A Tributary of the Mighty Skagit
The USGS DEM was interesting in that it encouraged upper Finney Creek to
flow into Deer Creek. A single pixel was tweaked to restore the
Extensive fractal analysis was done on this area. The results were
written up by Anicet Beauvais in ...
Calculating slope and contributing
area involves important decisions. Aside from the bag of worms of
scale dependency, there are many algorithms vying for our attention.
For the most part, we are content with the ARC/INFO implementation of a
bi-quartic surface fit through a point and it's eight neighbors. However, this is is inappropriate for calculating a stream slope from a DEM. (Of course,
the best stream slopes involve the
calculation of intercepts of mapped streams with contour lines.)
The next image shows the slope from any given river cell to its
lowest neighbor. (This is a good place to compare dithered and approximated
colors. Netscape users see Options:Preferences:Images and Security.)
The next image shows the slope of the river between junctions, derived by
querying the DEM for the elevation at each junction, and dividing the
distance by the stream length.
Grain size is proposed to be function of slope and contributing area. Below is
a plot of D50, the modeled grain size, for Finney Creek. We used the
junction-to-junction slope and the ARCINFO/TOPMODEL
for contributing area.
The constants were derived from field work at the site.