[Proj] Convergence

Mikael Rittri Mikael.Rittri at carmenta.com
Wed Sep 17 07:41:51 EDT 2008

I think you are right that PROJ.4 is inconsistent here. 

Before I read your post, I thought there was a standard sign convention
for meridian convergence, although I could never remember which way it went.  
But I have done some googling now, and the situation seems as bad as for 
the rotation angles in a 7-parameter transform.  

If I use a slash and a backslash to describe how a piece of a meridian 
looks in a projected grid, then: 

The National Land Survey of Sweden agrees with +proj=utm, and says:
  \ has positive meridian convergence, 
  / has negative.  

But the Ordnance Survey of Great Britain agrees with +proj=lcc, and says:
  \ has negative meridian convergence,
  / has positive.

I didn't search further.  Possibly, there might be a clear majority 
viewpoint among all national mapping agencies, but I got tired. 

Best regards, 

Mikael Rittri
Carmenta AB
Box 11354
SE-404 28 Göteborg
Visitors: Sankt Eriksgatan 5
Tel: +46-31-775 57 37
Mob: +46-703-60 34 07 
mikael.rittri at carmenta.com

-----Original Message-----
From: proj-bounces at lists.maptools.org [mailto:proj-bounces at lists.maptools.org] On Behalf Of Kees Krikke
Sent: den 15 september 2008 17:28
To: PROJ.4 and general Projections Discussions
Subject: [Proj] Convergence


I've made a piece of s/w that uses proj's convergence calculation to find the true north for a given grid north.

When testing with a utm and lcc projection i found some confusing results regarding the sign (direction) of the convergence.

To demonstrate this i modified proj.exe so it prints the convergence (in
deg) at the end of the line when using the -S flag:

C:\proj-4.6.1\src>proj -vS +proj=utm +datum=WGS84 +lon_0=3
#Universal Transverse Mercator (UTM)
#       Cyl, Sph
#       zone= south
# +proj=utm +datum=WGS84 +lon_0=3 +ellps=WGS84 +towgs84=0,0,0
2 52
431350.30       5761510.32      <0.999658 0.999658 0.999316 1.20783e-006 
0.999658 0.999658 -0.788041>
4 52
568649.70       5761510.32      <0.999658 0.999658 0.999316 8.54066e-007 
0.999658 0.999658 0.788041>
4 52.1
568496.62       5772632.28      <0.999658 0.999658 0.999315 0 0.999658 
0.999658 0.789115>

 From the above results it seems that for utm the convergence is 
negative when west and positive when east of lon_0.
If you calculate two utm positions along the same longitude east of 
lon_0, you see that the easting reduces when going northwards.

 From this is seems that TrueNorth = GridNorth + convergence

I used the same points for a lcc projection:

C:\proj-4.6.1\src>proj -vS +proj=lcc +datum=WGS84 +lat_0=52 +lon_0=3 
+lat_1=50 +lat_2=54
#Lambert Conformal Conic
#       Conic, Sph&Ell
#       lat_1= and lat_2= or lat_0
# +proj=lcc +datum=WGS84 +lat_0=52 +lon_0=3 +lat_1=50 +lat_2=54 +ellps=WGS84
# +towgs84=0,0,0
2 52
-68634.11       472.08  <0.999392 0.999392 0.998785 0.000355057 0.999395 
0.999389 0.788173>
4 52
68634.11        472.08  <0.999392 0.999392 0.998785 0.000355708 0.999395 
0.999389 -0.788173>
4 52.1
68481.15        11591.10        <0.999393 0.999393 0.998787 0.000434814 
0.999397 0.99939 -0.788173>

The convergence sign west and east of the origin are opposite of the 
signs found with the utm. But what confuses me is that when i calculate 
two positions along a longitude east of the origin, the easting reduces 
when going northwards just like with the utm projection.

 From this is seems that TrueNorth = GridNorth - convergence

I think that either the convergence for lcc or utm has the wrong sign. 
But i can't tell which one because i don't know how convergence is defined.

Any thoughts about this?


Kees Krikke

Proj mailing list
Proj at lists.maptools.org

More information about the Proj mailing list