GRASS Pixel Registration

Craig Anderson caa at noaacrd.Colorado.EDU
Tue Aug 24 17:49:07 EDT 1993


Fellow GU's

	I have run into a problem with how GRASS handles pixel registration/
coordinate referencing, my GRASS references have not been able to provide
the appropriate answer, thus the post here.

	1.  In contrast to IDRISI, which uses the lower left as an origin for
a particular pixel, GRASS seems to use the center of the pixel insetead.  Is 
this assumption correct?  When I set my database to 90N,90S,180W,180E and 
query the pixel in the upper left corner of my raster image, via d.what.rast,
the coordinates displayed are for the center of that pixel.


	2.  Secondly,  I am trying to convert station (point) data collected
at grid mesh points of lat/long into my database as a fine raster image,
theoretically the original points would tranlate into pixels which straddle
the bounding rectangle my region as defined by g.region.
	It appears that I have two options;

		a.  I can disregard and eliminate one pixel around the 
margin of my region and have a half pixel area of no data.
		b.  Or I can set the region to a slightly larger size
(half-pixel) and imply that the data wraps at the poles and at the Internation
Date Line.

Questions:
	
	a.  Can GRASS accomodate wraps at the poles and International Date 
Line?
	b.  If so, will this conflict with my existing database which is
set to merely touch these margins?

Thanks, Craig
caa at noaacrd.colorado.edu



More information about the grass-user mailing list