Conversion of lat-Lon rasters to UTM

[Malcolm Williamson]

On Thu, 17 Sep 1998, Bruce Nielsen wrote:

(lines deleted)
> 
> Method 2:  I remember some talk about a command r.proj a couple years
> ago, but I don't know if it was ever done.  I believe Dr. Simon Cox,
> also from Australia, was working on it.  Any other old-timers remember
> this?
> 

Simon wrote a shell script called r.proj, which I will be glad to pass 
along in its 1997 flavor. Below is a brief description of it. I also 
checked the ftp site on the moon (moon.cecer.army.mil) and found a 
different r.proj, written by Martin Schroeder of the University of 
Heidelberg, which claims to work under floating-point GRASS only.

Please note Simon's disclaimer regarding _large_ ascii intermediate 
files! Make sure that you consider the consequences of going from a 
run-length encoded binary file to an uncompressed ascii file. Also note 
that you must have s.sample on your system.

Regards,
	-Malcolm Williamson

-------
# r.proj 
#    copies a cell-map from one GRASS location to another
#    using the PROJ_INFO parameters to re-project the map
#    and the current region settings to determine the location, size and 
resolut
ion
#
#       The user must have write permissions in a couple of places
#
#       r.proj is a script:  it uses 
#       Gerry Evenden's PROJ program to convert the target cell-locations 
to 
#               the source projection
#       Darrell McCauley's s.sample program to resample the source map at 
the 
#               target cell-locations; 
#               defaults to nearest-neighbour resampling 
#                       (appropriate for "category" maps)
#               but can also use bilinear or cubic interpolation 
#                       (appropriate for "continuous" maps)
#
#       Only tested for a limited range of projections.  
#       The intermediate file is ascii so can be very large.
#       The s.sample stage takes a while (!) if the map is large.