EM 1110-2-2907
1 October 2003
(USPED) model. This model increases the accuracy of erosion and deposition prediction on
uneven terrain.
A = RKLSCP
6-3
where
A
=
estimated average soil loss in tons per acre per year
R
=
rainfall-runoff erosivity factor
K
=
soil erodibility factor
LS
=
slope length factor
C
=
cover-management factor
P
=
support practice factor.
(3) See http://www.iwr.msu.edu/rusle/about.htm for details on the Revised Universal
Soil-Loss Equation.
(4) The two models described above use statistical averages of hill slope segments for
the entire watershed, leading to inaccurate outputs. The SIMulated Water Erosion (SIMWE)
model, the third model used in this study, overcomes these shortcomings by adding a conti-
nuity equation. SIMWE is based on the solution of the continuity equation (solved by
Green's function Monte Carlo Method) that describes the flow of sediment over the land-
scape area. The factors included in the SIMWE model include measurements relating to
steady-state water flow, detachment and transport capacities, and properties of soil and
ground cover. The primary advantage of this model is its ability to predict erosion and depo-
sition on a complex terrain on a landscape-scale, thereby improving land use assessments.
c. Remotely Sensed DEM Data. In an effort to minimize environmental impacts at mili-
tary training sites, CERL scientists evaluated the effectiveness of applying standard soil loss
equations with the use of DEM at varying resolutions. The optimal pixel size for landscape
level erosion and deposition modeling ranges from 5 to 20 m. Most readily available DEM
data is at the 30-m resolution. Higher resolution DEM data are slowly becoming more
available ; for older DEM data sets and the easily accessible Landsat data, it is possible to
interpolate the low resolution data and resample the data at a finer resolution. For this study
the authors converted 30-m resolution data to 10-m resolution data by applying a regular-
ized spline with tension (RST) method, a spatial interpolation tool included in some GIS
software. The method is a smoothing function, which interpolates the resampled data from
scattered data (RST was developed by Lubos Mitas at North Carolina State University).
d. Study Results. The authors illustrated the issues associated with modeling soil loss
Topographic inputs into the models served as both a tool in evaluating erosion potential and
in determining the quality of the DEM. Low quality DEMs hold a high proportion of noise
in the data. The noise in the data creates two related problems: 1) the signals could easily be
interpreted as landscape features, and 2) large terrain features could be obscured by the
noise. Resampling and smoothing techniques using the RST reduced the noise and produced
a 10-m resolution DEM. This process better highlighted prominent topographic features.
(1) The potential for net erosion/deposition was calculated using two different resolu-
tions (the 30-m DEM and a 10-m DEM developed by the resampling of the 30-m data).
6-23
Previous Page
