[Proj] Accuracy of projections

[Clifford J Mugnier]

The mathematical accuracy of cartographic transformations used by PROJ4
such as the Gauss-Krueger Transverse Mercator are capable of yielding
excellent results (±0.1mm) when within ±3½º of the central meridian.
However, when ALSO introducing a datum shift into the transformation
process, the accuracy rapidly plummets to many ± meters!

The reason for this is that classical geodetic datums that have been
established by astronomical observations at origin points and then extended
through geodetic triangulation chains of quadrilaterals are often not
accurate within themselves to better than a meter!  No datum shift can
yield accuracies better than the basic accuracy of the datum itself.  All
datum shifts DEGRADE the accuracy of all points transformed because the
generation of transformation parameters is a "best fit" of all the
positions determined on a particular datum with relation to another datum,
commonly WGS84 or ITRFXX.

Datum shifts that use 3 parameters (common with EPSG info) will commonly
yield accuracies of better than ±25 meters.  How much better?  Depends on
the size of the area and the specifications of the original 19th/20th
century surveys.  Increasing the number of parameters may increase accuracy
over a small area. How much again depends on original accuracies of the
classical datum.

In regard to your question, better than one meter? Likely not.

Clifford J. Mugnier C.P.,C.M.S.
National Director (2006-2008)
Photogrammetric Applications Division
AMERICAN SOCIETY FOR PHOTOGRAMMETRY AND REMOTE SENSING
and
Chief of Geodesy
Center for GeoInformatics
LOUISIANA STATE UNIVERSITY
Dept. of Civil Engineering
CEBA 3223A
Baton Rouge, LA  70810
Voice:              (225) 578-8536
Facsimile:          (225) 578-8652
-----------------------------------------------------------
http://www.ASPRS.org/resources/GRIDS
http://www.cee.lsu.edu/facultyStaff/mugnier/index.html
-----------------------------------------------------------



Hello,

I'd like to transform GPS (DGPS) generated coordinates from WGS84
(lat/lon) to Gauss-Krueger coordinates on Bessel ellipsoid(epsg 31463). Is
it possible to receive an accuracy <1 meter using proj.4 standard epsg
file? (I don't really want to generate my own helmerttransformation and
transformationparameters...).

Kind regards
Tobias