[GRASS-user] Setting Initial Projection/Bounds

[Glynn Clements]

Rich Shepard wrote:

>    I keep tripping over new issues. If there's a doc discussing this, please
> point me to it and I'll go read it.
> 
>    My Nevada data come from a variety of different sources and have different
> projections (Alberts Conical Equal Area [aea] and Long/Lat [ll], if not
> more) and coverage boundaries. This for both vector and raster data.
> 
>    I want to put all these in a GRASS state data location and PERMANENT
> mapset. Once all data are in I can re-project and change regions as
> necessary.
> 
>    Importing the state and county boundaries I set the bounds by long/lat (in
> decimal degrees) and grass took the aea projection information from the .e00
> file. When I tried to import hydrologic unit boundaries from the National
> Hydrologic Database, v.in.ogr on a .shp file and using the -o option to
> override the defaults, failed becaue the projection of these data are
> lat/long.
> 
>    Is the appropriate response for me to set up a suite of locations, one per
> projection, to initially import all data and then reproject them into a
> common location? Or is there a more appropriate way to get all the disparate
> data into grass?

Each location has a single projection. If you have data in multiple
projections, you need to either re-project outside of GRASS (with e.g. 
gdalwarp) first, or create one location per projection.

Also, the region bounds must be in the coordinate system used by the
data. If you have lat-lon data, the bounds must be in lat-lon degrees. 
If you have AEA data, the bounds must be given in the AEA coordinate
system.

If the data has georeferencing information, the easiest solution is to
use "r.in.gdal -c ... location=..."; this will import the data into a
new location with the correct projection and bounds.

>From there, you can use r.proj to reproject data to the working
location.

-- 
Glynn Clements <glynn at gclements.plus.com>