[R-sig-Geo] Conversion of points from AGD66 to WGS84?

[Michael Sumner]

As far as I know, the EPSG 4202 init form is not sufficient, and you
need to specify the datum parameters explicitly.  Wherever in the
database "aust_SA" is defined it is not getting the actual parameters
for some reason - I don't know if that requires an install/build
configuration, but I've just done it manually in the past and stored
the values myself.

So,

pts <- matrix(c(145.1167, -37.83333), ncol = 2)

aust_SA <- SpatialPoints(pts, CRS("+init=epsg:4202"))

agd66 <- SpatialPoints(pts, CRS("+proj=longlat +towgs84=-127.8,-52.3,152.9"))

## no difference
spTransform(aust_SA, CRS("+proj=longlat +datum=WGS84 +ellps=WGS84
+towgs84=0,0,0"))

##SpatialPoints:
##    coords.x1 coords.x2
##[1,]  145.1167 -37.83333
##Coordinate Reference System (CRS) arguments: +proj=longlat
+datum=WGS84 +ellps=WGS84 +towgs84=0,0,0

## works now
spTransform(agd66, CRS("+proj=longlat +datum=WGS84 +ellps=WGS84
+towgs84=0,0,0"))

##SpatialPoints:
##     coords.x1 coords.x2
##[1,]   145.118 -37.83183
##Coordinate Reference System (CRS) arguments: +proj=longlat
+datum=WGS84 +ellps=WGS84 +towgs84=0,0,0

It doesn't seem to matter if I use the 3-parameter towgs84 argument
for AGD66 or the 7-param one listed here, but I don't know the details
of why:
http://www.osgeo.org/pipermail/gdal-dev/2006-February/008090.html

There's also this for what it's worth:
http://www.osgeo.org/pipermail/gdal-dev/2007-November/014762.html

I don't know apart from a matter of experience how to find out if the
EPSG presets are appropriate or not, so I'd just recommend doing some
reading on the topic to convince yourself.

(I checked that transformation above against the presets available in
Manifold GIS, which I happen to have verified for this quite some time
ago so I'm confident that it's right.)

There are several resources on the web for verifying this transformation, i.e.

http://www.ga.gov.au/earth-monitoring/geodesy/geodetic-datums/about.html

Cheers, Mike.

On Wed, Apr 4, 2012 at 2:11 PM, ADAM PETER CARDILINI
<apcar at deakin.edu.au> wrote:
> Dear All,
>
> I have a dataset of lon/lat coordinates with about two thirds in datum AGD66/84 and the other third in datum WGS84. From what I have read I think there is about a 200m different between these datum. I want to change all of the AGD66 coordinates to be correct in WGS84 to fix this difference.
>
> My data looks like:
> lon                          lat                           datum
> 145.1167              -37.83333             AGD66
> 144.9756              -38.36555             WGS84
>
> So far I have tried 1:
> # to choose the pts from the dataset
> pts66<-pts[which(pts$datum=='AGD66'),]
> # to assign the AGD66 datum
> pts66 = SpatialPoints(cbind(pts66$lon,pts66$lat), CRS("+init=epsg:4202"))
> # to transform this data to WGS I have tried two different methods
> pts66t = spTransform(pts66, CRS("+init=epsg:3033"))
> # and
> pts66t = spTransform(pts66, CRS("+proj=longlat +datum=WGS84 +ellps=WGS84 +towgs84=0,0,0"))
>
> The first of these transformations, using CRS("+init=epsg:3033"), transforms the results to the following:
> lon                          lat
> 11736428             3368292
>
> While the second, does not result in any transformation:
> lon                          lat
> 145.1167              -37.83333
>
> Why are these two transformations different? I had thought they would do the same thing because I have used "+proj=longlat +datum=WGS84 +ellps=WGS84 +towgs84=0,0,0" to reproject raster files from one CRS to WGS84.
> How can I get the points to transform and stay in the same lon/lat format?
>
> I just want to get the points that were taken in AGD66 to be at the right spot on the map when I use them in conjunction with raster files in WGS84. I think there must be some fundamental concept I am misunderstanding when working with this data. Can anyone help me out and/or point me in the right direction. Thank you very much for your time and help in advance.
>
> Kind regards,
>
> Adam Cardilini,
> PhD Candidate,
> Deakin University,
> School of life and environmental sciences,
> 75 Pigdons Rd, Waurn Ponds,
> Victoria, 3126
> Mob: 0431 566 340
> adamcardilini.wordpress.com<http://adamcardilini.wordpress.com/>
>
>
>        [[alternative HTML version deleted]]
>
> _______________________________________________
> R-sig-Geo mailing list
> R-sig-Geo at r-project.org
> https://stat.ethz.ch/mailman/listinfo/r-sig-geo



-- 
Michael Sumner
Institute for Marine and Antarctic Studies, University of Tasmania
Hobart, Australia
e-mail: mdsumner at gmail.com