[GRASS5] Module differences: r.flow/r.flowmd

GRASS GRASS developer
1

  In my naivte, it appears that there are no differences between r.flow
and r.flowmd.

  The former calculates slope curves (flowlines), flowpath lengths, and
flowline densities (upslope areas) from a raster digital elevation model.

  The latter produces slope lines (flowlines), flowpath lengths and flowline
densities from a raster digital elevation model using a modified multiple
directions algorithm.

  Both take the same input, so what makes each different? If the differences
are minor, I strongly urge that they be combined into a single module
(r.flow) for the 5.0-stable release.

Thanks,

Rich

Dr. Richard B. Shepard, President

                       Applied Ecosystem Services, Inc. (TM)
            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
                 Making environmentally-responsible mining happen.

2

There are quite a few differences between these two programs - the method
is generaly the same (vector-grid) but the implementation differs as well as
capabilities.
Also the results differ slightly. If there are any volunteers interested in
merging
those two programs both Jaro Hofierka and I will be happy to help.
Essentially r.flow is a complete rewrite of the original GRASS4.2 r.flow while
r.flowmd
is the original GRASS4.2 r.flow with added multiple directions capabilities for
computation
of upslope area, which improves the map of water flow (flowline density) in the
areas
with dispersal flow (such as shoulders). It was the lack of coordination on our
side
that resulted in the two versions. Jaro will be contributing s.flow program to
GRASS5.1
(it generates flowlines from the given site data) and I can talk to him about
the possibilities
to merge the two r.flow programs, but it is not as trivial as it appears to be.

Here are the main differences:

1. r.flow has an option to compute slope and aspect internally thus making the
program
capable to process much larger data sets than r.flowmd. It has also 2 additional
options
for handling of large data sets but I am not sure that they work properly.
2. the programs handle the special cases when the flowline passes exactly (or
very close)
through the grid vertices differently.
3. r.flowmd has the simplified multiple flow addition so the results are
smoother.

In conclusion, r.flowmd produces nicer results but is slower and it does not
support
as large data sets as r.flow.

I use r.flow and r.flowmd very often and I have tested both for Markus just a
few weeks ago
and I did not find any problems. Rich, your suggestion to rerun the analysis
with r.flow was
useful - because r.flow does not require slope and aspect, if the results are OK
the problem
may be in r.slope.aspect. Also, there may be a problem with the import of the
ANUDEM result.

Helena
Rich Shepard wrote:

  In my naivte, it appears that there are no differences between r.flow
and r.flowmd.

  The former calculates slope curves (flowlines), flowpath lengths, and
flowline densities (upslope areas) from a raster digital elevation model.

  The latter produces slope lines (flowlines), flowpath lengths and flowline
densities from a raster digital elevation model using a modified multiple
directions algorithm.

  Both take the same input, so what makes each different? If the differences
are minor, I strongly urge that they be combined into a single module
(r.flow) for the 5.0-stable release.

Thanks,

Rich

Dr. Richard B. Shepard, President

                       Applied Ecosystem Services, Inc. (TM)
            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
                 Making environmentally-responsible mining happen.

_______________________________________________
grass5 mailing list
grass5@geog.uni-hannover.de
http://www.geog.uni-hannover.de/mailman/listinfo/grass5

3

On Sun, 17 Jun 2001, Helena wrote:

There are quite a few differences between these two programs - the method
is generaly the same (vector-grid) but the implementation differs as well
as capabilities. Also the results differ slightly.

Helena,

  Thank you _very_ much for the explanation.

If there are any volunteers interested in merging those two programs both
Jaro Hofierka and I will be happy to help.

Jaro will be contributing s.flow program to GRASS5.1 (it generates
flowlines from the given site data) and I can talk to him about the
possibilities to merge the two r.flow programs, but it is not as trivial
as it appears to be.

  Permit me to suggest that the two of you are much better qualified to
merge the two because of your intimate familiarity with them. I can
understand how two slightly different versions came to be written rather
than one. But, if I understand correctly, the best features of each should
be retained in a single program. It would certainly make life easier for the
rest of us.

In conclusion, r.flowmd produces nicer results but is slower and it does
not support as large data sets as r.flow.

  This in itself argues for a single version. Big, fast and pretty. :slight_smile:

  BTW, when I offer suggestions, I intend no slights or offense. My
background includes using GIS (ARC/Info, Idrisi, and MapInfo) since 1987
even though I only now developing projects using GRASS. I will look at
things from the perspectives of the non-technical end user because much of
my time is spent translating complex technical issues to the decision-makers
who have the technical background of a banana slug (no offense to banana
slugs, for they are the state animal of Oregon).

Rich

Dr. Richard B. Shepard, President

                       Applied Ecosystem Services, Inc. (TM)
            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
                 Making environmentally-responsible mining happen.