Loading ...
Sorry, an error occurred while loading the content.

Re: [OziUsers-L] Re: Shatt-al-GPS

Expand Messages
  • Dennis Pogson
    Yeah! Believing anything you read in the Guardian is tantamount to believing the earth is flat. Add the fact that Landsat 7 images are about as reliable as a
    Message 1 of 20 , Apr 1, 2007
    • 0 Attachment
      Yeah! Believing anything you read in the Guardian is tantamount to believing the earth is flat. Add the fact that Landsat 7 images are about as reliable as a child's scribbling pad for navigation purposes, and this whole excercise becomes derisory.

      Agreed, the positions quoted are relative and therefore would show differences on even an inaccurate map. I would find it difficult to understand, if there was more than one handheld GPS on the British boats, why someone did not hit the "mark" key or the "man overboard" keys at the time of the attack, thus putting their position beyond all doubt. Maybe the Navy should have a special course aimed at teaching it's recruits how to use a GPS. It would sure relieve the boredom of playing Monopoly.

      Sounds like Garmin could exploit this one to their advantage, whichever way you look at it!

      Dennis.

      ----- Original Message -----
      From: David McKinnon
      To: OziUsers-L@yahoogroups.com
      Sent: Saturday, March 31, 2007 10:57 PM
      Subject: RE: [OziUsers-L] Re: Shatt-al-GPS



      Oh yeah, don't trust what you read in the newspapers, no matter how well
      regarded they are.
      Just a typo, but has a significant effect on matters



      _____

      From: OziUsers-L@yahoogroups.com [mailto:OziUsers-L@yahoogroups.com] On Behalf
      Of egdlym
      Sent: Sunday, 1 April 2007 6:11 AM
      To: OziUsers-L@yahoogroups.com
      Subject: [OziUsers-L] Re: Shatt-al-GPS

      Thanks to J.G. and Davo on the datum issue, which is pretty important
      stuff for ozie users.

      First, the UK newspaper "The Guardian" which I quoted from in
      starting this thread, has re-published the coords using the same
      numbers but now changing it to deg/min/sec whereas their earlier
      effort was deg/min.mm, thus 29 50.36' N 048 43.08'(they did use the
      minute apostrophe)is now replaced by 29 50' 36"N 048 43' 08" E ,
      Adding this to the Landsat/MrSid image I was using, after editing
      config/maps/lat-lon, gives a third position !! It is 451 metres North
      of the ddd mm.mm version. No wonder bombs go astray.

      Second, The Guardian showed a satpic at low tide and the navigable
      channel is clearly seen with a meander which is not followed by their
      overprint of the Iraq/Iran offshore boundary. Therefore any ship
      would have to curve into Iraqi waters on its way up the channel if
      this is correct. If on the other hand the boundary goes mid-channel
      round the bend, the ship's position would be much closer to the line.

      If anyone says they want it, I could post some images on the file
      section of this usegroup.

      regrda,

      Richard

      --- In OziUsers-L@yahoogro <mailto:OziUsers-L%40yahoogroups.com> ups.com,
      "rwcx183" <lgalvin@...> wrote:
      >
      > --- In OziUsers-L@yahoogro <mailto:OziUsers-L%40yahoogroups.com> ups.com,
      "egdlym" <rpgosnell@> wrote:
      > >
      > > Quite right Davo, we all use greenwich/equator for the origin.
      > > However, I understand that choosing different mathematical models
      > > (called spheroids)for the Earth's MSL shape causes lat/long, and
      > > grid, lines to appear to be 200 or more metres shifted among
      > > different models/sheroids. Obviously, using the true, exact
      shape
      > of
      > > the Earth would remove all ambiguity, but I gather that is too
      much
      > > for a whole-earth model. WGS84 is therefore a simple whole-earth
      > > compromise that has significant errors compared to the 'Truth'
      > > locally. Most countries have a local spheroid (UK uses the Airy
      > > Spheroid) which gives a much better match than WGS 84, but only
      > > within its assigned national coverage. If the Iranians' nautical
      > or
      > > topo maps were using a local accurate datum there would be a
      > similar
      > > discrepancy compared to WGS 84
      > >
      > > The Degree Covergence Project had to decide, for expediency, to
      use
      > > WGS 84 to keep everone sane. But going to say 52 00' 00"N, 002
      00'
      > > 00"W using OSGB 36 would put you 200 metres away from DCP
      > position,
      > > evn though the OSGB is in fact more accurate.
      > >
      > > Richard
      >
      > Richard,
      >
      > Actually, once the location of the center of the earth is
      > established, the ellipsoid can be changed at will, with no effect
      on
      > lat/lon coordinates. Lat/lon are just angles. The model of the
      > earth's shape cannot change these angles. Different ellipsoids
      were
      > indeed used with different datum definitions, but mostly this was
      > done to make altitudes accurate for a given local area. It's also
      > absolutely necessary when using non lat/lon grid coordinates like
      > UTM. To transform lat/lon coordinates from one datum to another,
      > mostly requires a simple x/y/z offset correction for the origin
      > (center of the earth) and it's not necessary to consider a
      different
      > ellipsoid. This is the so-called 3 parameter transform used by
      > OziExplorer. Ozi of course does take into account differing
      > ellipsoids for the purpose of displaying positions in non lat/lon
      > grid coordinates where ellipsoid really does matter.
      >
      > To anyone that thinks they understand all this, it will come as no
      > shock that the ground offset between 2 datums, is different for
      > lat/lon coordinates than it is for UTM or other grid coordinates.
      >
      > J.G.
      >

      [Non-text portions of this message have been removed]






      ------------------------------------------------------------------------------


      No virus found in this incoming message.
      Checked by AVG Free Edition.
      Version: 7.5.446 / Virus Database: 268.18.24/741 - Release Date: 31/03/2007 20:54



      [Non-text portions of this message have been removed]
    • rwcx183
      ... deg long, 0 ... displaying ... use grid ... Implicit in the definition of lat/lon coordinate system, are 2 reference points. One at the center of the
      Message 2 of 20 , Apr 1, 2007
      • 0 Attachment
        --- In OziUsers-L@yahoogroups.com, "David McKinnon"
        <david.mckinnon@...> wrote:
        >
        > So please explain for us simple map users what is the origin of 0
        deg long, 0
        > deg latitude.
        > If I visit a degree confluence point with my GPS set to wgs84 and
        displaying
        > lat/long is this considered a valid visit?
        >
        > Ps most of my work is to do with engineering type projects where we
        use grid
        > references to set out structures, roads, pipelines etc.
        >

        Implicit in the definition of lat/lon coordinate system, are 2
        reference points. One at the center of the earth and the other at
        0deg lon, 0deg latitude. The line connecting these two points, is
        the reference line from which all lat/lon angles will be measured.
        The issue is primarily with the point at the center of the earth.
        This exact location of that, with respect to surface of the earth,
        turns out to be quite a difficult thing to determine. As I mentioned
        in previous posts, prior to the advent of satellite based measuring
        systems, this was done mostly by astronomical observations. There
        are many sources of errors in these measurements. If the location of
        the center of the earth is determined and it's not quite accurate,
        then lat/lon coordinates based on the reference line, will be
        similarly inaccurate. There are also issues with the 0deg lat, 0deg
        lon line, which was also established by astronomical observations.
        Once you move significantly away from 0deg lat, 0deg lon reference,
        accuracy starts to degrade. Local datums were established as local
        reference points on which surveyors could base local land surveys.
        Each local datum had an implied reference point at the center of the
        earth. It was of course intended that this point was the same for
        every local datum around the world, but there was no way to prove
        that. Only with the advent of more modern measuring methods, such as
        satellite based systems and VLBI, did it become clear, that all of
        these local datums were not based on the same implied reference point
        at the center of the earth. Once the GPS system was operational, it
        became possible to find that mythical perfect reference point at the
        center of the earth. This point is actuall assigned an x/y/z
        coordinate of 0,0,0 and all other points on the earth can be
        represented as x/y/z offsets from that point. Of course, to
        establish a 3D x/y/z coordinate system, we need to establish 2
        reference lines emanating from the origin. The x axis is the line
        from the origin through 0deg lat, 0deg lon point. The y axis is the
        line from the origin through 0deg lat, 90 lon and the z axis is
        simply the polar axis. With all that established, for any given
        point on the earth, the x/y/z offset can be measured by GPS and in
        fact this is exactly what your GPS receiver does. Knowing your x/y/z
        offset, it's a small matter of math, to calculate latitude and
        longitude as angles. Once people started doing this, it was
        immediately noticed, that there were problems with the now accurately
        known lat/lon coordinates Vs. the traditionally measured lat/lon
        coordinates. As it turns out, it's not so difficult to figure out
        that for a given local datum, you can translate coordinates to the
        new reference system, by assuming that there's a fixed offset between
        the implied center of the earth reference point in the local datum
        and the new more accurately known center of the earth. Once you have
        that, it's some more simple math to establish a fixed x/y ground
        offset between the old local datum and the new ultra accurate WGS84
        datum. For local datums that apply to fairly small land masses, this
        works pretty well, but for a continent sized land mass, there are
        still problems and in such cases, high precision grid networks are
        developed to determine the offset between the older local datum and
        WGS84 at any given point. Here in the U.S., the offset between the
        old NAD27 datum and WGS84 varies some +/- 10 meters or so on the
        ground. Still, most GPS receivers can only directly calculate
        coordinates with respect to WGS84 datum and then they use a simple
        fixed x/y/z offset to translate coordinates to old datums. For large
        continents like North America, an average (mean) offset is used and
        this can be off by as much as 10 meters.

        It may come as a surprise to some, that even the gold standard WGS84
        datum has been "adjusted" over the years. Twice to be exact, as GPS
        equipment improved with each generation. These adjustments were much
        less than 1 meter. So, not much to worry about for the average GPS
        user.

        J.G.
      • rwcx183
        ... deg long, 0 ... displaying ... use grid ... After re-reading your question, I think that what you re asking is; how can we know where the north/south poles
        Message 3 of 20 , Apr 1, 2007
        • 0 Attachment
          --- In OziUsers-L@yahoogroups.com, "David McKinnon"
          <david.mckinnon@...> wrote:
          >
          > So please explain for us simple map users what is the origin of 0
          deg long, 0
          > deg latitude.
          > If I visit a degree confluence point with my GPS set to wgs84 and
          displaying
          > lat/long is this considered a valid visit?
          >
          > Ps most of my work is to do with engineering type projects where we
          use grid
          > references to set out structures, roads, pipelines etc.

          After re-reading your question, I think that what you're asking is;
          how can we know where the north/south poles are? Well, if you look
          straight up in the night sky and the stars revolve around that point
          straight up in the sky, then you know that you're at the north or
          south pole. The equator (zero deg lat) is of course, half way
          between the poles. Establishing a lon reference line was a lot
          easier, as it was done arbitrarily as the line connecting the poles
          that passes through the royal observatory in Greenwich England. The
          intersection of the equator and the Greenwich meridian, is the point
          of 0deg lat, 0deg lon. It's somewhere off the coast of Africa and
          you will need a boat or airplane, barring that you're a really good
          swimmer. To visit there with your GPS set to WGS84 is of course
          valid with respect to the degree confluence project, since that's the
          datum they've chosen. I expect that one has been visited quite a lot.

          J.G.
        • Matti Gronroos
          ... No. The the coordinates are based on a reference ellipsoid having certain characteristics: position on the centre and the lenght of the axis. The
          Message 4 of 20 , Apr 1, 2007
          • 0 Attachment
            --- In OziUsers-L@yahoogroups.com, "rwcx183" <lgalvin@...> wrote:

            > Implicit in the definition of lat/lon coordinate system, are 2
            > reference points. One at the center of the earth and the other at
            > 0deg lon, 0deg latitude.

            No. The the coordinates are based on a reference ellipsoid having
            certain characteristics: position on the centre and the lenght of the
            axis. The coordinates are then coordinates of this ellipsoid.

            The position of these ellipsoids vary. This is because even the term
            "the center of the earth" is ambiguous. The WGS-84 ellipsoid is
            positioned at the mass center of the earth.

            The earth's shape is approximated by an ellipsoid. Somewhere, the
            ellipsoid A is the most accurate approximation. Somewhere else, the
            ellipsoid B is a better choice.

            > Only with the advent of more modern measuring methods, such as
            > satellite based systems and VLBI, did it become clear, that all of
            > these local datums were not based on the same implied reference point
            > at the center of the earth.

            No!

            The mathematics of reference ellipsoids and datums dates back to
            1800's. The land surveyors knew very well what they were doing: They
            were creating local datums by best fitting the size, shape and
            location of the reference ellipsoid to the local region.

            The reason for this is obvious: The local measurements had to be as
            accurate as possible. The differences to the neigbouring countries is
            not a big issue, because the conversion from datum A to datum B can be
            done with a moderate accuracy.

            The "incompatibility" of the local datums was not an surprise to any
            professional on the area of geodesy.

            > Once the GPS system was operational, it
            > became possible to find that mythical perfect reference point at the
            > center of the earth.

            No. That work was, of course, done prior to the launch of GPS. Before
            WGS84, there were WGS60, WGS66, and WGS72, all based on gravitational
            measurements, and (with an exception of WGS60) on astrogeodetic data.

            The idea of WGS84 is to implement a global coordinate system, but NOT
            to replace local ones. The GPS is far from being accurate enough for
            land surveyours needing an accuracy to millimeters.

            The land surveyors are using terrestial reference frames for accurate
            measurements. They implement a fourth coordinate: the time. That is
            because the shape of the world is constantly changing. For example,
            the European ETRF reference network today almost equals to WGS84 (with
            the difference of centimeters), but this is not necessarily the case
            in the future: Thw WGS84 is tied to the mass center of the earth, and
            ETRF to the position of the Eur-Asian continental plate.

            Matti
          • egdlym
            Hi all. The best way to get a grip on the datums issue is to suck it and see.....with oziexplorer, or with your gps receiver s menu settings. Try displaying a
            Message 5 of 20 , Apr 2, 2007
            • 0 Attachment
              Hi all.

              The best way to get a grip on the datums issue is to suck it and
              see.....with oziexplorer, or with your gps receiver's menu settings.
              Try displaying a calibrated map in ozie, ideally of some place you
              know well, and then alter the datum settings to all sorts, and using
              map/display grid/display alt grid, see where the grid lines jump
              around to. Also, hovering the cursor over a waypoint displays its
              coords, but these will be different to the waypoint list version if
              the display datum, map datum and data file datum differ. Only if
              everything is set to WGS84 will values agree, but then you know the
              gridlines are 200 metres away from where they are by the more-
              accurate local datum. Time to tear out ALL hair!

              If you use a scanned image of a UK 1:50,000 map, there are little
              blue crosses over the map at 5-minute lat/lon intevals, so you can
              see whether ozie-generated lat/lon coincides. If you used ORD SRVY
              datum in calibration they will fit pretty well. Changing datum will
              show an offset. This experiment could be done with any other
              countries' maps that have lat/lon graticule or just the maps' corner
              lat/lon.

              Then repeat using metric grid squares (alt grid) and note how they
              jump to slightly different places relative to the lat/lon!! Tear out
              last strand of hair or start on another part!! Basically grid
              northing and eastings are conceived as though by tape measure across
              the (MSL)surface, (after the map projection was decided and
              executed), so the tape measure journey is longer or shorter by a
              small amount depending on the earth's shape assumed in the datum.
              So, measuring horizontally using a fixed length unit (the metre)
              progresses at a different pace over the (MSL) surface to using what
              is trying to be an angle from the earth's centre.


              Richard


              --- In OziUsers-L@yahoogroups.com, "rwcx183" <lgalvin@...> wrote:
              >
              > --- In OziUsers-L@yahoogroups.com, "David McKinnon"
              > <david.mckinnon@> wrote:
              > >
              > > So please explain for us simple map users what is the origin of
              0
              > deg long, 0
              > > deg latitude.
              > > If I visit a degree confluence point with my GPS set to wgs84 and
              > displaying
              > > lat/long is this considered a valid visit?
              > >
              > > Ps most of my work is to do with engineering type projects where
              we
              > use grid
              > > references to set out structures, roads, pipelines etc.
              >
              > After re-reading your question, I think that what you're asking is;
              > how can we know where the north/south poles are? Well, if you look
              > straight up in the night sky and the stars revolve around that
              point
              > straight up in the sky, then you know that you're at the north or
              > south pole. The equator (zero deg lat) is of course, half way
              > between the poles. Establishing a lon reference line was a lot
              > easier, as it was done arbitrarily as the line connecting the poles
              > that passes through the royal observatory in Greenwich England.
              The
              > intersection of the equator and the Greenwich meridian, is the
              point
              > of 0deg lat, 0deg lon. It's somewhere off the coast of Africa and
              > you will need a boat or airplane, barring that you're a really good
              > swimmer. To visit there with your GPS set to WGS84 is of course
              > valid with respect to the degree confluence project, since that's
              the
              > datum they've chosen. I expect that one has been visited quite a
              lot.
              >
              > J.G.
              >
            • rwcx183
              I think that you re missing the point. Latitude and longitude are angles. An angle is formed by 2 intersecting lines. A line is defined by 2 points. It
              Message 6 of 20 , Apr 2, 2007
              • 0 Attachment
                I think that you're missing the point. Latitude and longitude are
                angles. An angle is formed by 2 intersecting lines. A line is
                defined by 2 points. It does not matter if the earth was square
                shaped. The angles remain the same. You cannot make a measurement
                of an angle of a line without an intersection to another line. This
                intersection must be at the implied center of the earth. Simple
                basic geometry.

                --- In OziUsers-L@yahoogroups.com, "Matti Gronroos"
                <matti.gronroos@...> wrote:
                >
                > --- In OziUsers-L@yahoogroups.com, "rwcx183" <lgalvin@> wrote:
                >
                > > Implicit in the definition of lat/lon coordinate system, are 2
                > > reference points. One at the center of the earth and the other
                at
                > > 0deg lon, 0deg latitude.
                >
                > No. The the coordinates are based on a reference ellipsoid having
                > certain characteristics: position on the centre and the lenght of
                the
                > axis. The coordinates are then coordinates of this ellipsoid.
                >

                They would do just as well if the shape was a dodecahedron. Lat/lon
                angles have nothing to do with the an arbirtrary model of the earth's
                shape. They have only to do with a reference line.


                > The position of these ellipsoids vary. This is because even the term
                > "the center of the earth" is ambiguous. The WGS-84 ellipsoid is
                > positioned at the mass center of the earth.
                >

                The position of these ellipsoids vary because they could not
                accurately determine the mass center of the earth. They very much
                did want to do that.

                > The earth's shape is approximated by an ellipsoid. Somewhere, the
                > ellipsoid A is the most accurate approximation. Somewhere else, the
                > ellipsoid B is a better choice.
                >
                > > Only with the advent of more modern measuring methods, such as
                > > satellite based systems and VLBI, did it become clear, that all
                of
                > > these local datums were not based on the same implied reference
                point
                > > at the center of the earth.
                >
                > No!
                >
                > The mathematics of reference ellipsoids and datums dates back to
                > 1800's. The land surveyors knew very well what they were doing: They
                > were creating local datums by best fitting the size, shape and
                > location of the reference ellipsoid to the local region.
                >
                > The reason for this is obvious: The local measurements had to be as
                > accurate as possible. The differences to the neigbouring countries
                is
                > not a big issue, because the conversion from datum A to datum B can
                be
                > done with a moderate accuracy.
                >
                > The "incompatibility" of the local datums was not an surprise to any
                > professional on the area of geodesy.
                >
                > > Once the GPS system was operational, it
                > > became possible to find that mythical perfect reference point at
                the
                > > center of the earth.
                >
                > No. That work was, of course, done prior to the launch of GPS.
                Before
                > WGS84, there were WGS60, WGS66, and WGS72, all based on
                gravitational
                > measurements, and (with an exception of WGS60) on astrogeodetic
                data.
                >

                It was done with quite poor accuracy prior to the advent of GPS,
                which is my point.

                > The idea of WGS84 is to implement a global coordinate system, but
                NOT
                > to replace local ones. The GPS is far from being accurate enough for
                > land surveyours needing an accuracy to millimeters.
                >
                > The land surveyors are using terrestial reference frames for
                accurate
                > measurements. They implement a fourth coordinate: the time. That is
                > because the shape of the world is constantly changing.

                It isn't so much the shape of the earth that is changing, that
                requires the 4th dimension, as it is tectonic plate motion, a fact of
                which I deal with on a regular basis.

                For example,
                > the European ETRF reference network today almost equals to WGS84
                (with
                > the difference of centimeters), but this is not necessarily the case
                > in the future: Thw WGS84 is tied to the mass center of the earth,
                and
                > ETRF to the position of the Eur-Asian continental plate.

                Tectonic plate motion can happen with very negligible change to the
                shape of the earth. 2 different things.

                >
                > Matti
                >

                The shape of the earth has nothing to do with measuring lat/lon.
                Grid coordinate systems like UTM or other local grids, of course are
                profoundly effected by the shape of the earth.

                Don't get hung up on land surveying. The idea of lat/lon doesn't
                work well at all for that purpose. It was conceived for navigational
                purposes and the goal very much was to establish a global system to
                be used everywhere. Land surveyors did attempt to tie into that
                system through the establishment of local datums and it was intended
                that these local datums would fit well into the global lat/lon
                system. Unfortunately, they had to make adjustments, since they
                could not even figure out a true straight up and down line (due to
                gravity effects) and this distorted their astro measurements. It
                didn't really matter at the time, since continent to continent
                accuracies were plenty good enough for navigational purposes.

                Land surveyors today cling to the old local datums for historical
                reasons. If there were none and they were to be devised today, there
                would be just one global datum. Of course this isn't good enough for
                millimeter accuracy. Nobody is claiming that it is. Continue on
                with differential GPS, but don't forget that you would not have that
                at all, if it weren't for WGS84.

                J.G.
              • Matti Gronroos
                ... No. The latitude of a point P on the Earth is defined as a geodetic latitude, which is the angle between the equatorial plane and the normal to the
                Message 7 of 20 , Apr 2, 2007
                • 0 Attachment
                  --- In OziUsers-L@yahoogroups.com, "rwcx183" <lgalvin@...> wrote:
                  >
                  > I think that you're missing the point. Latitude and longitude are
                  > angles. An angle is formed by 2 intersecting lines. A line is
                  > defined by 2 points. It does not matter if the earth was square
                  > shaped. The angles remain the same. You cannot make a measurement
                  > of an angle of a line without an intersection to another line. This
                  > intersection must be at the implied center of the earth. Simple
                  > basic geometry.

                  No.

                  The latitude of a point P on the Earth is defined as a geodetic
                  latitude, which is the angle between the equatorial plane and the
                  normal to the reference ellipsoid surface on that position. That angle
                  is NOT the same as the angle between the equatorial plane and the line
                  drawn from the center of the mass to the point P. The difference
                  depends on the latitude and it varies from zero to about 0.2 degrees,
                  i.e. to about 20 kilometres.

                  Unfortunately, is not just simple basic geometry, because are not
                  talking about spherical coordinates. Thus, forget the geocentric
                  angles in this context.

                  > They would do just as well if the shape was a dodecahedron. Lat/lon
                  > angles have nothing to do with the an arbirtrary model of the earth's
                  > shape.

                  Yes, they do. If you need to find a position of a point P in a 3D
                  coordinate system relative to a 3D shape, you need to know the metrics
                  of that shape.

                  In the GPS positioning (and other positioning), that 3D shape is the
                  reference ellipsoid, which is a approximation of the irregular shape
                  of the Earth. The position lat/lon given by a GPS device consist of
                  the following coordinates:

                  1) geodetic latitude (which is not equal to the geocentric angle)
                  2) geodetic longitude (which is equal to the geocenteric angle)
                  3) vertical distance from the sufrace of the reference ellipsoid

                  To calculate 1) and 3), you need to know the position, orientation and
                  the size of the ellipsoid.

                  > The position of these ellipsoids vary because they could not
                  > accurately determine the mass center of the earth. They very much
                  > did want to do that.

                  This definitely is not true. They wanted to find an ellipsoid best
                  matching to the area they were mapping.

                  > Tectonic plate motion can happen with very negligible change to the
                  > shape of the earth. 2 different things.

                  No. The movement currently is about 2.5 centimeters a year. That is
                  why WGS84 is complemented by terrestial reference systems implementing
                  a non-moving coordinate system.

                  Matti
                • rwcx183
                  ... are ... measurement ... This ... angle ... line ... degrees, ... Yes, you re right. I don t know what I was thinking. I did know this at one time, but
                  Message 8 of 20 , Apr 3, 2007
                  • 0 Attachment
                    --- In OziUsers-L@yahoogroups.com, "Matti Gronroos"
                    <matti.gronroos@...> wrote:
                    >
                    > --- In OziUsers-L@yahoogroups.com, "rwcx183" <lgalvin@> wrote:
                    > >
                    > > I think that you're missing the point. Latitude and longitude
                    are
                    > > angles. An angle is formed by 2 intersecting lines. A line is
                    > > defined by 2 points. It does not matter if the earth was square
                    > > shaped. The angles remain the same. You cannot make a
                    measurement
                    > > of an angle of a line without an intersection to another line.
                    This
                    > > intersection must be at the implied center of the earth. Simple
                    > > basic geometry.
                    >
                    > No.
                    >
                    > The latitude of a point P on the Earth is defined as a geodetic
                    > latitude, which is the angle between the equatorial plane and the
                    > normal to the reference ellipsoid surface on that position. That
                    angle
                    > is NOT the same as the angle between the equatorial plane and the
                    line
                    > drawn from the center of the mass to the point P. The difference
                    > depends on the latitude and it varies from zero to about 0.2
                    degrees,
                    > i.e. to about 20 kilometres.
                    >
                    > Unfortunately, is not just simple basic geometry, because are not
                    > talking about spherical coordinates. Thus, forget the geocentric
                    > angles in this context.
                    >

                    Yes, you're right. I don't know what I was thinking. I did know
                    this at one time, but seem to have lost this knowledge over the years.

                    As a practical reality, for a given point on the earth, I can define
                    2 lat/lon coordinate systems, using different ellipsoids, but both
                    with the same origin. Both ellipsoids provide a reasonable fit to
                    the shape of the earth in a given local area. Converting lat/lon
                    coordinates from one to the other, results in zero change in
                    longitude coordinates and an extremely tiny change in latitude
                    coordinates. I'll grant you that the extremely tiny change may be
                    very important for land surveying at the millimeter level, but it's
                    less than negligible for navigation purposes. The point I'm trying
                    to make here, is that a far larger source of error between datums, is
                    the location of the origin and not the difference in ellipsoids for 2
                    datums.


                    > > They would do just as well if the shape was a dodecahedron.
                    Lat/lon
                    > > angles have nothing to do with the an arbirtrary model of the
                    earth's
                    > > shape.
                    >
                    > Yes, they do. If you need to find a position of a point P in a 3D
                    > coordinate system relative to a 3D shape, you need to know the
                    metrics
                    > of that shape.
                    >
                    > In the GPS positioning (and other positioning), that 3D shape is the
                    > reference ellipsoid, which is a approximation of the irregular shape
                    > of the Earth. The position lat/lon given by a GPS device consist of
                    > the following coordinates:
                    >
                    > 1) geodetic latitude (which is not equal to the geocentric angle)
                    > 2) geodetic longitude (which is equal to the geocenteric angle)
                    > 3) vertical distance from the sufrace of the reference ellipsoid
                    >
                    > To calculate 1) and 3), you need to know the position, orientation
                    and
                    > the size of the ellipsoid.
                    >

                    Not to nitpick here, but the prime result of a standalone GPS
                    position calculation, is a position in 3D ECEF XYZ coordinates. Any
                    translation to other coordinate systems, is not really anything to do
                    with the GPS system per se, but more to do with putting the
                    coordinates into a more human friendly form.


                    > > The position of these ellipsoids vary because they could not
                    > > accurately determine the mass center of the earth. They very
                    much
                    > > did want to do that.
                    >
                    > This definitely is not true. They wanted to find an ellipsoid best
                    > matching to the area they were mapping.
                    >
                    > > Tectonic plate motion can happen with very negligible change to
                    the
                    > > shape of the earth. 2 different things.
                    >
                    > No. The movement currently is about 2.5 centimeters a year. That is
                    > why WGS84 is complemented by terrestial reference systems
                    implementing
                    > a non-moving coordinate system.

                    I'm not following your thinking here. The plates can move around and
                    that may change the x/y shape of the land mass, but the shape of the
                    earth's surface doesn't change much because of it. True there are
                    very minor changes in the geoid, but these are much smaller than the
                    simple x/y ground position changes.

                    J.G.
                  Your message has been successfully submitted and would be delivered to recipients shortly.