Soil Erosion and Conservation - Environmental Modelling: Finding Simplicity in Complexity - page 355

relationships have also been developed, as we shall see,
between general precipitation data and erosivity. If we
attempted to conduct such a broad-scale study with the
WEPP model, we would quickly find ourselves with com-
plicated sets of analyses, which we would then need to
compose back to the general trends that RUSLE and the
USLE provide directly. There would also be a data prob-
lem in this case, because WEPP requires certain details
of precipitation that are not available from the global
circulation models used to predict future climate change.
In the second study we review here, the objective was
to determine the specific effects of changes in rainfall
erosivity that might occur as a function of changes in the
number of rain days in the year versus erosivity changes
that are expected to occur when precipitation amounts
per day and associated rainfall intensities change. In this
study, the USLE and RULSE would have been largely
ineffective, because these changes are related to process
changes within the system which USLE and RUSLE do
not take into account. We shall see that in this case the
detailedprocess interactions withinWEPP enable us to see
some quite interesting and important system interactions
which significantly impact the results.

affected by extreme precipitation events ( > 50.8mm in a
24-hour period). According to statistical analyses of the
data, there is less than one chance in a thousand that this
observed trend could occur in a quasi-stationary climate.
Karl et al . (1996) also observed in the weather records an
increase in the proportion of the country experiencing a
greater than normal number of wet days.
Atmosphere-ocean global climate models (see
Chapter 9) also indicate potential future changes in rain-
fall patterns, with changes in both the number of wet days
and the percentage of precipitation coming in intense
convective storms as opposed to longer duration, less
intense storms (McFarlane et al ., 1992; Johns et al ., 1997).
Rainfall erosivity is known to be strongly correlated
with the product of the total energy of a rainstorm
multiplied by the maximum 30-minute rainfall intensity
during a storm (Wischmeier, 1959). The relationship first
derived by Wischmeier has proved to be robust for use
in the United States, and is still used today in the RUSLE
(Renard et al ., 1997).
A direct computation of the rainfall erosivity factor, R,
for the RUSLE model requires long-term data for rain-
fall amounts and intensities. Current global circulation
models do not provide the details requisite for a direct
computation of R-factors (McFarlane et al ., 1992; Johns
et al ., 1997). However, the models do provide scenarios
of monthly and annual changes in total precipitation
around the world. Renard and Freimund (1994) recently
developed statistical relationships between the R-factor
and both total annual precipitation at the location and a
modified Fournier coefficient (Fournier, 1960; Arnoldus,
1977), F, calculated from monthly rainfall distributions.
The example study that we want to examine here was
conducted by Nearing (2001), who used the erosivity
relationships developed by Renard and Freimund (1994)
to estimate the potentials for changes in rainfall erosivity
in the United States during the twenty-first century under
global climate-change scenarios generated from two cou-
pled atmosphere-ocean global climate models. The two
coupled atmosphere-ocean global climate models from
which results were used were developed by the UKHadley
Centre and the Canadian Centre for Climate Modelling
and Analysis.
The most current UK Hadley Centre model, HadCM3
(Gordon et al ., 2000; Pope et al ., 2000; Wood et al ., 1999),
is the third generationof atmosphere-ocean global climate
models produced by the Hadley Centre. It simulates a 1%
increase in greenhouse gases for the time period studied,
as well as the effects of sulphate aerosols. The model
also considers the effects of the minor trace gases CH 4 ,

22.3.1 Potential changes inrainfall erosivity in
theUnitedStatesduringthe twenty-first
century

Soil-erosion rates may be expected to change in response
to changes in climate for a variety of reasons, includ-
ing, for example, changes in plant biomass production,
plant residue decomposition rates, soil microbial activity,
evapo-transpiration rates, soil surface sealing and crust-
ing, as well as shifts in land use necessary to accommodate
a new climatic regime (Williams et al ., 1996). However,
the direct, and arguably the most consequential, effect of
changing climate on erosion by water can be expected
to be the effect of changes in the erosive power, or ero-
sivity, of rainfall. Studies using WEPP (Flanagan and
Nearing, 1995) have indicated that erosion response is
much more sensitive to the amount and intensity of rain-
fall than to other environmental variables (Nearing et al .,
1990).Warmer atmospheric temperatures associatedwith
potential greenhouse warming of the earth are expected
to lead to a more vigorous hydrological cycle, with the
correspondent effect of generally more extreme rainfall
events (IPCC, 1995). Such a process may already be taking
place in the United States. Historical weather records ana-
lyzed by Karl et al . (1996) indicate that since 1910 there
has been a steady increase in the area of the United States