Convert a layer from WGS 84 to WGS 84 / Pseudo-Mercator during GRASS import

Question

I have a bash script that does the following:

export PATH=$PATH:/usr/local/grass/bin
GRASSDATA=/output/grassdata && \
MAPSET=PERMANENT && \
MAP=outputAmtrakMap && \
rm -rf $GRASSDATA && \
(grass -c /input/kx-amtrak-stations-SHP/amtrak-stations.shp -e $GRASSDATA && \
grass $GRASSDATA/$MAPSET/ --exec v.in.ogr --verbose --overwrite input=/input/kx-amtrak-stations-SHP/amtrak-stations.shp output=$MAP snap=0.0001) 2>&1 | tee /output/logfile && \
MAP=outputInstitutions && \
grass $GRASSDATA/$MAPSET/ --exec v.in.ogr --verbose --overwrite input=/input/Colleges_and_Universities/CollegesUniversities.shp output=$MAP snap=0.0001 2>&1 | tee /output/logfile
MAP=outputAirports && \
grass $GRASSDATA/$MAPSET/ --exec v.in.ogr --verbose --overwrite input=/input/World_Airports/world_airports.shp output=$MAP snap=0.0001 2>&1 | tee /output/logfile

I've verified that the import of the first two shapefiles works well. The problem is that the import of the third shapefile results in this error:

   ERROR: Projection of dataset does not appear to match current location.

   Location PROJ_INFO is:
   name: WGS 84 / Pseudo-Mercator
   a: 6378137
   es: 0
   proj: merc
   lat_ts: 0
   lon_0: 0
   x_0: 0
   y_0: 0
   k: 1
   nadgrids: @null
   wktext: defined
   no_defs: defined

   Dataset PROJ_INFO is:
   name: WGS 84
   datum: wgs84
   ellps: wgs84
   proj: ll
   no_defs: defined

   Difference in: proj

   In case of no significant differences in the projection definitions,
   use the -o flag to ignore them and use current location definition.
   Consider generating a new location from the input dataset using the
   'location' parameter.

Is there a way I can have my script convert the WGS 84 world_airports.shp layer to WGS 84 / Pseudo-Mercator without risking corrupted data?

Answer 1

If you use v.import instead of v.in.ogr then GRASS will reproject the input ogr datasource to match the MAPSET.

I would, however, suggest to review your workflow. It seems that you are

1. Deleting and recreating a MAPSET. Why is this necessary?
2. Opening three sessions of GRASS in the same MAPSET. That's not recommended.

Usually you would create a MAPSET once. Then start GRASS in that MAPSET, and run a script at the GRASS prompt that: imports your data, and runs whatever analysis you need. HTH

Responding to the comment:

No, that won't work. When you start GRASS it spawns a totally new shell. So you can't just chain together GRASS commands with the call to start a GRASS session. What is definitely possible, tho' is to start a GRASS session, and have it run a script of commands that you prepare in advance. So you could setup a short script (let's call it my_script.sh) that contains:

MAP=outputMap
v.import --verbose --overwrite input=/input/kx-amtrak-stations-SHP/amtrak-stations.shp output=${MAP}_1 snap=0.0001
v.import --verbose --overwrite input=/input/Colleges_and_Universities/CollegesUniversities.shp output=${MAP}_2 snap=0.0001
v.import --verbose --overwrite input=/input/World_Airports/world_airports.shp output=${MAP}_3 snap=0.0001

Make that script executable. Now you can get GRASS to run this script in one of two ways: Either, as you showed, using the --exec option. i.e.

grass $GRASSDATA/$MAPSET/ --exec my_script.sh

or, you can set the GRASS_BATCH_JOB environment variable, then call GRASS. It will run the commands in the script, then exit. i.e.:

export GRASS_BATCH_JOB="$HOME/my_script.sh"
grass $GRASSDATA/$MAPSET/
# DOn't forget to unset the env variable
unset GRASS_BATCH_JOB

More detail are on the wiki