[GRASSLIST:2493] Problem with Cylindrical Equal Area Projection

[David Orme]

Hi all,

This is partly to answer a question I asked the list last year about a 
mismatch in Cylindrical Equal Area projection coordinates values 
between grass and other GIS programs (ArcGIS and GeoTrans). I've 
finally got around to chasing it up and talked to Frank Warmerdam and 
Gerald Everden - there is a bug in CEA projections using proj4.4.7. The 
problem looks, at a cursory glance, to only affect CEA projections with 
latitudes of true scale >0 - the higher the lat_ts, the greater the 
discrepancy (see below). I've used Gerald's libproj4 and lproj test 
program - this gives the same values provided by ArcGIS and GeoTrans. 
I've posted this on proj4's bugzilla site at Frank's request - this is 
just for information for anyone using CEA in grass.

Thanks,
David Orme

> doibook:~ dorme$ proj +proj=cea +lat_ts=0 +ellps=WGS84 # PROJ CEA
> 0 70
> 0.00    5976952.48
> ^C
> doibook:~ dorme$ lproj +proj=cea +lat_ts=0 +ellps=WGS84 # LPROJ CEA
> 0 70
> 0.00    5976952.48
> ^C
> doibook:~ dorme$ proj +proj=cea +lat_ts=30 +ellps=WGS84 # PROJ BEHRMANN
> 0 70
> 0.00    6901590.24
> ^C
> doibook:~ dorme$ lproj +proj=cea +lat_ts=30 +ellps=WGS84 # LPROJ 
> BEHRMANN
> 0 70
> 0.00    6895812.59
> ^C
> doibook:~ dorme$ proj +proj=cea +lat_ts=45 +ellps=WGS84 # PROJ GALL 
> ORTHO
> 0 70
> 0.00    8452687.25
> ^C
> doibook:~ dorme$ lproj +proj=cea +lat_ts=45 +ellps=WGS84 # LPROJ GALL 
> ORTHO
> 0 70
> 0.00    8438529.02
> ^C