| 2004 
itation 
it with 
; used 
and of 
and a 
ording 
hich it 
rosion 
station 
Iber of 
nts: m 
1.67. 
for n. 
in and 
:duced 
reased 
| is in 
iety of 
‚2004; 
timate 
wry et 
useful 
r drift, ' 
International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part B7. Istanbul 2004 
  
The Thornes erosion model was then applied for both dates of 
the two sites using the GRID module of ArcInfo v.8.1. By 
subtracting the runoff and erosion estimates of the earlier dates 
from their respective counterparts of the later dates, a 
qualitative comparison could be made and a possible increasing 
or decreasing trend could be identified. The results for the area 
of Xaló are shown in Figure 4 below: 
a. % change in runoff 
8 >50% increase (2 10-50% increase © <10% change G 10-50% de e m >50% decrease 
    
b. % change in erosion 
= >50% increase © 10-50% increase — <10% change 1 10-50% decrease ® >50% decrease 
  
S CELUM dus ed 
Figure 4: Xalo area (a) runoff estimate of 1978 subtracted from 
runoff estimate of 2000 and, (b) erosion estimate of 
1978 subtracted from erosion estimate of 2000 
Overall, there appears to be little change (<10%) in runoff and 
erosion throughout the Xaló catchment. Nevertheless, the 
figures also help identify areas where runoff rates and the 
erosion potential have both shown a considerable decrease of 
more than 5094, e.g. areas within eclipses A and B in Figure 4, 
possibly due to reforestation practices. Other areas have 
increased their runoff and erosion rates. This is mainly due to a 
significant documented change in landuse in the form of 
agricultural terrace abandonment (area C), or changes in land 
cover in the form of forest fires which are dominant in this part 
of Spain (García-Haro et al. 2001, areas D and E). Finally, in 
some areas downstream, new ‘urbanizaciones’ or urbanized 
areas now dominate the landscape (area F), leading to an 
increase in runoff due to the replacement of the vegetated cover 
with paved roads or pavements. The actual effect this change 
might have on erosion rates and land degradation, is 
questionable since it greatly depends on the quality of the 
construction and the flood prevention engineering work carried 
out. Nevertheless, the ‘visual’ degradation of the environment 
in coastal Spain, with all its consequences on the sustainability 
and the livelihoods, is already quite pronounced and is thought 
to have long reached alarming rates. 
The evolution of the runoff and erosion potential for the island 
of Lesbos is shown in Figure 5 below. Overall, there seems to 
be a greater amount of areas with >50% change in runoff and 
erosion rates than in the case of Xaló. So far the focus was on 
the western part of the island (through a series of 
MEDALUS....199? projects) due to desertification that is 
taking place in this area. However, this study shows that, in 
relative terms, some areas in the south-eastern part of the island 
also appear to be at risk with increased amounts of runoff and 
erosion and therefore, prone to land degradation. The most 
threatened areas seem to be those where forest fires have 
destroyed either the coniferous and deciduous forests (areas A, 
B and C in figure 5) and are now covered by sparse trees and/or 
maquis. The areas where a small change or decrease in both 
557 
runoff and erosion is more pronounced are those in the central 
part of the (areas D, E and between). 
    
    
à. % change in runoff 
EN >50% decrease 
C] 10-50% decrease 
{1 «10% change 
E] 10-5096 increase 
WU >50% increase 
b. % change in erosion 
KM >50% decrease 
C1 10-5096 decrease 
{_J] «1096 change 
(73 10-50% increase Ü 5 10 20. Kiometers 
BN >50% increase 
Figure 5: Lesbos (a) runoff estimate of 1975 subtracted from 
runoff estimate of 1999 and, (b) erosion estimate of 
1975 subtracted from erosion estimate of 1999 
4. CONCLUSIONS 
This study has looked into the possibility of combining 
remotely-sensed with ancillary data, in order to study the 
potential relationship between LULC changes and land 
degradation in the Mediterranean region. This was achieved 
using two largely recognised indicators of degradation: surface 
runoff and sheetwash soil erosion. The results, produced using a 
high rainfall event, have identified areas where runoff and 
erosion figures are increasing as well as areas where these 
figures are less now than they were in the 1970s. 
The methodology can be used in the decision-making process of 
the selection of areas that need mitigation measures to be 
adopted which will reduce or even reverse their degrading 
potential. Specifically, future work should focus on problems 
faced here with the specific datasets, such as: (a) the 
phenological difference of the MSS and the TM data as well as 
the calibration data, (b) the scarceness of the soil samples and 
the actual method of interpolating these over the area. More 
samples should be collected and a more reliable interpolation 
method such as the geostatistical method of kriging, should be 
applied, rather than extrapolating the point values over the 
lithological units. (c) With additional soils data, the soil 
erodibility factor of Thornes should be calculated to replace the 
k of USLE used in the case of Xaló. Finally, issues related to 
the propagation of errors from the combination of data derived