[Proj] gauss-kruger projection

[Mikael Rittri]

Hello, Pieter.
 
-----Original Message-----
> From: Pieter Bosch [mailto:pieter.bosch at za.saabgroup.com] 
> Sent: den 6 februari 2009 11:37
> To: Mikael Rittri; proj at lists.maptools.org
> Subject: RE: [Proj] gaus-kruger projection[Scanned]
> 
> Thanks for your reply
>
> Well I'm generating maps for the African continent with spheroid/ellipsoid wgs84 and a datum called "Hartebeesthoek 94".  > UTM gridzone 34 is one that I will be using  

Hartebeesthoek 94 is, for almost all practical purposes, the same datum as WGS84.
The differences would be caused by contintental drift. Hartebeesthoek 94 is attached 
to the African continent, and follows it as it drifts along.  WGS84 does not. But 
the difference should be less than 1 meter or so. 

> is there a diff between gauss zones and Utm zones?

Yes, UTM zones are standardized. If someone says "UTM zone 34", 
everyone agrees what it means. The scale factor on the central meridian is 0.9996.

For Gauss-Krüger zones, the scale factor is usually 1.0 exactly, 
but there are many conflicting zone systems.
There are two different global ones (used at least in Europe and Asia), 
and many national ones as well.  So, if someone says just "Gauss-Krüger zone 4", 
you cannot really tell what they mean, without more information.

> What do you mean with "what instance of the Gauss-Kruger projection" ?
> give me a example of a instance of the Gauss-Kruger projection. Pls

Well, Gauss-Krüger is a general method to construct a map projection, 
which can be customized to be good for a particular place on Earth. 
So, by an instance, I mean one particular map grid, used for one particular
part of the word.  That is to say, +tmerc, where the parameters have 
been given concrete numerical values.  

I mean, you cannot just ask the proj list 

  "give me the numerical values for the +tmerc parameters, so that I get Gauss-Krüger".

There are an infinite number of possible numerical values, and the proj list cannot know 
which one you want!  But if you can say, for example,
    "Hartebeesthoek 94 with UTM zone 34, southern hemisphere",
or 
    "Hartebeesthoek 94 with South African Survey Grid zone 23, using South-Oriented Transverse Mercator",
that would be enough. 

The file nad/epsg, in the proj distribution, is full of instances of coordinate references systems.

> An example of my values 
> 
> CM =23 deg East
> Datum = hartebeesthoek 94
> Ellipsoid = wgs84
> UTM zone= 34J
> Lat To convert= -28
> Long to convert=23

If you look in nad/epsg, you will find 
  # WGS 84 / UTM zone 34S
  <32734> +proj=utm +zone=34 +south +ellps=WGS84 +datum=WGS84 +units=m +no_defs  <>

which is used for the southern hemisphere, and might be what you want. 

However, if CM means Central Meridian, note that the Central Meridian 
of UTM zone 34 is 21 deg East, not 23 deg East as you wrote. 
So, perhaps the Coordinate Reference System you want is really not 
based on a UTM projection, but is a CRS that is missing from nad/epsg:

  # Hartebeesthoek94 / Lo23
  # Unable to translate coordinate system EPSG:2050 into PROJ.4 format.

But no worries. I think the following should work: 

  +proj=tmerc +lat_0=0 +lon_0=23 +k=-1.0 +x_0=0 +y_0=0 +ellps=WGS84 +datum=WGS84 +units=m +no_defs

I got the parameters from the EPSG database (www.epsg.org), except that I negated k. 
The negative sign on k ensures that the projected coordinates grow westward and southward,
as they should in the South-Oriented version of Transverse Mercator that Hartebeesthoek94 / Lo23 uses.
(However, if your map appears upside down, just remove the minus sign and try again.)  

> As I understand (correct me if I'm wrong) there is a Lambert Conformal
> (preserves the areas-I think) projection 

Yes, there is a map projection called Lambert Conformal Conic. But it preserves 
angles, not areas.  "Conformal" means angle-preserving. 

> and a Gauss projection (preserves the angles). 

Yes.

> Both of them are Transverse Mercator projections (or a form of it).

No, Lambert conformal conic is not a Transverse Mercator. 
But I do think Lambert was the first to invent a Transverse Mercator projection, 
using formulas for a spherical Earth. 

> the one is based on spherical and the latter is based on ellipsoidal shape of the earth.
 
You must be thinking of the spherical Transverse Mercator that Lambert invented. I have 
never heard that anyone uses it, these days.  

> So how will I know if I'm actually using the Gauss projection. 

I beg your pardon? If you use +proj=tmerc or +proj=utm in proj4, then I promise
that you get a Gauss-Krüger projection. (Please don't omit Krüger from the name!)
But if you are asking, how do I know what kind of coordinates my geodata file 
has, I say: ask the geodata provider.  He/she ought to know. 

> Is using wgs84(a ellipsoidal) enough to secure this.
Only if you also use +proj=tmerc or +proj=utm. Not otherwise. 

I suggest you read Furuti's site on map projections,
    http://www.progonos.com/furuti/MapProj/Normal/TOC/cartTOC.html 
After that, you can learn about geodetic datums from
    http://www.ordnancesurvey.co.uk/oswebsite/gps/information/coordinatesystemsinfo/guidecontents 

Good luck,

Mikael Rittri
Carmenta AB
SWEDEN
www.carmenta.com