back to article Mysterious sat-pic China desert markings - EXPLAINED

The grids of white lines in China's Gobi desert that have got the world's conspiracy theorists in a lather for a week, are actually calibration targets used to help China's spy satellites, says a NASA researcher. Since the 65ft-wide white line patterns were spotted on Google Earth, it has been speculated that they were …


This topic is closed for new posts.
  1. Anonymous Coward
    Anonymous Coward

    Could Also Be

    ... a testing site for the optical targeting sensor of a missile.

    1. Aaron Em

      Dumbest thing I've heard all day

      Granted, the day is young, but still -- the only way what you've said could even possibly make sense is if you were talking about an ICBM, and the idea of optical targeting on an ICBM-delivered warhead is risible to say the least.

      Next time you're moved to comment, distract yourself with another bong hit or two instead. We'll all thank you for it.

      1. serviceWithASmile

        yeah ok, optical targeting *ALONE* on ICBMS might be a bit silly...

        no need to be a dick about it though.

        just saiyan.

        secondly, it's china. you think they'd really mind if their ICBMs weren't completely on target?

        when they inevitably take over the world all they have to do is *miss* china

        1. Aaron Em

          *ALONE* qualifier unnecessary

          Do a little reading. 'Fast' doesn't begin to describe the terminal profile of an ICBM-launched warhead, which is unguided -- hence "ballistic" -- anyway; even if someone were mad enough to build fins and optics into one, which they wouldn't because it'd be a waste of mass better used for payload or RV structure, it wouldn't have time to maneuver before fusing.

          As for the rest, it is every sensible person's responsibility to shame arrant ignorance out of blithely opening its uninformed mouth in public.

          1. Anonymous Coward
            Anonymous Coward

            As for the rest

            No, you're just a dick.

          2. hayseed

            Ballistic Warhead Targeting

            Tricks could be done shifting the aerodynamic response of a fast-moving shell by shifting around the center of mass, for instance - no fins needed (or wanted at that speed). In fact, terminal evasiveness is supposed to be a characteristic of warheads from the superpowers nowadays.

          3. hayseed

            Terminal Guidance Not Impossible

            Apparently the Indians can do it to.


            Agni. The 2500-kilometer Agni technology demonstrator uses the SLV-3

            booster for its first stage and a liquid-fueled Prithvi for its second

            stage. Three test shots were conducted before the U.S. successfully

            pressured India into suspending testing (1994). Of particular

            interest, the Agni tests demonstrated that India can develop a

            maneuvering warhead that incorporates endo-atmospheric evasive

            maneuvers and terminal guidance in the reentry vehicle.

          4. James O'Shea

            errm... Aaron, m'lad, you don't seem to have heard of 'maneuverable re-entry vehicles'. Look 'em up, Google is your friend. There _are_ RVs which can, and do, modify their flight path from a purely ballistic path, and apparently at least one such type of RV uses an optical system, not for targeting guidance, but to help evade possible intercepting missiles. (Google is your friend with that one, too.) Note that you are the only one who mentioned 'fins', which almost certainly wouldn't work at RV velocities. Methods which would work include gyroscopes and reaction control thruster packages. (Yes, really.) (Google is your friend there, too.)

            I leave as an exercise for the student the matter of exactly who should take a hit off his bong before ever again blithely opening its uninformed mouth in public.

          5. This post has been deleted by a moderator

          6. Stuart Castle Silver badge


            It's every sensible person's responsibility to point out that something is wrong. Not be a dick or take the piss.

  2. Aldous
    Thumb Up

    out of curiosity

    are any other nations grids visable. USA's will probably be buried at white sands or nevada witha blackout against google but what about uk, france etc?

    1. Fredrick Smith
      Thumb Up

      England has a big white horse.

    2. Anonymous Coward
      Anonymous Coward

      we just use someone else's.

    3. John McCallum

      Optical Targets

      This may sound a bit daft but why should the US make a special target site just use Manhatten Island most of the streets are in a gridand as for the UK Milton Keynes all those roundabouts must be good for something?

      1. Stevie Silver badge


        I've always assumed that Harlow New Town was designed as some sort of nuclear test site. Big sign on the outskirts announcing it as a "Nuclear Free Zone", yet begging for some sort of large, very hot event to remove it from the memory of mankind forever. Has to be irony from on high.

    4. Grease Monkey Silver badge

      Has it occurred to you that nations allied with each other probably share such things?

    5. Shades

      Does what now?

      "satellite calibration patterns" on US military bases in the UK which are visible on Google Earth. According to the US mil they are actually motorcycle training areas but that sounds far too reasonable... so let's not let the "truth" get in the way! :)

    6. Destroy All Monsters Silver badge

      Nah, the US calibrates by looking at car's license plates.

      1. Martin Huizing

        New word of the day: "Lanoitpo"

        The US calibrates during their bombing of overseas targets. Two birds one rocket...

    7. Anonymous Coward
      Anonymous Coward

      ah the secret is out about

      Hampton Court Maze, psst You didn't hear it from me.

    8. Lockwood

      There's a lovely diamond shaped object with a slice taken out of the top of it that the UK uses to calibrate their spy sats just off the South coast.

      The notch in the top is used to calibrate direction.

    9. Cameron Colley

      RE: out of curiosity

      I read a post on Google Sightseeing a while back about Abu Dhabi’s Sheikh Hamad bin Hamdan Al Nahyan having "HAMAD" in large letters using canals -- apparently a guy in the US called Luecke went even bigger and NASA use his tree-graffiti for satellite calibration.

  3. Anonymous Coward
    Anonymous Coward

    Go on then...

    Someone please explain why the weirdy pattern?

    1. Steve Knox

      Why the weirdy pattern.

      Because a regular pattern would be 100% useless for orientation purposes. To fully orient a camera in 3 dimensions from a two-dimensional image, you need an image that appears unique however it is rotated in any of those three dimensions, If there are any two rotations that create the same (or roughly similar depending on the effective resolution) two-dimensional image, then the camera can't be sure how it's oriented.

      It's also important that the image be unique in comparison to other subjects of the camera, so that, for example, a satellite camera doesn't mistakenly try to orient itself to the ridge patterns of a mountain range or the street patterns of a major city.

    2. pepper

      low res

      If its a low res image then its quite possibly easier to find its direction/rotation by the use of crazy angles that make it look unique.

  4. Aaron Em

    Chronic unemployment not without consolation

    After all, you can sit home all day doing bong hits and babbling conspiracy-theory nonsense about something some red-eyed twat found on Google Earth when he had nothing better in mind to do at three in the morning last Wednesday.

    Speaking as somebody who's been working his very testicles off for the last month without a break, I have to say the thought comes with a certain appeal, even if I would chop my own balls off with a cleaver in public before I'd sink so low as to haunt conspiracy theory forums.

    1. Stevie Silver badge


      It took you a whole month to paint those lines in the Gobi Desert? Just so the Chinese could test their missiles?

    2. Tom Maddox Silver badge

      Clearly . . .

      "Speaking as somebody who's been working his very testicles off for the last month without a break"

      . . . you've gone mad from overwork, because nothing you posted actually makes the least amount of goddamn sense.

    3. majorursa

      I'd say chop'm... quicker then working them off. Besides a dick like you doesn't need balls.

  5. Mike Richards Silver badge

    Spy satellites

    You should also add Israel, Japan and Russia to the list of countries operating spy satellites.

  6. bitmap animal

    Very straight edges..

    The thing that struck me when I first saw this a few days ago was how straight the edges were. It looks more like an overlay, some of the ground they are over is pretty rough so top marks to their painters!

  7. Michael H.F. Wilkinson
    Black Helicopters

    They just want you to think they are calibration patterns!

    where's my tinfoil hat!!

  8. Yet Another Anonymous coward Silver badge

    @out of curiosity

    Yes the UK uses a circular target with lines about the thickness of a road. For secrecy these are combined with the normal road network - the use of 1000s of them suggest a massive fleet of UK spy satelites.

    however the Hanger Lane gyratory is still believed to be a message to aliens

    1. Smudger 1

      ...and the M25

      is purportedly the Dread Sigil Odegra...

      1. Grease Monkey Silver badge

        I think you'll find that Good Omens is humourous fiction.

  9. Anonymous Coward
    Anonymous Coward


    the little girl and the clown?

  10. GettinSadda


    Seeing as how there are also a number of interesting painted "airport shapes" nearby, and some large squares that appear to be covered in craters, could this be to do with developing a system that can target bombs or shells based on visual cues rather than GPS? The crazy-paving patterns could be the start of getting a system to be able to track location based on city streets.,93.468995&spn=0.001611,0.003484&t=h&z=19&vpsrc=6

    1. Kevin McMurtrie Silver badge

      Or not craters

      The painter went around small rocks that were too big to pick up and sloppily covered the tops of larger rocks that could be walked on.

  11. Tegne
    Black Helicopters

    If china have their own spy sattellites...

    I'm surprised they haven't leaked photos of all the sensitive areas blocked out by Google earth

    1. Grease Monkey Silver badge

      What would they gain from doing that? They already have images of those areas and I'm sure they will have shared them with their allies. Why would they care if anybody else saw them or not?

  12. Anonymous Coward
    Anonymous Coward

    UK spy sats

    Really, thought we borrowed time from the yanks, after the Z****** affair?

  13. Generic User

    Best guess: A straight grid would confirm the focus but give you a 1 in 4 chance of having the correct orientation. Some straight lines with the odd intersecting lines create a distinct shape/pattern that only looks "right" if you are calibrated correctly and at the right angle..

  14. Wupspups

    They've been practicing bombing airfields too

    The whole area 14Km ENE (east north east) 40.490N 93.468E is fully of bomb craters and missile grazes. Looks like they had a whole load of whoosh bang blow things up fun there.

    If I was to guess you got a bombing range with areas of ground marked out to simulate airfields and towns.

    1. Grease Monkey Silver badge

      And that would be unusual why? Most militarized countries have bombing ranges and firing ranges and the like.

  15. PrescottS

    QR Code For ET

    It is abviously a QR Code for ET to be able to pull up a wiki on their current interstaller location using an app on their etPhone

    1. /dev/me

      Earth (planet): mostly harmless

      [citation needed]

  16. TheElder

    Calibrations targets??? Utter BUNK. Orient satellite? Hahahahahahahah. Ever heard of star sensors? Paint? If you check it out with Google Earth and drop back to 2005 historical imaging the "paint" is stored in a large pile near some buildings. If you followed the link to the LLM site Mr Hill also gives another "example" of a "satellite calibration target" located in Arizona. It's a Maltese cross near what is now a trap shooting range that appears to have a long disused helicopter runway. Yes, many helicopters use runways for safety reasons. The Maltese cross is a standard FAA symbol for a holding point for helicopters landing. If the runway is occupied they are directed to the cross to wait until the runway is clear.

    It is clear that Mr. Hill either doesn't want to say what it is or just doesn't know and doesn't want to appear ignorant. I strongly suspect the latter.

    I have no idea what it is but I do know what it isn't. It most certainly doesn't have anything to do with "calibrating" satellites. They don't require "calibration". Orientation is done by other means and has nothing to do with something on the ground passing by at 5 miles per second every 32nd orbit.

    1. hayseed

      Star Sensors

      A star pattern would look the same to a satellite tilted at the same angles 100 miles away - after all, the stars are VERY far away. The point of these satellites is to look down, thus the calibration on the ground makes sense.

      1. TheElder

        Not so good at geometry eh? Three sensors looking in X, Y and Z give an absolute attitude fix. The rest is just mathematics. A ground target is useless since it isn't visible most of the time. A spy satellite needs to know where it is pointed ALL of the time.

        1. hayseed

          Ground Target not Useless

          <A ground target is useless since it isn't visible most of the time.>

          math illiterate yourself.

          The star field is a 2-D map. One might thus figure out angles with respect to the ecliptic plane, but the field would look the same translated *very far* in Z perpendicular to the plane, since the stars are so far out.

          A calibration only has to be done when doable, as for most instruments. Error accumulates due to the the limited precision of the calibration, over time, as one calculates positions using dynamics and the result of the last calibration.

        2. hayseed

          Pretty Good at Geometry (Non-Euclidean, Also!)

          In other words (and angles with respect to some other plane than the ecliptic can be used - you are possibly confused by the difference between that and what you get as an end result, that is, that the satellite lies on some *line* through the solar system which can be characterized by the angle it makes through some arbitrary plane. You cannot add information by specifying a different plane - it will be the same line described by different angles, which could have been calculated from math once you know the orientation of one plane with respect to the other. Thus, you get two degrees of freedom, regardless. The third degree has to come from elsewhere.

          1. hayseed

            More Precision

            To be more precise (since I do have training in physics) : a perpendicular unit vector can describe the plane we use for observation. The observed orientation of the craft in the star field observed at that angle can also be described, since it is only a direction, as a unit vector. An observation in any other direction is related to the first by a unitary rotational transformation of the axis. The observed orientation of the spacecraft will be related by the corresponding contravariant transformation. Thus, there is no new information, and certainly nothing about an absolute position - you could translate many millions of miles in any direction and still see the same star field on any but Hubble-sized optics.

            You *only* get orientation information from star observations. This has to be combined with other measurements to actually know where you are at.

  17. ManW/NoName

    More to see?

    There's are a couple of other interesting things in that area also

    A similar grid...,93.742118&hl=en&ll=40.455177,93.390999&spn=0.072494,0.10334&sll=37.0625,-95.677068&sspn=44.25371,67.412109&vpsrc=6&t=h&z=13

    and other bombed buildings?. Hard to miss that shade of blue...,93.510738&spn=0.002264,0.003229&t=h&vpsrc=6&z=18

  18. Zot

    Look to me like...

    Someone's been driving about in random directions harvesting the top soil.

    Either that or it's test burns from satellite fired defense laser! : ) They need to sort out that joystick control though!

  19. Volker Hett

    One looks like a 1:1 representation of a taiwanese air field

    chinese painting

    real air field

  20. Brian Miller

    Satellite lens made by Holga

    Of course the lines are huge! Holga made the lens, so of course it's "low resolution." Probably also has vignetting and light leaks, too.

  21. Sam Therapy

    You're all wrong

    It's China's failed attempts to hide itself with Dazzle.

    You numpties.

  22. Stevie Silver badge


    This is obviously a place where they train people to use the white line painting machine before they let them out on the roads for real.

  23. TheElder

    The lines aren't painted. They are made with salt from the enormous nuclear powered salt refining plant about 250 km to the west on the edge of the Taklamakan dune fields.

    See here 40°25'53.71" N 90°50'05.89" E

  24. Herby

    I am reminded...

    Of a building nearby (where they monitored spy satellites) that was printed on the top of the buildings. It said "If you can read this, you are 10 years behind us". It was written in Cyrillic and was originally a few feet high. It was reduced each year. I have no idea if it is still there after the fall of the Soviet Union.

    My guess is that the Chinese are just getting started.

  25. Godwhacker

    Opportunity to start a war!

    £10 to the first person to do a copy on Hamstead Heath

  26. Christian Berger

    It would make sense for automated systems

    It's fairly easy to detect lines and their cross sections. So if you wanted to build a satellite which automatically 'resets' its idea of its position, you can use such a pattern which can easily be precisely located even if visibility is low.

    A simple grid or regular pattern is far harder to recognize as regular pattern can also occur by chance. That's less likely with such a pattern.

    The great advantage of doing it this way is that there is no "calibration signal" which needs to be sent to the satellite for longer amounts of time. Instead you only need to send it the coordinates where it should take the picture and the coordinates where it should dump the data do. This could be done via many ways, even inconspicuously via RDS on FM-broadcast radio. Nobody would be any wiser where the control centre is.

    You wouldn't even need to have a special monitoring station for that satellite since it will always report it's position. Normal monitoring (which everybody does) would be enough.

  27. mky


    This explains crop circles then?

  28. This post has been deleted by its author

  29. kosh

    It's a map of central <major western industrial city>.

  30. Anonymous Coward
    Anonymous Coward

    It spells out...

    "Made in China, batteries not included"

    1. Lockwood

      Naught to three sad onions.

  31. Martin Huizing

    It's a road marker on cafeine!

  32. JeffyPooh

    Mysterious runes...

  33. Bernard M. Orwell

    Are we certain....

    ....that the pattern doesn't match the layout of any existing cities?

  34. AnonyCow

    So does that mean we can copy it and ruin their calibration?

  35. TheElder

    [quote]math illiterate yourself.

    The star field is a 2-D map. One might thus figure out angles with respect to the ecliptic plane, but the field would look the same translated *very far* in Z perpendicular to the plane, since the stars are so far out.

    A calibration only has to be done when doable, as for most instruments. Error accumulates due to the the limited precision of the calibration, over time, as one calculates positions using dynamics and the result of the last calibration.[/quote]

    I said "attitude fix", not positional fix. The description of attitude is independent of the description of position. It is also independent of the orbit. Further, once launched the satellite is not "calibrated", it is characterized. Calibration is placing a system in a preferred known state. Characterizing is discovering what state a system is in. A system is first calibrated and then characterized. The term "calibration" is one of the most often misused technical terms.

    Calibration of the optical system once on orbit is no longer possible with the sole exception of focus. Once in orbit the satellite optical system must be characterized to discover how it deviates from the pre launch calibration. Also, the orbit must also be characterized since it directly affects the optical system image scale. That cannot be done with a single or a few ground targets.The Earth equipotential gravitational field isn't a globe or even a spheroid. It resembles a lumpy potato. Characterizing the orbit is done with numerous targets world wide which have been characterized by ground truth observations. A very large database of well characterized ground objects is used to constantly update satellite orbital parameters.

    The attitudes fix provided by the star sensors is used together with the imaging data to characterize the pointing accuracy of the optical system. This permits the determination of boresight deviation from from the subzenithal point of the optical system to account for angular compression of apparent dimensions.

    A single target tells you only what is correct for that target, hardly a very useful parameter. Since it isn't practical or even possible to go about painting targets all over the planet real world objects with well know dimensions are used instead.

    As for what the strange hieroglyphs really are, the most like explanation has nothing to do with satellites. They are probably used as optical ground mapping test figures for cruise missile navigation systems.

    1. Grease Monkey Silver badge

      Do you always have to have the last word?

    2. hayseed

      No one has argued that one can't figure out the attitude of a spacecraft from the stars. The trick is to determine the exact position of what one is looking at in earth terms.

      There is no distinction between characterization and calibration here. You characterize things so that your software will read "0" when it is at "0" here. That can be the same as calibration - - for instance, calibration you adjust a readout, maybe, so that the readout reads "0: when it should, when it is found by characterization that a certain voltage should read "0"

      Lots more things are adjusted than the focus - I just read a paper that goes to the many steps that are needed for an accurate geolocation calibration. The idea is to establish known control points for accuracy (and darkness/lighness also) as well as focus so it is known how light/dark a target is, where it is, and to see it sharply.

    3. hayseed

      A Single Target is not useless

      Instead of just using targets you don't control, adding a target that allows you to investigate many parameters at once, under your own control, is not necessarily useless. Of course, they can make use of any number of uncontrolled targets also.

    4. hayseed

      Test Figures

      <<As for what the strange hieroglyphs really are, the most like explanation has nothing to do with satellites. They are probably used as optical ground mapping test figures for cruise missile navigation systems.>>

      Do you recognize that the same logic applies? Why draw test figures when the missile is already passing over known terrain? Could there be additional utility in characterizing all the different transfer functions [there are probably several measurements that they would like to make that I haven't even thought of, all with their own characterizations to be used to calibrate the software readout of measurements. [See, characterization is used *for* calibration - calibration is not misused as a word either by the scientists who wrote the paper I read or myself). Actually, I saw your comments elsewhere.

  36. TheElder

    re: Pretty Good at Geometry (Non-Euclidean, Also!) #

    I fail to see your point. What I said is the star sensors provide a fix on **attitude** as you now seem to agree. Attitude is not position and position is not attitude. However, knowing absolute attitude relative to a "fixed" reference such as the "fixed stars" (astronomy term) is necessary in order to calculate the transformation matrix that describes the pointing of the optical system. That is why star sensors exist.

    So what is your point?

    1. Grease Monkey Silver badge

      Who knows? But you've proved mine.

  37. hayseed

    Gravitational Anomalies

    BTW, data about gravitational anomalies, etc. would have been gathered from many other satellites, especially ones carrying atomic clocks like GPS satellites. Like I said, calculating orbits using differential equations taking into account the anomalies can be done to figure out position elsewhere. The anomalies do NOT have to be rediscovered anew. So orbital mechanics can still tell how the error propagates, and thus two measurements at different times of the same object will still tell you a lot.

  38. This post has been deleted by its author

  39. hayseed

    Maybe checking CCD alignment with image motion?

  40. hayseed

    <<Characterizing the orbit is done with numerous targets world wide which have been characterized by ground truth observations. A very large database of well characterized ground objects is used to constantly update satellite orbital parameters>>

    The paper you and I saw did calibration with 50k strips in the shape of a cross from an aerial survey, dense with control points. They *measured* targets around the world, but the main calibration was done rather locally (this is for a civilian satellite). This was for geolocation. The CCD calibration also seems to require several local points, in the paper. Image distortion and alignment seems to be something that one would want to check out with an entire small region of carefully surveyed points.

    For a military satellite, let's just say that I have heard it said that they have pretty big station-keeping engines. Thus, for this, it may also be necessary to have a wealth of compact information for a quick recalibration.

  41. hayseed

    Paper more closely

    I looked at the last paper more closely in an expanded version, and it was an image registration/alignment paper - it is useful to go over the same area multiple times. As far as alignment, it could probably be done at any of a number ot targets in the world. For a general characterization of the distortion of the imaging system, especially more nonlinear parts, it is probably best (as a picture in another paper of the same guys shows) to have many, many known points in the same image.

  42. hayseed

    Gravitational Anomalies II

    Yes, of course one can do a Green's function description of the gravitational anomalies. 1/r is harmonic except at r=0, so the gravitational field is harmonic outside of the surface. Once one fits a complete orthogonal set of harmonic functions to the description of the gravitational anomaly at the level of the GPS satellites, that set will continue to correctly describe the anomaly at other levels.

This topic is closed for new posts.

Biting the hand that feeds IT © 1998–2021