|
|
|
Using
Soil Erosion Modeling for Improved Conservation Planning:
A GIS-based
Tutorial
|
Deriving erosion modeling
inputs from a land cover/land use map
Land use reflected by land cover
has a profound impact on erosion and deposition. Surface cover,
such as vegetation or plant residue may intercept raindrops
(reducing their energy and ability to dislodge soil particles),
increase infiltration, slow down runoff and reduce
transporting capacity of water flow.
Among the factors influencing erosion (e.g. rainfall, topography,
soil properties, cover), land cover is the one most influenced by human
activities and it is also the one that is the easiest to manage
and change. Land cover is therefore the major
factor considered for conservation planning.
In the RUSLE3D and USPED models the impact of land cover
is represented by the C-factor.
C-factor is the ratio of soil loss from land under specified
conditions to the corresponding soil loss from clean-tilled
continuously fallow land under identical conditions. An
increase in the cover factor represents a decrease in protection
provided by vegetation and an increase in the estimated erosion rate.
Typical values for well protected land would be 0.005 to 0.1.
Because C-factor is a relative, empirical number, the best resource for obtaining
values for a given area is the local office of the Natural Resources
Conservation Service (NRCS). Many considerations go into
estimating C-factor for a given land cover and area including: canopy (plant
cover above the soil), crop rotation
schedule, tillage practices, organic residue, surface roughness,
and soil moisture. While some of these variables will vary throughout
the year due to seasonal changes, the C-factor is an average annual value.
Here we provide general guidance based on
our experience and links to relevant internet resources.
Sources of digital land cover data suitable for erosion modeling
-
The
USGS National Land Cover Data (NLCD) based on LANDSAT TM at 30m resolution
provides land cover classified into 21 classes.
The new NLCD 2000,
based on Landsat 7 ETM+ which is mostly 15m resolution, is a cooperative
effort involving the USGS, EPA, USFS and NOAA and is due to begin production
in 2002.
The data are suitable for erosion estimates at regional and larger watershed
scales.
-
Digital Orthophoto Quarter Quadrangles (DOQQ), black and white and infrared,
have 1m resolution (and higher) and can be used to derive land use/land cover
by image processing and/or digitization and field validation.
Estimating the C-factor
Estimating the C-factor for GIS-based data is different from traditional
field estimates in that percent cover is estimated for precise location(s).
To cover larger areas one needs to rely on the land cover layer and known values
of C-factor for certain vegetation and land cover types.
The C-factor ranges estimated for case studies reported in this tutorial is based on land use classes.
The land use map is simply
reclassed using a C-factor chosen for each land use type. Any additional GIS
layers containing more detailed information that affects C-factor
(e.g. dirt roads not depicted by the land use maps, recently disturbed areas, etc.)
can be utilized to improve the C-factor map. A detailed description of
field estimation of C-factor can be found in the RUSLE manuals at the USLE/RUSLE official web site.
The US Army Engineer Research and Development Center, Construction Engineering
Research Lab (ERDC-CERL) has developed a new C-factor protocol that takes users through steps in combining ground surveys, satellite image
analysis, vegetative index calculations and regression analysis in order to
estimate cover at high resolution. See also: US Army Environmental Center (USAEC)
Conservation Technology pages.
Up
Computing erosion modeling inputs from a DEM
Deriving erosion modeling inputs from a land cover
Erosion modeling inputs
|
 HOME
H. Mitasova, et al., Geographic
Modeling Systems Lab, UIUC
|