[GRASS-user] Creating mapset and/or importing data based on a proj4 string

[Ken Mankoff]

Dear List,

Most of the time when I work with data in GRASS it is a provided on a
standard GIS grid (EPSG code) and I can work with it easily. I create a
location using the "-c" option when I start GRASS, or I import it into an
existing mapset with r.import.

I'm now working with data and all I know is this:

+proj=stere +lat_0=90 +lat_ts=71 +lon_0=-39 +k=1 +x_0=0 +y_0=0 +datum=WGS84
+units=m +no_defs”

In the past, when working with a few variables and this information, and
lat/lon grids in the NetCDF file, I would export the data (using Python) to
lat,lon,data, and then r.in.ascii and project it that way. However, now I'm
working with 14,000 high resolution daily records on the above projection.

It seems the only way to do this efficiently is to work in the projection
of the data, and r.import the few other variables to this projection. Or
perhaps cdo or gdal or some other 3rd party tool can quickly re-project
everything for me to a known EPSG code. I've solved this before with
vectors, ogr2ogr, and arbitrary proj4 strings.

Can anyone here suggest how to best get this data into GRASS?

I've tried using gdalwarp:

gdalwarp -of GTIFF -s_srs "+proj=stere +lat_0=90 +lat_ts=71 +lon_0=-39 +k=1
+x_0=0 +y_0=0 +datum=WGS84 +units=m +no_defs" -t_srs EPSG:3413
NETCDF:"${file}":RU out.tif

But am concerned because when I then run gdalinfo on the out.tif file, it
reports strange bounds such as:

Origin = (-909.965548780931840,1546.900601999830769)
Pixel Size = (7.477558376763198,-7.477558376763198)
Corner Coordinates:
Upper Left  (    -909.966,    1546.901) (165d27'58.34"E, 89d59' 0.36"N)
Lower Left  (    -909.966,   -1264.661) ( 80d44'10.54"W, 89d59' 8.22"N)
Upper Right (     794.918,    1546.901) (107d48' 8.56"E, 89d59' 2.20"N)
Lower Right (     794.918,   -1264.661) ( 12d50'53.17"W, 89d59'10.36"N)
Center      (     -57.524,     141.120) (157d10'37.34"E, 89d59'54.94"N)

Which seems to be very close to the pole (both lower and upper at >89 deg).
This should be covering a much large portion of the Arctic.

Thanks,

   -k.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.osgeo.org/pipermail/grass-user/attachments/20171006/3eab2bfe/attachment.html>