Geology Reference
In-Depth Information
20
The Future of Soil Erosion
Modelling
M.A. NEARING
1
AND P.B. HAIRSINE
2
1
USDA-ARS, Southwest Watershed Research Center, Tucson, AZ, USA
2
CSIRO Land and Water Division, Canberra, Australian Capital Territory, Australia
Niels Bohr reputedly commented that 'Prediction
can be very difficult, especially about the future'
(Rosovsky, 1991). Certainly that statement is
true for prediction of processes in nature, includ-
ing soil erosion. It also may hold true when
attempting to predict the future direction of a
field of research. In 1990 Nearing
et al
. published
a paper in the
Soil Science Society of America
Journal
entitled 'Soil Erosion Prediction Research
Needs'. A retrospective assessment of that paper
indicates that it was largely unsuccessful in out-
lining the important advances and changes in
this field of science since it was published. One
might want to keep that in mind when reading
this chapter.
The Nearing
et al
. (1990) review of research
needs was written during a time when the devel-
opment of process-based soil erosion models was
at the forefront of the science. This was a line of
research that began sometime in the late 1960s
and early 1970s (Meyer & Wischmeier, 1969), and
was near its peak of effort at the time. A team of
scientists from the USDA was developing the
Water Erosion Prediction Project (WEPP) model
(Nearing
et al
., 1989; Laflen
et al
., 1997), which
relied on a numerical, steady-state solution of the
sediment continuity equation, and which focused
heavily on modelling inter-storm variations in
the determinant system properties such as soil
erodibility, soil moisture, soil surface conditions,
plant canopy and ground cover. Another team
from Europe, in a project funded by the European
Union, developed a model called EUROSEM
(Morgan
et al
. 1998), which used a dynamic solu-
tion to the sediment continuity equation driven
in part by a hydrological model based on the kin-
ematic wave equation (Woolhiser
et al
., 1990).
EUROSEM is a single storm model that focuses on
infiltration, runoff and erosion from individual
storms, and allows the user to define initial system
conditions for storms. In Australia, Hairsine and
Rose (1992a,b) developed a dynamic solution to
the sediment continuity equation that encom-
passed what were at that time novel and important
descriptions of fundamental erosion mechanics
not explicitly included in other models. These
were based on the concept of balancing simulta-
neous entrainment, deposition and re-entrain-
ment of particles rather than relying on an
independent sediment transport equation. A sim-
plified product from this line of research was later
introduced as the GUEST model (Misra & Rose,
1996; Yu
et al
., 1997). In the Soviet Union,
Larionov (1993) had been working on yet another
process-based description of soil erosion for the
purposes of prediction.
Certainly there has been a great deal of value
derived from the development of process-based
soil erosion models, both in terms of practical
application and advancement of the science. The
engineers who worked on the clean-up of the
Rocky Flats Superfund site in Colorado claimed
