Hi Bryon,
I graduated (M.S. degree) in physical geography last year and my
thesis was 'Interpolation, morphometric analysis and water erosion
modeling'. I hope I can be of some assistance for you.
If you want to get reasonable results for LS factor either for
USLE or Revised USLE you should concern on two main problems:
quality of your input data and quality of used techniques.
Is your USGS DEM adequate for this kind of research
and mainly your aim of work/thesis (which is not quite clear for me)?
You mentioned DEM with 100m horiz. resol. and 1m vertical resol.
resampled to 28.5 m. What did you use for resampling?
It is VERY important to have the best quality of input data - the
finer resolution the better. Resampling methods are also
important especially on such flat area. Don't forget that
grass handles integers (you have vertical resol. 1 m).
2. I am not familiar with r.watershed and on-line manual does not
give full information about techniques used in the program.
If you know something about RUSLE and especially LS factor
you may know how it is computed (slopes are segmented, different
expressions for different categories of slope steepness, soil
status, slope length factor depends on ratio for rill and interill
erosion (aproximated through the slope, etc.).
Until now, I don't have very good references for this program.
Anyway, I can propose you one possible solution. There is
a paper of Prof. Moore and Wilson: "Length-slope factors for the
Revised Universal Soil Loss Equation: Simplified method of
estimation" published in Journal of Soil and Water Conservation 47, 1992,
pp. 423-428.
Unfortunately, I don't have this paper (only other papers
from conferences) but the main point is that LS factor for
RUSLE can be estimated using a modified LS factor (slope
length is replaced by upslope contributing area).
This LS factor can be computed easily by r.mapcalc.
Recently released r.flow (available in /grass/incoming directory
on moon server) computes upslope c. areas (flowline accumulation -
see manual page for details).
Also consider the use of better interpolation/resampling methods.
Perhaps other people would give their opinions and experiences on
this issue.
If you have more questions, feel free to contact me.
Cheers,
Jaro
------------------------------------------------------------------------
Jaro Hofierka
Dept. of Cartography, Geoinformatics & Rem. Sensing
Comenius University
842 15 Bratislava E-mail: hofierka@devin.fns.uniba.sk
Slovakia hofierka@geoinfo.fns.uniba.sk
------------------------------------------------------------------------