Soil Erosion and Conservation - Environmental Modelling: Finding Simplicity in Complexity - page 361

and to track changes in erosion over time can be done
with any of the models discussed above. Often a statisti-
cal sampling scheme is used to take random points over
the area of interest, and to apply the erosion model to
each point (USDA, 1996). In this case, too, we are not
so concerned about the individual prediction for each
point of application, but rather the ability of the model
to predict overall averages of soil loss in a quantitatively
accurate manner. While we know that none of these mod-
els will necessarily predict erosion for a particular site to
the quantitative level of accuracy we would like to see
for survey assessment purposes (Nearing, 2000), each of
the three models does predict the averages for treatments
quite effectively (Risse et al ., 1993; Rapp, 1994; Zhang
et al ., 1996). As with the case discussed above, the issues
related to choosing the correct model are related to the
information desired and the available data.
Conservation compliance, governmental policy mak-
ing, and regulation of land-users' actions follow the same
guidelines as for the other two applications: information
desired and data availability are again the keys to choice
of model. In this case, however, the argument is often
given, most often by the farmer who is being regulated,
that if we know that there are uncertainties in the ero-
sion predictions for individual applications, how can we
be sure that his field is being evaluated accurately. The
answer is, of course, that we cannot be sure. If the model
predicts that the farmer's field is eroding at a rate in excess
of what our society's policy indicates to be acceptable, the
model could well be wrong for this particular field. This
problem is really no different from that faced by insurance
companies as they set rates for insurance coverage. My
home may be more secure from the possibility of fire than
my neighbour's home because I ammore careful than my
neighbour. But unless my home falls in a different cate-
gory (for example, better smoke-alarm protection), I will
not have much luck in going to my insurance company
and asking for a lower payment rate. Likewise, if I am the
farmer, I cannot expect to give a coherent argument for
lower soil loss than the model predicts unless I conduct
some practice, such as reduced tillage or buffers, which
arguably reduces erosion.
Complexity and uncertainty are key issues relative
to the development, understanding, and use of erosion
models for conservation purposes. They are inevitable
considerations because of the many complex interac-
tions inherent in the erosional system as well as the enor-
mous inherent variability inmeasured erosion data. These
issues do not, however, prevent us from using models
effectively for conservation planning. In fact, the scientific

evidence indicates that choice of models, which implies
choice of model complexity, is more a matter of the type
of information desired and the quality and amount of data
available for the specific application. If our goal is to know
to a high level of accuracy the erosion rate on a particular
area of ungauged land, we cannot rely upon the models.
Natural variability is too great, and uncertainty in predic-
tions is too high (see Nearing et al ., 1999; Nearing 2000).
For appropriate and common uses, such as those dis-
cussed above, models can be effective conservation tools.

22.5 Acknowledgements

The precipitation data from the HadCM3 model for the
period 2000-2099 was supplied by the Climate Impacts
LINK Project (DETR Contract EPG 1/1/68) on behalf of
the Hadley Centre and the UKMeteorological Office. The
precipitation data from the CGCM1 model, GHG

+

A1
scenario, for the period 2000-2099 was supplied by the
Canadian Centre for Climate Modelling and Analysis.

References

Arnold, J.G., Williams, J.R., Nicks, A.D. and Sammons, N.B.
(1990) SWRRB: A Basin Scale Simulation Model for Soil and
Water Resource Management , Texas A&M University Press,
College Station, TX.
Arnoldus, H.M.L. (1977) Methodology used to determine the
maximum potential average annual soil loss due to sheet and
rill erosion in Morocco. FAO Soils Bulletin , 34 , 39-51.
Baffaut, C., Nearing, M.A. and Nicks, A.D. (1996) Impact of
climate parameters on soil erosion using CLIGEN and WEPP.
Transactions of the American Society. of Agricultural Engineers ,
39 , 447-57.
Boer, G.J., Flato, G., Reader, M.C. and Ramsden, D. (2000) A
transient climate change simulation with greenhouse gas and
aerosol forcing: experimental design and comparison with the
instrumental record for the 20th century. Climate Dynamics , 16
(6), 405-25.
Chu, S.T. (1978) Infiltration during an unsteady rain. Water
Resources Research , 14 (3), 461-6.
Cusack, S., Slingo, A., Edwards, J.M. and Wild, M. (1998) The
radiative impact of a simple aerosol climatology on the Hadley
CentreGCM. Quarterly Journal of the RoyalMeterological Society ,
124 (551), 2517-26.
Edwards, J.M. and Slingo, A. (1996) Studies with a flexible new
radiation code. 1: Choosing a configuration for a large scale
model. Quarterly Journal of the Royal Meterological Society , 122 ,
689-719.
Flanagan, D.C. and Livingston, S.J. (1995) USDA-Water Ero-
sion Prediction Project: WEPP User Summary. NSERL Report
No. 11. USDA-ARS National Soil Erosion Research Laboratory,
West Lafayette, IN 47097-1196.