International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part B7. Istanbul 2004
TDM(kg/ha)
SEELE LEEE
ac
4
0 50 100 150 200 250 300 3556 400
Julian day
+ Real TDM (kgtha) —— Simulated TDM (kg/ha) |
Figure 7.- Variation of the total dry matter (DM)
accumulation according to the Julian day. Simulated values
(. )and the real values (*) measured in field experiment.
Hidango, season 1992.
Under this scheme, we have used a model divided in two levels
of organization. First of them is related to the information that
feeds the model, its processing and the elaboration of results
archives. In a second level, the biophysical aspects are located,
for which use becomes of four sub-models: Hydric balance,
Phynologic, senescence and growth of the semi-arid
Mediterranean prairie. In Figure 7, the real and simulated
values appear, corresponding to the accumulation of the dry
matter (DM) for the zone of Hidango. When analyzing this
figure is given off that this model, in spite of its simplicity,
reproduces the tendency of the variation in this variable. When
relating, by means of regression analysis, the simulated values
(y) and the measured real values in land (X), permitted to obtain
a significant regression (P< 0,05), with a coefficient of
determination of 97.6% and a standard error of 292.1 kg has”.
Respect to the coefficients of the line of calculated regression,
these parameters did not differ significantly from 0 and |,
respectively (P> 0.05), which indicates the slant non-existence.
The RMSE between the simulated values and the land measured
ones was of 306.6 kg has!, which represents a percentage
deflection of a 12.496 respect to the average of the average
values in land. From these results the dynamics of the prairies is
easily identifiable and to distinguish from shrubs. First they
-have an inter-annual variability; however the others do not
show it among the year. The growth of the prairies begins with
the first rains of autumn, reaching its Maximum appraises in the
months of August and September. By the end of September
they enter in an accelerated process of senescence, which agrees
with the formation of seeds and its later maturation. The
production of annual dry matter is variable; according to if it is
greater or smaller the degree of hydric deficit during the season.
Esteem that accumulation DM of this prairie; would not surpass
the 800 to 2500 kg had'! in normal years.
5. CONCLUSIONS
The present work provides an analysis method oriented to the
space characterization of the inter-annual dynamics of the
vegetation. This analysis allows making the monitoring of the
vegetation when only the NDVI is known. The advantage of the
method is that the data for the classification are extracted from
the series of time of NDVI. Nevertheless, the complication is in
that a great number of images is needed to be able to make it. If
images are used this type of analysis is very advantageous,
562
because it would allow viewing the variation of the spatial
homogeneity index among years. By the way, this variation in
the borders of the zones of SHI would allow identifying the
impact of a certain year in the spatial distribution of the
vegetation. On the other hand, with the use of simple models of
simulation it was possible to consider the beginning of the
growth of the annual Mediterranean prairie, the date in which
the flowering and the length of phynological cycle of this type
of prairies take place. It was possible to simulate the
accumulation of DM and the variation of the water content in
the radial zone of the sub humid Mediterranean annual prairie,
using a simple model, by means of a suitable election and
calculation of the main parameters with physiological
interpretation. Therefore, it was possible to relate the NDVI and
the accumulation of dry matter of the semi-arid Mediterranean
prairie. From this type of models and land information also it is
possible to find simple algorithms that relate the calculated
productivity of the prairie and radiometric indices from satellite
images.
6. ACKNOWLEDGMENTS
This work was funded by the international Spanish agency for
cooperation (AECI) in the frame of the project A/0229/03.
7. REFERENCES
Castellaro G.G.1988. Elaboración de un modelo de simulación
del crecimiento y fenología de la pradera mediterránea anual.
Tesis Ingeniero Agrónomo. 155 p. Universidad de Chile.
Facultad de Ciencias Agrarias y Forestales. Escuela de
Agronomía. Santiago, Chile.
Castellaro, G.G., M. Silva y F. Santibáfiez. 1994. Efecto de la
radiación solar y la temperatura sobre las fenofases de las
principales especies del pastizal mediterráneo anual. Av. Prod.
Anim. 19: 65 —75.
Curran, P. 1980. Remote sensing systems for monitoring crops
and vegetation, Progress in Physical Geography, 4: 315 - 341.
Chavez, P. S. (1996). Image-based atmospheric correction-
revisited and improved. Photogrammetric Enginneering and
Remote Sensing, 6(9) 1025-1036.
Diallo, O., Diouf, A., Hanan, N., Ndiaye, A . y Y. Prevost.
1990. AVHRR monitoring of savanna primary production in
Senegal, West Africa: 1987-1988. International Journal of
Remote Sensing, 117: 1259 - 1279.
Friedl, M. A., and Davis, F. W., 1994, Sources of variation in
radiometric surface temperature over a tallgrass prairie, Remote
Sens Env, 48, 1-17.
Gong, N. S., & J. R. Miller, 1995, Coniferous forest leaf area
index estimation aling the oregon transects using compact
airbone spectrografic imager data, Photogrametric Engineering
& Remote Sensing, 61, n° 9, 1107-1117.
Goward, S. N., Cruickshanks, G. D., and Hope A. S., 1985,
Observed relation between thermal emission and reflected
spectral radiance of a complex vegetated landscape, Remote
Sens Env, 18, 137-146.
Inte
Inı
Ne
Hu
Re
Jai
As
Ka
soi
gr
Ka
ve:
Re
La
CO!
Sp:
Se
La
ve:
an:
Lit
mc
atn
45
Lo
19
Col
Lai
Shi
veg
site
Mc
19°
tec
(Ec
sus
ma
Em
Sar
clir
[VS
Ap
Nei
resi
