Also, with regard to a raster GIS, I assume that the length of slope, one
input to figuring the factor 'LS', is the distance across each individual
cell, or the diagonal distance. How do you decide which to use? Does it
matter?
Definitely yes. It's difference to use 8 possible directions tracing
flow lines on a raster DEM or e.g. 360 directions. The first one predicts
the longer slopes as the reality is. I suggest to use r.flow program
placed in grass/incoming directory on moon. It uses a raster-vector
approach in construction of flow lines on raster-based DEMs. That means,
it constructs smooth flow lines and computes their lengths. I advise
to read a manual to get more information on this.
What algorithm does GRASS use to compute slope anyway, and is this
distance question related?
The best way is to use s.surf.tps or s.surf.2d /r.resample.tps/
from interpolation library placed also on the moon. They compute
elevation, slope and aspect maps required by r.flow in appropriate
way.
When you have slope lengths and slope steepness raster maps you
can use r.mapcalc to compute LS factor for USLE according to
its definition.
Jaro
--------------------------------------------------------------------
Jaro Hofierka
Dept. of Cartography, Geoinformatics & Rem. Sensing
Comenius University
842 15 Bratislava
Slovakia E-mail: hofierka@devin.fns.uniba.sk
--------------------------------------------------------------------