ap of 
] very 
's and 
; been 
plants, 
of the 
o. The 
ve the 
partly 
o from 
1ght to 
ing the 
ogenic 
of the 
s. The 
ogenic 
ry has 
INVENTORY AND THE FLORA CARTOGRAPHY METHODOLOGY USING THE COMBINATION OF 
REMOTE SENSING AND GIS . CASE OF THE AIN RICH REGION (ALGERIA). 
F. BENHAMOUDA', M. CHOUIEB", M. CHIKH'& A. LATTOUI" 
* National Center of Spatial Techniques;, BP 13 Arzew 31200, ALGERIA. 
** National Institute of Higher Education in Agronomy; Mostaganem, ALGERIA. 
ISPRS Commission VII, Working Group 5 
Keywords : Steppe ; Desertification ; Remote Sensing ; GIS ; Inventory ; Cartography ; Flora ; Vegetation. 
ABSTRACT 
The Algerian subsaharian zones are subject to an ecological unbalance; of human and climatic origins; more and more 
threatening and it risks to restrain seriously the future and the socio-economical blossom of the steppe. The spatial observation 
of these territories constitute a diagnostic mean for desertification. The design of an integrated approach using remote sensing ( 
Landsat TM ) and the GIS constitute an important step, in the building up of data bases on the observation stations chosen in 
advance from optimised space images, to the thought generalization for cartographic terrain surveys and the setting up of an 
observation system "discountinued" of the regions subject to desertification. With an area of 146107. 8 ha, the region of AIN 
RICH was chosen as a pilote zone for the methodology application. The latter has two objectives : the flora inventory and the 
cartography of vegetation at a medium scale. The ecological analysis of the result obtained shows a general tendency of 
degraded pathways. The best Stipa Tenacissima cover represents just 2.63% of the region area, then 22.2% of the area 
occupied by degraded steppes of Stipa Tenacissima and Artemesia herba Alba. However, the degraded steppes ( Noaea 
Mucronata ) without a great pastoral interest, occupy 16.7% of the total area. The result obtained are promising and can be put 
high, if a continuous daily observation is ensured (e.g : Earth Observing System and SPOT Vegetation). 
1. INTRODUCTION. 
The development of the semi-arid zones constitute a strategic 
challenge of ALGERIA of the third millennial. These zones 
apart from their dimensions, contain natural facilities in 
terms of enormous potentialities. They are currently 
compromised to an ecological unbalance; of human and 
climatic origins; besides in addition threatening and which 
risks hindering seriously the future and the socio - 
economical blossom of the steppe. The spatial observation of 
these territories constitute a mean of diagnosis in matter of 
desertification. The conception of an embedded approach 
using remote sensing data (TM of Landsat 5) and GIS 
constitute an important step, in order to ally two objectives: 
the inventory of the flora and the cartography of the 
vegetation of the pilot zone of Ain Rich. It consists the 
development of data bases on the observation ground 
stations, previously chosen from optimized images, to the 
generalization argued to cartographic finality from ground 
plotting and to setting up a system of observation « 
discontinuous» for a region subject to desertification. 
2. METHODOLOGY. 
The problematic of choice of the methods of studies of the 
flora, which are based on the ground sampling and therefore 
on the statistics (Factoriel Analysis of the Correspondences.), 
is approached by the selection of the localities of taking on 
the basis of the criterias: of localization, of accessibility and 
of spatial distribution of the stations on the space of survey. 
For the number of the ground ploting, conditioning the 
representativity of the vegetation inventory, we choose 11 
stations (of which 09 only were « visited», see table 1) on the 
basis of a spectral introverted homogeneity on the TM (TM 
4, 3 and 1) color composite, which translates the dominant 
plant species as well as its state of degradation. The spectral 
International Archives of Photogrammetry and Remote Sensing. Vol. XXXII, Part 7, Budapest, 1998 
homogeneity is correleted to some similar conditions of the 
habitat in all its extent, that it acts on the physical medium or 
some biotics interferences (see Figure N° 1). 
Inside every station, several ground plotting were 
accomplished on some placets on the numbers 3 to 5. The 
unit of sampling represents a hectare, the places (of circular 
shapes having a radius of 17.84 m) is situated on the four 
(04) extremities of the station and in the middle, which are 
distant from the four first at 200m. The stratified sampling 
adopted allowed us to get some information (a hundred of 
parameters) which reflect the environment conditions of the 
vegetation and the flora. 
These information were introduced in a data base of a 
Geographical Information System Mapinfo 4.0. It is bound 
to the layer of information representing the stations and 
places. The data, in particular those concerning the steppic 
grouping and its rate of recovery, respectively identified and 
appreciated on the terrain, served to the automatic 
classification of 06 TM channels (TMI, 2, 3, 4, 5& 7). The 
result of the classification is well correleted with regard to 
the interactive interpretation from satellite images. The scale 
of representation of the cartographic results is the 1 : 50,000 
(amount held of the spatial resolution of TM data 30 x 30 m 
and of the scale of sampling plan which is 1: 50,000). A 
representation of the information result to the scale the 1 : 
200,000 made possible with the ease offered by the G.LS, 
permitting thus a synthetic analysis. 
669