.

    1. AndRand said,
       April 3, 2014 at 12:41 am
       if you use 900Mhz of AES frequency to calculate doppler shift it quite matches the BFOs from the recorded part of track.

       I have to update this statement: it fits the planned flightlog https://flightaware.com/live/flight/MAS370/history/20140307/1635Z/WMKK/ZBAA/tracklog
       With flightlog updated to this guestimation http://skyvector.com/?ll=4.746514696181505,98.71435545889493&chart=302&zoom=9&plan=A.WM.WMKK:F.WM.GUNBO:G.5.7530039670591675,103.1850585976314:F.WS.IGARI:F.WM.GUNIP:F.WM.VAMPI there is significant difference:

       [img]http://i.imgur.com/LAUjjoR.png[/img]

       My version of Hamster3null spreadshet here: https://docs.google.com/spreadsheet/ccc?key=0AhvpxNRGOuapdERwQ0hPdzlnVDE4djRBWVM2dURVMXc&usp=sharing

15. You can still find the PDF (or something very like it) here:

    http://www.malaysiaairlines.com/content/dam/mas/master/en/pdf/Annex_I_images.pdf

    There is also the Inmarsat ‘Newsroom’ page which essentially reproduces the same information here:

    http://www.inmarsat.com/news/malaysian-government-publishes-mh370-details-uk-aaib/

    1. As I wrote, I am sure that everyone has seen the later 3 pages already; but not the first (cover) page.

       1. Agreed that it is potentially interesting that the ‘Inmarsat cover page’ appears to have been removed. Sorry about the duplicate link, I was just trying to give a bit of wider context.

       2. My intent in putting that in is more a matter is establishing a chronology as this sorry saga continues.

          The guys were working on a Sunday, credit to them; similarly the information was conveyed from AAIB to the Malaysian Government promptly.

       3. Ah – I get you – the *date* is on the cover page. Yes – I see what you mean (I was thinking that it might be something to do with the ‘politics of attribution’ and/or not wanting the word ‘Inmarsat’ to appear on it for some reason – though why that would be I can’t think – your explanation is much simpler!)

Comments are closed.