ea t t nt, n m
um RUE —————
en REE a MEERES VR ai netu
—— ——
In this application only four categories were retained: coniferous forest.
deciduous forest, grassland and cultivated land. These categories were
also mentioned as mapping units in the legend of the NGI 1/25.000 topo-
graphical map (4) of the area, though dated 1966. Indeed, our Landsat
information is dated 1973 and 1976 which may account for some discrepan-
cies to be discussed later.
Such a four category classification, exclusive of areas built over or
otherwise thoroughly influenced by human activity, may seem somewhat sim-
plistic. It is applicable only to thinly populated areas where inclusion
of small hamlets in one or other of the categories mentioned would cause
only small errors at least in a first approximation.
The method was therefore first applied to an area iñ southern Belgium
devoid of major highways and industry, with a rural population living in
small villages that blend harmoniously in a landscape, forested to a large
extent and for the rest dedicated to agriculture.
1. Description of the test site (Fig.1.) (5)
The almost rectangular test site of 7 by 8 km (5 670 ha) lies in the Dinant
arrondissement, Namur province with its central point at approximately
50°15'N, 5°05'E. It is part of a small natural region called the Famenne,
elongated in a SW-NE direction along the western Ardennes piedmont formed
by Eo-devonian outcrops. To the north-west it is bordered by the loess
covered Condroz plateaux. The climate is suboceanic temperate: mean
annual rainfall 850 - 1 100 mm; mean annual temperature 8 - 9? C.
The Famenne can be subdivided into three parts. The south-eastern part
mainly calcareous, is not represented in the test site. The central part,
an 8 to 10 km wide depression 150 to 250 m above sea level, has been cut
in soft Neo-devonian schists. On level sites, its soils are very humid
during rainy periods, dry and cracking during dry spells, due to the imper-
vious clayey subsoil. On sloping sites shallow stony clay soils are found.
About three quarters of the test site are representative of this central
part of the Famenne. The north-western part, of which the remaining
quarter is representative belongs to a 4 km wide swath of higher lying
land - up to 300 m above sea level - which due to the psammitic nature of
its rocks, has better resisted the erosion. It is dissected by steep val-
leys and forms the transition towards the Condroz.
Approximately two thirds of the test site is covered by edaphic oak-horn-
beam forest (Carpinion betuli) treated as copsewood or as mixed copsewood
and timber. Especially in its southern part plantations of conifers have
been installed, mainly Picea abies. Several of these are too small in
extent to be recorded on the topographic map otherwise than symbolically.
The larger forests are rather important for our work as their general
aspect and the form of the surfaces they occupy do not vary much over the
years. They thus provide the necessary landmarks for the positioning
of arable lands and pastures, whose form and destination is of course more
variable in time. The monitoring of the variation of the latter may be
considere
central d
dence of
area is n
2. Data p
The two s
follows:
Sce
1-242-
1-213/
Of the M:
made by ¢
important
natural
The main
to some 1
ble on tl
Computat:
of Leuvei
Universi
into ORS]
was used
ORSER NM
3. Softw
NMAPW T
ft kh £ 5
STATS
et nn 6 Me rn0t 0 Zod
CANAL
f 3