s of
jugh
id as
t the
atial
ered
and
into
| and
ry to
1 the
le to
n the
t has
ch of
ed to
lue
Porres, Maru
oscillate between —0.09 and 0.09—with the development of vegetation either. The values, as seen with the curvature
parameters, do not only present a very low correlation coefficient, but also in some cases, contrary to prediction, show
negative correlations. From the results obtained, it can be deducted that the hydrological factors show very little
significance in the distribution of the vegetation, regardless of the lithology in which it is located.
4.3. Slope
The slope of the land shows little direct correlation —oscillating between —0.15 and 0.32—, with values higher than
those of K,, Ky, H and CA, but in any event always low and, more importantly, with different tendencies. In reality, the
slope should not establish a direct relationship but instead its influence should only be significant in places with very
steep slopes, where these inclines impose a limitation for the development of life. In any event, assuming that the values
are significant in areas where the slope is indeed a limitation, that significance is underestimated, as the dimension of
the pixel imposes a certain degree of generalisation of the land forms. In this way, the map of slopes does not come to
recognise spaces with slopes greater than 45? when, in reality, they exist. On the other hand, the amplification of the
area does not elevate the significance in any way.
4.4. Potential direct solar radiation
This, of all of the parameters analyzed, is the only one that shows well defined correlation indices and, in some cases,
show a clear influence of the distribution of the vegetation, as has been shown in previous works (Pardo et al., 1999,
Pardo et al., 2000, Urbano, 1999). From the figures obtained, ideas can be deducted that signal the importance of (i) the
solar radiation values for the growth of vegetation, (ii) the lithology and (iii) the time passed since the fire.
Table 3. Correlation coefficient’s potential solar radiation vs. NDVI of the burned areas
dif-ndvi | ndvi87 | ndvi94 dif-ndvi | ndvi87 | Ndvi94 dif-ndvi | ndvi87 | ndvi94
FIRE-1978 |solar radiation -0,25 -0,12 | -0,24 X X X X X X
FIRE-1979 |solar radiation -0,30 -0,42 | -0,50 -0,35 -0,35 | -0,46 -0,23 | -0,21 | -0,37
FIRE-1984 | solar radiation -0,30 -0,18 | -0,32 -0,40 -0,53 | -0,64 -0,38 | -0,10 | -0,33
FIRE-1985 |solar radiation -0,11 -0,23 | -0,23 -0,39 -0,34 | -0,46 -0,65 | -0,33 | -0,66
First, we should emphasize that once sufficient time has lapsed a clear relationship is established between the solar
radiation values and the values of NDVI. In all of the cases it has been observed that the areas that receive more energy
are those that register lower NDVI values. In the graph xx the relationship between the mean NDVI of the studied zone
and the mean solar radiation values is displayed. The regression line that shows the signaled tendency is exhibited.
In any event, significant difference according to the type of underlying material is observed: the areas in which
impermeable materials are found -marl, clay - although interspersed between permeable materials, show a more clear
relationship between solar radiation and NDVI than others in which only permeable materials are found (such as
limestone and dolomite).
NDVI vs. direct solar radiation limestone
maris and
limestone
36000 2 A Keuper
34000 4 e
^ . .
320004 |: mt = |_ineal (limestone)
2 30000 4 e € X ar = lineal (marls and
& 28000 7 ore UO limestone)
e :
* 26000 + ~ — = = - Lineal (Keuper)
x Sonny
24000 ” = zu y TO"*WU,UX UOTZOTY
22000 - ~ R?= 0,0114
20000 : rs y = -28284x + 37866
0,15 0,25 0,35 0,45 0,55 R'z 0,8492
y = -78128x + 50927
NDVI valuess R? » 0,9814
Fig. 1
To assess the significance of the solar radiation, first the regenerative process of the vegetation covering itself should be
characterized. In areas close to the studied zone although somewhat more humid, as is the Garraf plot, close to
Barcelona in Catalunya , it has been observed that in two or three years the coverage has reached 90%. In even more
humid areas like the forests of Pinus sylvestris of Ripollés, in the Catalonian Pre-Pyranees it has been observed five
years after the fire that the foliage index has reached its maximum level (Gracia and Sabaté, 1996). The study of the
NDVI evolution in on our zone of work is separate from this, the process of regeneration is much more slow, observing
a tendency to increase the NDVI even after 15 years have passed. This expansive tendency of vegetation is larger with
International Archives of Photogrammetry and Remote Sensing. Vol. XXXIII, Part B7. Amsterdam 2000. 1175
