Geographic boundaries in GRASS raster/vector files

GRASS GRASS user
1

  I've read the programmer's manual (4.2) over and over, and I still can't
answer my own question. So, I'll let you smarter folks answer it for me. :slight_smile:

  The layer boundary is defined by four values: "north", "east", "south" and
"west". The values (represented in the documents and the spearfish data set)
are in meters and represent the geographic position of the outside edges of
the cells (not the center). That I understand. However, GRASS originally
developed in a simple world where all data sets were projected in UTM. This
results in nice, rectangular map layers with perfectly horizontal and
vertical edges.

  However, almost all of our work is in State Plane Coordinates. As a matter
of fact, several regulatory agencies specify this projection. Well, SPC
results in a layer which is not orthoginal to the sides of the monitor in
which it is displayed. For example, the upper left corner has a northing
value less than that of the upper right corner (same for the lower corners).
At the same time, the upper left corner has an easting with a lesser value
than the lower left corner.

  Given this lack of uniformity in values, how to I enter values when
asked by the program? In other GIS software, the bounds are represented by
the (easting, northing) values of the corners, not the edges. Wouldn't it be
nice if GRASS understood corners, too?

Thanks,

Rich

Dr. Richard B. Shepard, President

                       Applied Ecosystem Services, Inc. (TM)
              Making environmentally-responsible mining happen. (SM)
                       --------------------------------
            2404 SW 22nd Street | Troutdale, OR 97060-1247 | U.S.A.
+ 1 503-667-4517 (voice) | + 1 503-667-8863 (fax) | rshepard@appl-ecosys.com

2

On Fri, Apr 07, 2000 at 07:33:32PM -0700, Rich Shepard wrote:

  The layer boundary is defined by four values: "north", "east", "south" and
"west". The values (represented in the documents and the spearfish data set)
are in meters and represent the geographic position of the outside edges of
the cells (not the center). That I understand. However, GRASS originally
developed in a simple world where all data sets were projected in UTM. This
results in nice, rectangular map layers with perfectly horizontal and
vertical edges.

  However, almost all of our work is in State Plane Coordinates. As a matter
of fact, several regulatory agencies specify this projection. Well, SPC
results in a layer which is not orthoginal to the sides of the monitor in
which it is displayed. For example, the upper left corner has a northing
value less than that of the upper right corner (same for the lower corners).
At the same time, the upper left corner has an easting with a lesser value
than the lower left corner.

  Given this lack of uniformity in values, how to I enter values when
asked by the program? In other GIS software, the bounds are represented by
the (easting, northing) values of the corners, not the edges. Wouldn't it be
nice if GRASS understood corners, too?

There's no difference between corners and edges. Other software just
takes the least/greatest values for min(x,y),max(x,y) to form the
bounding box -- grass refers to edges. So you just take the extrema and
use them as your edges. You might add a little to get an even cell
size. Maybe make a MASK layer to block out those areas outside your
area of interest -- you could use an SPC boundary map to create one.

I've yet to get the State of California (as a whole) to fit in a UTM
projection ;-). So, we use an Albers Equal Area Projection.

--
+----------------------------------------------------+
| Eric G. Miller egm2@jps.net |
| GnuPG public key: http://www.jps.net/egm2/gpg.asc |
+----------------------------------------------------+

3

On Fri, 7 Apr 2000, Eric G . Miller wrote:

There's no difference between corners and edges. Other software just
takes the least/greatest values for min(x,y),max(x,y) to form the
bounding box -- grass refers to edges. So you just take the extrema and
use them as your edges. You might add a little to get an even cell
size. Maybe make a MASK layer to block out those areas outside your
area of interest -- you could use an SPC boundary map to create one.

Eric,

  I suspected as much, thank you for confirming it. Now for the translation:

  north = max(y)
  south = min(y)
  west = min(x)
  east = max(x)

Correct?

I've yet to get the State of California (as a whole) to fit in a UTM
projection ;-). So, we use an Albers Equal Area Projection.

  Well, we all know that California is different. :slight_smile:

  Almost all of our projects are small. I think the largest area covered 440
acres. *Our* problem is finding data of sufficiently large scale to provide
the detail we need. Well, ... actually, when it comes to in-river work,
we've covered several miles all at once. But, nothing state-wide so far.

  We are preparing some demos which cover the entire Willamette River
valley, a linear distance of about 150 miles. I know that somewhere there is
a dividing line between the Oregon North Zone and the Oregon South Zone in
the State Plane Coordinates system, but I don't know just were it is. Both
Oregon and Washington handle the problem of two zones by projecting one onto
the other. Hey, that works for us, too!

Thanks again,

Rich

Dr. Richard B. Shepard, President

                       Applied Ecosystem Services, Inc. (TM)
              Making environmentally-responsible mining happen. (SM)
                       --------------------------------
            2404 SW 22nd Street | Troutdale, OR 97060-1247 | U.S.A.
+ 1 503-667-4517 (voice) | + 1 503-667-8863 (fax) | rshepard@appl-ecosys.com