670
FIGURE 2Scheme of the main stages for processing and sampling the information related to the
erosion hazard mapping
conservation measures are taken in the area.
In this context erosion hazard is the combined
effect of all erosion factors: climate, relief,
soil, land use and management (.Bergsma, 1981 )..
For the evaluation of the erosion hazard
each geomorpho1ogica1 association was sub
divided into mapping units defined on . I'goul d
plotter'.' images at scale 1:100.000 by a dis
tinctive land use pattern and minor land
forms. Jn every mapping unit field measure
ments of the factors of the Universal Soil
Loss Equation were made: rainfall erosivity
(R), soil erodibility (K), relief influence
(LS), and coverage (C). Finally, ranks of soil
loss in tn/ha/year corresponding to different
erosion hazard categories were establish fol
lowing the criteria applied by S.A. El Swaify
(1977) for the evaluation of soil loss in
Hawaii (FIGURE 2). The innacuracy derived from
the computation of ayerage values of the USLE
not disminishe the usefu11 ness of this ap
proach for the erosion mapping at regional’
level. With respect to this Janssen ( 1 983 : 1 20)
said, "in many areas a rough potential erosion
map based on the USLE and covering large areas
might be useful to organizations which intend
to undertake erosion control measures to im
prove agricultural production".
THE LANDSAT INFORMATION
The use of Landsat images gave an acceptable
planimetrie base considering the poor rei la
bility of the regular cartography at small
scales. In the same way the "gould plotter"
enlargements permitted the clear definition
through visual interpretation of mapping units
integrated by typical landforms ( f . e . a 11uvia 1
fans, fluvial valleys, etc.) generally cov
ered by uniforms land use pattern.
The coverage evaluation from mu 1t i - seasona1
images reflecting different biomass densities
was obtanined both by digital processing and
visual interpretation. The use of those ap
proaches depends both on the availability of
adquate equipment and the cost-benefit of
their application taking into account the ex
tension of the studied area and the requiered
mass of field information. According with
Townshend ( 1 981) a visual interpretation can
prove to be much more cost effective for many
task than computer implemented methods. The
amount of data in a single Landsat frame is
enormous and consequently the computational
time for even a large computer system can be
very high. If small scale map production is
required, then pixel by pixel classification
may be un.n e cesa r i 1 y detailed.
RAINFALL EROSIVITY
The rainfall erosivity indexes are parameters
which dérive from the characteristics of the
rain; for their direct correlation with var
ious erosive processes (splash, sheet, gully,
etc.), they are used in the predictions of
the soil loss. The one which is best know has
been developped by Wischmeier and Smith ( 1 978)
who based their investigation on the relation
ships between soil loss and the characteris
tics of the rainfall (quantity, intensity,
impact and d tod moment). The index expresses
the product of the cynetic energy and the
maxima 30 minutes of intensity of a rainfall
1 imitatn
records,
the deve
f i c u 1 t y
was used
fall and
the a d ve
gica1 da
values o
such a ci
eq ua tion
"a" and 1
a 1 c1 ima
ted in t
a high ci
values o
Fo 1 lowin'
cu rves wi
( FIGURE ;
«astern |
eros i v i t’
towa rd 11
crease o
the i n t e
mountain!
thus re s|
ra i n fa 1 1
between |
has been
SOI L ER0I
The sens*
d i f fe ren1
erosion"
erosion <
by the s
than by 1
soil. Ne'
eroded t1
factors ;
internal
"e rodibi
Among t h€
tion, the
n omo g ra p1
Equation,
e rod i b i 1 i
t i ve type
b i 1 i t y c c
tion reqi
centage c
tage of ;
structure
The va1
area are
v i u s re 1 e
the o r i gi
mate. The
t reme wee
presence
mine a me
slope of
q u T e s , u r
decrease
en ce of t
influence
soils. Or
increase
loess a n c
t i on. In
°f e r o d i t
must be a
presence
original
