[Talk-GB] Calling all projection experts: WGS84 parameters

[Colin Smale]

On 2013-03-14 23:18, Grant Slater wrote: 

> On 14 March 2013 21:17,
Steve <stephen.peters1 at sky.com> wrote:
> 
>> Hello fellow OSMers A
hopefully straightforward newbie question for you... I'm currently
uploading various new polygons via shapefiles created in Quantum GIS.
These use the WGS84 (EPSG:4326) projection, which Quantum defines as:
+proj=longlat +ellps=WGS84 +datum=WGS84 +no_defs +towgs84=0,0,0 A fellow
mapper is doing similar, great stuff mapping administrative boundaries
(Parishes etc) from Ordnance Survey, and setting the projection with the
following parameters: +proj=longlat +ellps=WGS84 +datum=WGS84 +zone=30
+no_defs Because of the different parameters (i.e. I have
+towgs84=0,0,0, and he has +zone=30), there is a slight offset in data
we are creating in the same areas. The question is: which one of us is
correct?
> 
> Ordnance Survey data is (normally) EPSG:27700
>
OpenStreetMap is: EPSG:4326
> 
> Zone is used for creating an offset
origin point. Doubt that is what
> you want...

It sounds like the
(incorrect) use of a "zone" value would create an error of the order of
many degrees, not a few metres as Steve and I are seeing. The offset
between his and my line is mostly a matter of 5m or so; "my" line is
further north. If it wasn't for the fact that our lines are juxtaposed
in a particular area, the difference might not even be noticeable to the
"untrained eye" (like mine). 

Is it possible that the +zone=30 is being
ignored because it is not specifying +proj=utm? The zone business is
typical UTM isn't it? 

I got the 

Colin 
 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/talk-gb/attachments/20130315/c364dc2f/attachment.html>