[R-sig-eco] MCP {adehabitatHR} across multiple utm zones

[Michael Sumner]

Hello,

I would recommend not using UTM at all. There are plenty of good
choices of projection, but the right one depends on the extent of your
study area and your purposes.

I would usually choose an equal-area projection centred on my area,
but you'll want to investigate what makes sense for distance
calculation across the various scales it's needed. (This is not
something that seems to get done though it's just as important for UTM
as any other.)

The incantation for spTransform or project in rgdal will look
something like (default unit will be metres):

CRS("+proj=laea +ellps=WGS84 +lon_0=147 +lat_0=-42")

with appropriate values for lon_0 and lat_0. There's a lot more to
this, such as choosing units and providing a false offset X/Y to move
the centre from 0,0. At least note the limitation of project that
cannot transform between datums. The rgdal package is the most
versatile resource in R for this, giving access to the PROJ.4 library,
though there is also proj4 if rgdal is too much.

spatialreference.org is a great resource for finding out about
specific local projections

This paper is a good introduction, but there are plenty of internet
resources about projections:

@ARTICLE{Gudmundsson1998,
  author = {Gudmundsson, Gudmundsur A. and Alerstam, Thomas},
  title = {Optimal map projections for analysing long-distance
migration routes.},
  journal = JAB,
  year = {1998},
  volume = {29},
  pages = {597-605}
}

A UTM zone really is only accurate in its narrow (6 degrees longitude)
band and switching between zones is not something you can do easily.
It could be done but it's a lot of pain for very little, and still you
need to choose a single coordinate system at the end.

These are guides about the broader family of projections that I think are good:

http://www.geo.hunter.cuny.edu/~jochen/gtech201/lectures/lec6concepts/map%20coordinate%20systems/how%20to%20choose%20a%20projection.htm

http://www.georeference.org/doc/guide_to_selecting_map_projections.htm

http://msdn.microsoft.com/en-us/library/cc749633(v=sql.100).aspx

Cheers, Mike.

On Wed, Nov 7, 2012 at 4:36 AM, Kylie Scales <klscales10 at gmail.com> wrote:
> Hi all,
>
> Does anyone know of a way to create minimum convex polygons (mcp,
> {adehabitatHR}) from UTM co-ordinates (SpatialPoints class) that span
> multiple UTM zones?
>
> Or, whether there is another way to pass lat/lon co-ordinates to the
> argument 'unin'?
>
> unin
>
> the units of the relocations coordinates. Either "m" (default) for meters or
>  "km" for kilometers
>
> Any assistance much appreciated.
>
>         [[alternative HTML version deleted]]
>
> _______________________________________________
> R-sig-ecology mailing list
> R-sig-ecology at r-project.org
> https://stat.ethz.ch/mailman/listinfo/r-sig-ecology



-- 
Michael Sumner
Hobart, Australia
e-mail: mdsumner at gmail.com