Geology Reference
In-Depth Information
more at the end of the storm, when the aggregates
are dominant, whereas on the Cambisol, with a
higher proportion of larger diameter particles,
detachment continues to the end of the storm.
The parameter values estimated by WEPP for
the Cambisol were higher than those for the
Argisol, except for critical shear strength (
t
c
),
which was the same for the both soils. For the
Cambisol, values for
K
i
and
K
r
were overestimated
and values for
t
c
and
K
e
were underestimated. For
the Argisol,
K
i
and
t
c
values were overestimated
and the
K
r
and
K
e
values were underestimated
(Table 10.3; Fig. 10.4). Nevertheless,
K
i
was
slightly overestimated for the Argisol (1.65%) and
strongly overestimated for the Cambisol (161%).
The factor
K
e
was underestimated for both the
Cambisol (
because it had higher rill and inter-rill erosion
values. Furthermore, the calculated critical shear
strength and effective hydraulic conductivity val-
ues are less for the Cambisol than for the Argisol.
Thus, the WEPP equations overestimated inter-
rill and rill erodibility values and underestimated
the effective hydraulic conductivity and critical
shear strength of the Cambisol. For the Argisol,
the inter-rill erodibility and critical shear strength
values were overestimated, and rill erodibility
and effective hydraulic conductivity values were
underestimated. The use of the experimentally-
derived soil parameters as input data to WEPP
improved the soil loss estimates by 306% and
runoff estimates by 135%.
We conclude that there is a great need to use
local data to calibrate the values of the input
parameters used in WEPP to describe the soil
properties in order to achieve better predictions.
The lack of accurate data, such as rainfall inten-
sity, may also cause considerable errors when we
compare measured and estimated values. We also
conclude that for tropical soils, which usually
have higher clay contents than European and
North American soils, there is a trend of overes-
timation, both for runoff and soil loss, when we
compare measured data with data predicted by
WEPP.
90.41%). WEPP
underestimated the critical shear strength for the
Cambisol (
−
81.20%) and Argisol (
−
15.25%) but overestimated it for the
Argisol (496%). Inter-rill erosion was overesti-
mated for the Cambisol (34.26%) and underesti-
mated for the Argisol (
−
47.76%).
The hydraulic conductivity underestimation is
probably related to soil characteristics, which
generated the equations used for the WEPP pre-
diction. The differences are likely to be associated
with the high cation exchange capacity of the
Argisol and high sand content of the Cambisol.
−
Acknowledgements
10.4 Conclusions
We are grateful to Professor Michael Fullen
(University of Wolverhampton, UK) for reading
the first version of this chapter and making useful
corrections and suggestions.
In the two case studies it is evident that there are
considerable differences between measured and
model values. In case study 1, the values for runoff
predicted by WEPP were higher than those meas-
ured. The values of soil loss measured on the 20 m
and 40 m length segments on a 1% slope angle were
higher than those predicted by WEPP. The values
for soil loss measured on the 20 m and 7% slope
were very close to the ones predicted by WEPP, dif-
fering by only 6.7%. On the other hand, on the
40 m and 7% slope, the measured values were lower
than the ones predicted by WEPP. This demon-
strates the need for further model calibration.
In case study 2, we conclude that the Argisol
has higher topsoil erodibility than the Cambisol,
References
Chaves, H.M.L. (1990)
Uncertainty analysis of a steady
state erosion model
. PhD thesis, Purdue University.
Chaves, H.M.L. (1994a) Novidades sobre o Water
Erosion Prediction Project - WEPP. In Pereira, V. de P.,
Ferreira, M.E. & da Cruz, M.C.P. (eds),
Solos
Altamente Suscetíveis à Erosão
. Jaboticabal - SP:
UNESP/SBCS: 207-42.
Chaves, H.M.L. (1994b) Adaptação do modelo WEPP para
as condições brasileiras. In Pereira, V. de P., Ferreira,
