[Proj] towgs params and epsg code for Italy regions

[Mikael Rittri]

Hello Massimo,
You wrote
 
> when I don't need more accuracy I use the following +towgs params :
> 
> all the Italian region :
> towgs84 = −225.000, −65.000, 9.000 
> 
> Italian peninsular part :
> towgs : −104.1, −49.1, −9.9, 0.971, −2.917, 0.714, −11.68 
>
> Sardinia :
> towgs84 = −168.6, −34.0, 38.6, −0.374, −0.679, −1.379, −9.48 
> 
> Sicily :
> towgs84 = −50.2, −50.4, 84.8, −0.690, −2.012, 0.459, −28.08 

...

> I agree I can manually add it to the epsg file, but maybe there is a reason 
> because these parameters are not already inside it?  A license problem? 

The country-wide transform is not defined in the EPSG database, as far 
as I know.  Where did you find it?  But the other three are defined, 
plus several more for offshore areas around Italy. 
  So, the problem is ambiguity. There are many possibilities, but there 
is only one Rome40 datum. The policy of Proj.4 is that when a datum has 
several alternative towgs84-shifts in the EPSG database, then Proj.4 
cannot know which one is best, so no one at all is chosen. (Until you 
add some towgs84 parameters manually, I mean.)   

Other GIS software makes other design choices. I have read about some 
- I think it was GeoTools - that attempts to choose the best datum shift
of several possibilies (or at least an arbitrary one).  
But that optimistic approach is sometimes dangerous.  

Best regards,

--
Mikael Rittri
Carmenta AB
SWEDEN
www.carmenta.com 

________________________________

From: proj-bounces at lists.maptools.org [mailto:proj-bounces at lists.maptools.org] On Behalf Of massimo di stefano
Sent: den 9 mars 2009 16:14
To: proj at lists.maptools.org
Subject: [Proj] towgs params and epsg code for Italy regions



Hi All,

usually i need to reproject Italian data
 
from
Gauss Boaga rRome 40 

to 
Wgs 84 

and inverse 

when i need a good precision on a local scale extent, i use a set of control points (IGM95 vertex) 
and apply last sqare methods to perform 2 different geometrics transformation :

1)  conforme transformation (4 parameters valid for only planimetry) ->  

Tx (longitude - tanslation) 
Ty (latitude - translation) 
λ (scale factor) 
α (rigid rotation)

2) affine transformation (6 parameters valid for only planimetry) ->

Tx (longitude - tanslation) 
Ty (latitude - translation) 
λx (longitude scale factor) 
λy (latitude scale factor) 
α (rigid rotation) 
ϒ (angular sliding)


to transform vector files i apply the transformation parameters using pythn-gdal
but maybe (really i don't know) ...
it is possible to use these params directly in proj 
(like we can do with cs2cs or including it inside an epsg code?)

...

an oher question :-/ ...

when i don't need more acuracy i use the follow +towgs params :

all the Italian region :

towgs84 = −225.000, −65.000, 9.000 


Italian peninsular part :


towgs : −104.1, −49.1, −9.9, 0.971, −2.917, 0.714, −11.68 

Sardinia :

towgs84 = −168.6, −34.0, 38.6, −0.374, −0.679, −1.379, −9.48 

Sicily :

towgs84 = −50.2, −50.4, 84.8, −0.690, −2.012, 0.459, −28.08 

i have taken the above params directly from grass gis and i include it in the epsg file that come with the proj source code.
Using it inside the epsg code 3004 - 3003 the errors is drastically reduced.

.. my question :

instead of setting manually a customized projection,
is it possible to define new epsg code based on the +towgs paprameters inside the epsg file 
and redistribuite it with the proj source code ?

... a lot of software student and newbie gis users can beneficiate of it.

i agree i can manually add it to the epsg file, but mybe there is a reason beacouse these parameters are not already inside it ?
a license problem ? 

Thanks to All for any answers,


Massimo Di Stefano
massimodisasha at yahoo.it

epifanio on   irc.freenode.net  /join gfoss











Massimo Di Stefano
massimodisasha at yahoo.it

epifanio on   irc.freenode.net  /join gfoss