4,8 and 2 using the seasonal botanical variations
or the seasonal spectral variations.
4 DISCUSSION AND CONCLUSION
The classification of grasslands using seasonal bota
nical and spectral data must be developped according
to the following organigram given in Figiire 4.
Phytosociological releves
PltLftoAoCsLoloqZcaZ tizlzvzA
anatyAiA modoJL
Agro-ecoloqical units
definition
Spectral behaviour in
red and near infra-red.
SzaAonaZ Apzztn.aZ bzka-
vZqua modzZ.
I
Spectral and botanical classification
of grasslands.
Figure 4 .Orqanigram of botanical and spectral
classification of grasslands.
It is necessary to describe phytosociological rele
vés and classify them according to sigmatist method.
Once this step acheived, the more abundant species
are selected. They enable to define and describe
the agro-ecological units.
Phytosociological units my be converted in agro-
ecological units, but reverse is impossible as there
has been a loss of informtion meanwhile.
The agro-ecologiacl units are characterized by spec
tral behaviour, particularly in red and near infra
red wavelengths. Once their spectral behaviour deter
mined, it is possible to cluster any new relevé or
spectral set of data to one or another agro-ecologi
cal unit.
On the contrary it is not possible to define first
a spectral behaviour and then try to create agro
nomical units as the worse agronomical value.units,
on very dry or on muddy soils show spectral beha
viour similarity and are difficult to separate.
It is the first time that a combined botanical and
spectral classification is tempted. It has been ela
borated in order to use reflectance values measured
in the field for green biomss evaluation, and sate
llite data for grassland classification and survey
over large areas.
REFERENCES
Braun-Blanquet, J. & J. Pavillard 1925. Vocabulaire
et sociologie végétale. Bull. Soc. Bot. Fr. 72:
910-911.
Brown. R.J. F.J. Ahern, 1983. Alberta rangeland as
sessment using remotely sensed data. Centre Cana
dien de Télédétection, Research report 83,1:128 p.
Curran P.J 1983. Estimting green LAI from multispec-
tral aerial photography. Photogramm. Eng. Remote
Sensing. 49: 1709-1720.
Everitt. J.H. 1985.Using aerial photography for de
tecting blackbrush (Acacia rigidula) on South Texas
rangelands. J. Range Management. 38 (3): 228-231.
Gates D.M. H.J Keegan, J.C.Schleter,V.R. Weidnerl965.
Spectral properties of plants. Applied Optics.4(1):
11-20.
Gates.D.M. 1970. Physical and physiological proper
ties of plants.In Remote Sensing with a special
reference to agriculture and forestry. National
Academy of Sciences, Washington D.C:224-252.
Grouzis.M. M. Méthy.1983. Détermination radiométrique
de la phytomsse herbacée en milieu sahélien: per
spectives et limites. Acta Oecologica. Oecologia
plantarum.4(18),n°3:241-257.
Guinochet.M. 1973. Phvtosociologie. Masson.227p.
Knipling. E.B. 1970. Physical and phvsioloqical basis
of the reflectance of visible aœd near infra-red
radiation from vegetation. Remote Sensing of Environ.
1 (3) : 155-159.
Richardson. AJ. J.H. Everitt, H.W. Gausmn. 1983.Radio-
metric estimation of biomss and nitrogen content
of Alicia grass. Remote Sensing of Environ.13:179-
184.
Thalen. D.C.P. J.A.Leemburq,R.Beck. 1980. Estimting
aerial biomss in semi-natural vegetation from
spectral reflectance measurementsi)Preliminary
experiences. Acta Bot. Neerl. 29(5/6) :565-578.
Tucker.C. J, C. L. Van Prae t, A.Gas ton, E.Boerwinkel. 1983
Détermination par télédétection de la oroduo tion
primire herbacée aérienne au Sahel sénégalais.
Proiet pilote d'inventaire et de surveillance des
écosystèmes pastoraux sahéliens.FAO-UNEP.
Van der MaarelE. 1979. Transform tion of cover-abun
dance values in phvtosociologv and its effects on
community similarity. Vegetatio.39(2):97-114.
Sympoi
The it
mapp:
A.R.Jon*
Departmen
ABSTRACT:
on the mai
ital forma
of the Lan
ner (MSS)
site in sc
such data
Examples c
ysis and u
1 INTRODUC
One of the
ribes it a
pande d to
or form,,b
surficial
es current
ogist has
ranging fi
pedology a
atic recor
form of a
of discipl
from const
exploratio
is to depi
ure and di
and indica
the develo
Numerous
mapping ex
not forget
(Verstappe
geoirorjholo
have been
most of th
images. In
purposes,
photograph
Mitchell 1
1980, Vers
and deploy
on the Lan
images hav
ologist to
(1974) use
Krinsley (
Iran and M
ery for so
ty of sate
igations h
false colo
deal of ge'
can be ext
assisted
surficial :
cessed dat
as an effe
This pap'
i'ole that
in future
strating t]
