to FAO,1984.
. Both methods (RUSLE and FAO) measure the following
factors ,which affect the amounts of soil loss by water
erosion in the field : erosivity of the rains;erodability of the
soil ;topography; vegetation and tillage. The interaction of
the above factors affect water erosion .
Since "erosivity" has a low value in the studied area
(R:40.6) according to RUSLE , and no conservation
practices (e.g. tillage) are applied the other three factors
determine the severity of the process
Keeping in mind the importance of the relief factor in
the water erosion process,the following algorithm was espe-
cially developed for this study,in order to highlight it:
1. [((TM 3- TM 2)*500]/(TM 3 * TM 2)
This index was displayed in the red channel and
combined with :
2) TM 3/TM 4 displayed in the green channel and
3){[ ( TM4-TM3)/(TM4 + TM3)]+0.5}*0.5 (SAVI,Huete
,1988)in the blue channel.
The first channel highlights relief. **
The second channel (combination of red and NIR reflec-
tance) gives information on surface changes due to rain-
fall (2) which is correlated with bare soil (figure 4) and
the third channel is the Soil Adjusted Vegetation Index
(8)which is strongly linked with the vegetation cover measu-
red in the field (12) in Winter and Summer, at the same date
of the satellite path and results are shown in figure 3.
Relief,erodability and vegetation are in this case the
most important parameters in water erosion process; and all
of them are enhanced by the set of indexes (WEEN)
A triangle scheme (figure 5) was developed to predict the
different levels of water erosion through the colours
appearing on the screen, following the variations of colours
along the sides of the triangle.
A visual comparison was performed between the conven-
tional water erosion mapping units and the WEEN enhace-
ment results can be seen in figure 2.
**This relief index was assigned to the first channel
(red) when displayed in the screen because human eyes
can separate features better in this color even with
medium to low contrast. (Pratz, 1978).
S.CONCLUSIONS:
In the wet season the sites from moderate degree on of
water erosion are best identified using the WEEN set of
indexes .The importance of highlighting this level is that the
largest area is affected by a moderate water erosion degree
(Table 4).The sites with severe erosion degree can also be
detected using Principal Components.
In the dry season,only PCA enhacement shows water
erosion from severe degree onwards in the test
sites. Moderate degree was not detected by any
enhacement.
It was found that WEEN was the best algorithm to detect
the studied process in its early stages because it
highlights relief, erodability and vegetation principal active
factors in this area. This possibility is fundamental for the
planning of erosion prevention practices.
TABLE 1:Soil loss tons/ha and year
Less than 10 Tn/ha/year: slight
10 to 50 Tn/ha/year: moderate
50 to 200 Tn/ha/year: severe
more than 200 Tn/ha/year: very severe
TABLE 1:
MAP HAZARD ACTUAL
UNITS
R*K*LS
1 13 5.85
2 29.8 110
3 280 280
4 108 48.6
5 15 0.5
6 254.6 62.4
7 78 58,7
8 0 0
386 Intemational Archives of Photogrammetry and Remote Sensing. Vol. XXXII, Part 7, Budapest, 1998
2: Fi
DI
|
LE
