[Proj] Datum shift for Dutch maps 1850-2000

[Jan Hartmann]

This certainly makes a difference! Adding +nadgrids=@null to the 
original definition gives a map with a 120m South shift, still too much 
but nearing a solution.  Isn't this a bug in PROJ? I did not request a 
datum shift, and neither the from-projection, nor the to-projection had 
a datum defined, only ellipsoids. I am using Proj 4.6.1.

I looked again at the 19th century documentation, and it says the map is 
in the "modified Flamsteed or French" projection. All later 
documentation refers to "Bonne" as the projection for this map and its 
successors; they are actually still called "Bonne maps" here. The 
Flamsteed projection is the classic sinusoidal projection, which is 
certainly not the one for this map, and the "modified Flamsteed 
projection" seems to be some sort of conic projection, but I haven't 
found any information on this. It could be the same as the Bonne 
projection, which is conic too. Does anyone have any information on this?

Jan

Frank Warmerdam wrote:
 > Jan Hartmann wrote:
 >> Well, changing the ellipsoid seems to help. The ellipsoid that is 
officially documented for this map, and the one I have been using has 
+a=6376950.4 +rf=309.65.
 >> This gives the 650m Nord shift. Replacing this ellipsoid by the 
Bessel ellipsoid (a=6377397.155, rf=299.1528128) gives a 50m South 
shift. Again, I am reprojecting to the present projection which has a 
Bessel ellipsoid. I can see that the first ellipsoid has a significantly 
smaller semi-major axis than the second one. Does this explain the  Nord 
shift?
 >
 > Jan,
 >
 > A large north-south shift would be consistent with proj.4 attempting an
 > automatic datum change via the ellipsoid (converting from latlong on one
 > ellipsoid to geocentric, and then to lat/long on the other ellipsoid).
 >
 > Best regards,