International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part B7. Istanbul 2004
Ikonos-Multispectral satellite and an image of sensor ASTER of
the satellite Terra.
All the images had been geographically registered so that the
cells of each image were coincident; also the possible geometric
errors had been corrected, as well as, converted for physical
values. Initially, the digital numbers had been converted for
radiancia and, later, converted for reflectance at satellite level
(Robinove, 1982; Markham & Barker, 1986; Hill, 1991;
Mesquita Jr. & Bitencourt, 1997). During the conversion for
reflectance, the values of radiáncia are corrected to the
illumination effect, caused for the relative position of the Sun.
The images of each date had been transformed into IVDN
images (Rouse e/ al. 1973) and were selected only the area of
study inside of the Pé-de-Gigante.
In the sample plots of 30x30m were sampled the index of foliar
area registered (IAF), with optic sensor LAI-2000/LiCor, in
June of 2001 and January of 2002, when the percentages of
covering of the projection of the coverege of the trees had been
quantified, in each plot. The figure 3 showed IVDN images of
the satellites: Landsat-TM, Terra-ASTER and Ikonos-
Multispectral and the position of the 39 sample plots in the
study area.
Figure 3. The sample plots position geografically associated
with the Terra-ASTER and Multitemporal-Ikonos
NDVI image.
c
coefficients of correlation was the
3. RESULTS AND DISCUSSION
The available images and the following parameters of the
vegetation had been carried through correlation analyses
between the IVDN and: height, basal area, cronw area, density
of individuals, vegetation cover, volume of the cronws,
cylindrical volume, total volume and index of foliar area, of
each parcel of 30x30m. In table 1, are presented the coefficients
of correlation between the parameters and the IVDN. The
parameters of the vegetation and images IVDN had been
ordered in the table based on the coefficients of correlation, the
greater for the minor, respectively, of the right for left and from
top to bottom. The parameter that presented the best
coefficients of correlation with all the
cylindrical volume, the image that presented the highest
Landsat-TM image of
images was the
September of 1994.
Volume
Volume Área Densidade Area de Volume
Cilindrico — Altura Basal de Copa — Cobertura Copas total
(m?) (m) (m^) Individuos (m^) (Yo) (m!) (m) JAF Total
Lundsat- 1
set/1994 0.82 0,89 0.80 0,82 0,88 0.73 0.71 0.69 0.24 : f
Landsat-
jul/1996 0.79 0.84 0,80 0.80 0.86 0,80 0 0.74 { t
1
Ikonos- 1
abr/2001 0.67 0.69 0,80 0.80 0.61 0.40 0.75 0.76 0.74 i 5
Landsat- '
set/2000 0.76 0.80 0.72 0.73 0,81 0.78 0.61 0.62 025 + 6.08
Terra- '
'
nov/200
2 0.47 0.43 0.62 0.58 0.41 0,14 0,59 0.61 074 !
'
Lundsat- 1
ngo/1995 0,53 0.53 0.48 0.45 0,54 0.50 0.29 0.28 0.04 | 3,64
Landsat- '
jun/1997 0,42 0.49 0,39 0.44 0.44 0.40 0.30 0.28 02^ | T
Landsat- ;
nov/199 0.03 1
5 0.54 0.53 0.40 0.38 0,46 0,40 0,30 0.24 HIN
1
y |
Landsat- 0.07 ,
jun/1995 0.57 0.50 0.43 0.36 0.39 0.43 0.22 0.14 (3
1
Landsat- '
jul/1995 0.36 0.28 0,47 0.39 0.27 0.06 0.27 0.30 33 I
(^15
Landsat- 1.03 a
jan/1996 0.39 0.27 0.22 0,11 0.15 0,22 0.00 [-0.0 NE
Total’ 6.32 6.25 6.13 3.86 3.82 4.86 / 6
396
Ikonos-Multspectral of 2001 and the Landsat-TM images of
September 1994 are distinguished of others, therefore they had
had the maximum values for 4 parameters. The image of
September of 1994, mainly with the parameters of the structure
of the trunk, and the Multispectral-Ikonos of 2001, mainly with
the structure of leves.
The gotten results had been similar to the results gotten in other
works with Cerrado with Bitencourt-Pereira (1980) and Santos
(1988). Although the amostral area has been relatively small for
the Landsat-TM images, the gotten values of IVDN had been
similar to the gotten ones for Mesquita Jr. (1998). The control
of the taking of data in the field and the work multiscales, with
the Terra-ASTER images and of the [konos-Multispectral, they
make possible qualitative and quantitative profit of the relations
between the parameters of the vegetation and the IVDN.
Internat
0,6
o o
= TC
correlation coeticient
Figure 4
The valu
amount
metcorol
municipa
correlate
rain, of
graphical
point of t
For the c
curves ha
of accum
phase lag
displacen
temporal
coefficier
phase in t
period arc
The gotte
data and
dependen:
the case o
space res
best resul
Multispec
other auth
another p
depthness
crown of t
Beyond th
sampling
including
epifts alre
et al., 199
IVDN int