Dear GRASS experts:
I report a strange behavior of GRASS 7.0.3 on Ubuntu 16.04.
It happens in r.slope.aspect. When I tried to create a
slope map using a resolution bigger than the original DEM's
resolution, the result becomes strange.
Here is a simple example (please look at the code at the
end of this e-mail). First, it creates an imaginary DEM
on which "elevation" is a column number in 100 pix x 100 pix
region of 0N~1N, 0E~1E . It is called testdem.
Using that DEM, r.slope.aspect gives right answer
(test_slope =0.51 deg. It is equal to atan (100 m/((40000000 m)/360)).
Please take a look at the attached picture.
However, after setting the resolution twice as big (0.02 deg) as
the original one (0.01 deg), r.slope.aspect happens to give a
wrong answer (test_slope1 =0.0257, which is ohe-half of the right answer).
However, when I resample the original DEM by r.mapcalc
command, and try r.slope.aspect command using the
resampled DEM (called test_dem2), this time it gives right
answer (test_slope =0.51 deg).
Obviously these three results should be the same because the
original DEM is a homogeneous slope.
I suspect something wrong in the treatment of resolution in
r.slope.aspect commands.
(code)
g.region w=0 e=1 s=0 n=1 rows=100 cols=100
r.mapcalc expression="testdem=col()" --o
r.slope.aspect elevation=testdem slope=test_slope0 --o
g.region res=0.02
r.slope.aspect elevation=testdem slope=test_slope1 --o
r.mapcalc expression="testdem2=testdem" --o
r.slope.aspect elevation=testdem2 slope=test_slope2 --o
d.mon wx0
d.rast test_slope0
d.rast test_slope1
d.rast test_slope2
---------------
Kenlo Nishida Nasahara
Faculty of Life and Environmental Sciences
University of Tsukuba, Japan 305-8572
24dakenlo@gmail.com
(attachments)
Sorry, I found that it is avoided by "-a" option.
It is described in the manual.
"If the user really wants the raster elevation map resampled to
the current region resolution, the -a flag should be specified. "
https://grass.osgeo.org/grass73/manuals/r.slope.aspect.html
Kenlo Nasahara
On Mon, 15 Jan 2018 12:28:41 +0900
NASAHARA Kenlo <24dakenlo@gmail.com> wrote:
Dear GRASS experts:
I report a strange behavior of GRASS 7.0.3 on Ubuntu 16.04.
It happens in r.slope.aspect. When I tried to create a
slope map using a resolution bigger than the original DEM's
resolution, the result becomes strange.
Here is a simple example (please look at the code at the
end of this e-mail). First, it creates an imaginary DEM
on which "elevation" is a column number in 100 pix x 100 pix
region of 0N~1N, 0E~1E . It is called testdem.
Using that DEM, r.slope.aspect gives right answer
(test_slope =0.51 deg. It is equal to atan (100 m/((40000000 m)/360)).
Please take a look at the attached picture.
However, after setting the resolution twice as big (0.02 deg) as
the original one (0.01 deg), r.slope.aspect happens to give a
wrong answer (test_slope1 =0.0257, which is ohe-half of the right answer).
However, when I resample the original DEM by r.mapcalc
command, and try r.slope.aspect command using the
resampled DEM (called test_dem2), this time it gives right
answer (test_slope =0.51 deg).
Obviously these three results should be the same because the
original DEM is a homogeneous slope.
I suspect something wrong in the treatment of resolution in
r.slope.aspect commands.
(code)
g.region w=0 e=1 s=0 n=1 rows=100 cols=100
r.mapcalc expression="testdem=col()" --o
r.slope.aspect elevation=testdem slope=test_slope0 --o
g.region res=0.02
r.slope.aspect elevation=testdem slope=test_slope1 --o
r.mapcalc expression="testdem2=testdem" --o
r.slope.aspect elevation=testdem2 slope=test_slope2 --o
d.mon wx0
d.rast test_slope0
d.rast test_slope1
d.rast test_slope2
---------------
Kenlo Nishida Nasahara
Faculty of Life and Environmental Sciences
University of Tsukuba, Japan 305-8572
24dakenlo@gmail.com
--
奈佐原 顕郎 （旧姓西田）
筑波大学生命環境系
〒305-8572 つくば市天王台1-1-1
電話・FAX： 029-853-4897
Kenlo Nishida Nasahara
Faculty of Life and Environmental Sciences
University of Tsukuba, Japan 305-8572
24dakenlo@gmail.com