[OSRS-PROJ] Prime Meridian Support in PROJ.4

Clifford J Mugnier cjmce at lsu.edu
Thu Oct 3 15:16:10 EDT 2002


Here's my two cents worth of opinion on this topic:

The defining parameters of a classical horizontal geodetic datum are:

Astronomic Latitude of the origin,
Astronomic Longitude of the origin,
Ellipsoid Semi-major Axis,
Ellipsoid Semi-minor Axis or Reciprocal of Flattening,
Astronomic Azimuth from origin to another point,
Deflection of the Vertical in the N-S direction,
Deflection of the Vertical in the E-W direction,
Ellipsoid Height of the origin,
Elevation of the origin (or Geoid Height/separation)

The first five of the above are ALWAYS defined.  The remainder are used
with mature datums that generally have been through at least one iteration
within a country.  Many times they are equated to zero, at least on the
first set of datum defining parameters.

Now please notice the second parameter listed above.  It is the Astronomic
Longitude of the origin.  If that value is zero for a country, then by
definition it is that country's Prime Meridian for that country's datum.
Therefore, ALL cartographic coordinate systems that are on that datum use
that point as the Prime Meridian.

However, if a classical datum DOES use a Prime Meridian other than the
Conventional International Origin (CIO) corresponding to Herstmonceaux
Castle at Greenwich, England; a geodetic transformation from that classical
datum to say, WGS84 Datum requires a FOUR-parameter shift.  The fourth
parameter then is an initial rotation about the polar axis to come into
coincidence with Greenwich.

The most common "other" Prime Meridians used are Paris & Ferro (Canary
Islands), but over a dozen of others exist.  As more individuals from all
over the world become users of this "PROJ4" system; the need for such
accommodation will certainly increase.

The primary coordinate system within a country is the DATUM.  Projections
are ALWAYS secondary whenever coordinate precision of a position is better
than 100 meters. In other words, if tens of meters or meters make a
difference in the data set; you have to worry about this stuff.

So there.  Whew!

Cliff

Prof. Clifford J. Mugnier (cjmce at LSU.edu)
Chief of Geodesy
CENTER FOR GEOINFORMATICS
Department of Civil Engineering
LOUISIANA STATE UNIVERSITY
Baton Rouge, LA  70803
Voice and Facsimile:  (225) 578-8536
Pager:              1-(888) 365-5180
================================
http://www.ASPRS.org/resources.html
http://www.ce.LSU.edu/~mugnier/
================================


Frank Warmerdam wrote:
 > In short, I don't think you can operate on coordinate systems with
 > non-greenwich prime meridians properly at this time.
 >
 > The following bugzilla enhancement request addresses this issue:
 >
 >  http://bugzilla.remotesensing.org/show_bug.cgi?id=186 Luis Mota wrote:

Luis Mota wrote:
>  From what I see on this link, you are the person this enhancement has
> been assigned to. Also, I suspect that this enhancement is not a
> priority to you...
>
> The thing is that it is quite important to me, since almost all my data
> is expressed on one of this non-greenwich-prime-meridian-datum. It seems
> to me that this enhancement is not too complicated: it would be a
> question of subtracting the pm value to an input datum and adding it to
> the output, isn't it? Or is it more difficult?
>
> If this is the case, what about if I tried to solve this question? You
> would send me the code and I'd try to contribute with this
> improvement... Is the code written in C? Would you guide me through the
> process and check that the results are correct?

Luis,

I have moved this discussion to the PROJ.4 mailing list, since that is
more appropriate.

PROJ.4 is indeed a C library and you are certainly welcome to prepare a
patch yourself.  If so, please attach the completed and tested patch to
the bug report and I will (after a bit of review) apply it to the core.

If you want to do this, please prepare the patch against the current CVS
version of PROJ.4.  Various information about PROJ.4, including anonymous
CVS access to the source is available at:

   http://www.remotesensing.org/proj

One concern I have is whether the prime meridian should be considered to
have any relationship to longitude parameters for a projection.  For
instance, if a TM projection uses a geographic coordinate system with a
Lisbon prime meridian, is the central meridian supposed to be specified as
degrees from Lisbon or Greenwich?  What is common practice?

For practical purposes in PROJ.4 I would have preferred that prime meridian
issues remain completely divorced from the projection routines (pj_fwd,
pj_inv) and only addressed at the point where transations are done from one
lat/long coordinate system to another.  But this sort of implies that all
longitude parameters to a projection with an alternate prime meridian are
implicitly relative to that pm ... that may be a peculiar convetion.

I would like the feedback of the PROJ.4 community on this issue before we
commit to an approach to prime meridian support.

Best regards,
--
---------------------------------------+--------------------------------------
I set the clouds in motion - turn up   | Frank Warmerdam,
warmerdam at pobox.com
light and sound - activate the windows | http://pobox.com/~warmerdam
and watch the world go round - Rush    | Geospatial Programmer for Rent


----------------------------------------
PROJ.4 Discussion List
See http://www.remotesensing.org/proj for subscription, unsubscription
and other information.


----------------------------------------
PROJ.4 Discussion List
See http://www.remotesensing.org/proj for subscription, unsubscription
and other information.



More information about the Proj mailing list