The basin output from r.watershed frequently contains narrow, parallel
basins that are not realistic. This, of course, affects channel
determination, too. What might be the reason for this when the source DEM
does not display such inconsistencies?
Sample maps can be provide if desired.
Rich
On Fri, 8 Jun 2012, Rich Shepard wrote:
Sample maps can be provide if desired.
Attached is such a map. The base DEM has a 10mx10m resolution (source:
National Elevation Dataset). The black lines are the HUC-12 sub-watershed
boundaries from the National Hydrographic Dataset. The red lines are the
vectorized bounds from r.watershed with a threshold of 125,000.
I'd like to understand why some of the r.watershed basins look like
squashed worms.
Rich
r.watershed.output.pdf (83.7 KB)
On Fri, 8 Jun 2012, Rich Shepard wrote:
I'd like to understand why some of the r.watershed basins look like
squashed worms.
By trial-and-error with different threshold values I think my original
value (125,000) was much too small. The threshold for this area should be
~400,000.
Rich
Rich,
On Sat, Jun 9, 2012 at 12:21 AM, Rich Shepard <rshepard@appl-ecosys.com> wrote:
On Fri, 8 Jun 2012, Rich Shepard wrote:
I’d like to understand why some of the r.watershed basins look like
squashed worms.By trial-and-error with different threshold values I think my original
value (125,000) was much too small. The threshold for this area should be
~400,000.
Try also multiple flow direction (MFD) algorithm ( -f flag in r.watershed)
regards,
madi
–
Dr. Margherita Di Leo
On Sat, 9 Jun 2012, Margherita Di Leo wrote:
Try also multiple flow direction (MFD) algorithm ( -f flag in r.watershed)
madi,
That was much worse. The basins were even less realistic. I also tried the
-4 option, but that made rectangular basins. I thought the MFD would produce
better results than the SFD, but apparently not; at least, in this one
situation.
Thanks,
Rich
--
Richard B. Shepard, Ph.D. | Integrity - Credibility - Innovation
Applied Ecosystem Services, Inc. | Helping Ensure Our Clients' Futures
<http://www.appl-ecosys.com> Voice: 503-667-4517 Fax: 503-667-8863
On Sat, Jun 9, 2012 at 6:38 PM, Rich Shepard <rshepard@appl-ecosys.com> wrote:
On Sat, 9 Jun 2012, Margherita Di Leo wrote:
Try also multiple flow direction (MFD) algorithm ( -f flag in
r.watershed)madi,
That was much worse. The basins were even less realistic. I also tried the
-4 option, but that made rectangular basins. I thought the MFD would produce
better results than the SFD, but apparently not; at least, in this one
situation.
These nearly straight red lines indicate that the DEM is flat in these
areas. What kind of pre-processing did you apply? Also, to reproduce
the results, the output of r.info of the DEM, g.region and g.proj
would be helpful.
Markus M
On Sun, 10 Jun 2012, Markus Metz wrote:
These nearly straight red lines indicate that the DEM is flat in these
areas.
Markus,
These are mostly the valleys on either side of the mountains.
What kind of pre-processing did you apply?
None.
Also, to reproduce the results, the output of r.info of the DEM, g.region
and g.proj would be helpful.
g.region -ap rast=dem_proj
projection: 99 (Transverse Mercator)
zone: 0
datum: nad83
ellipsoid: grs80
north: 28704673.96487485
south: 28563905.79283388
west: 464582.93917488
east: 628305.08857358
nsres: 29.98895868
ewres: 30.00222639
rows: 4694
cols: 5457
cells: 25615158
g.proj -p
-PROJ_INFO-------------------------------------------------
name : Transverse Mercator
proj : tmerc
datum : nad83
ellps : grs80
lat_0 : 34.75
lon_0 : -115.5833333333333
k : 0.9999
x_0 : 200000.00001016
y_0 : 8000000.000010163
no_defs : defined
-PROJ_UNITS------------------------------------------------
unit : US survey foot
units : US survey feet
meters : 0.3048006096012192
r.info map=dem_proj
Layer: dem_proj Date: Fri Mar 16 17:00:56 2012
DataBase: /home/rshepard/grassdata
Title: ( dem_proj )
Type of Map: raster Number of Categories: 255
Data Type: FCELL
Rows: 4694
Columns: 5457
Total Cells: 25615158
Projection: Transverse Mercator
N: 28704673.96487485 S: 28563905.79283388 Res: 29.98895868 |
E: 628305.08857358 W: 464582.93917488 Res: 30.00222639
Range of data: min = 1654.234 max = 3143.268
Data Description:
generated by r.patch
Comments:
r.patch input="dem2,dem3" output="dem_proj"
Rich