. An
short
ution
ta at
| and
the
ance
)ace)
| the
and
IARE
arted
land,
ed at
then
on (7
apper
on. À
and
The
isses
isses
12 for
" and
Such
jle by
puter
in of
. The
] and
te an
juities
. The
cams,
abase
Ss.
cently
pean
there
cover
ed the
of the
our of
ngary,
t was
launched to extend this type of information for national
level needs. As a first step to satisfy these needs the
scale 1/50 000 and a minimum interpretation unit of 4ha
was adopted. The merged SPOT P (resolution 10 m)
and Landsat TM data integrated in false. A colour
images represent the input data for photointerpretation.
New 4th level of interpretation was introduced: 25
classes of 3rd level (scale 1/100 000) were divided into
63 classes of the 4th level (scale 1/50 000), while 19
classes of 3rd level could not be divided further in the
sense of delimitation criteria (or they do not occur on the
territory of the four proposing countries), their names of
the 4th level as well as definitions agree with the 3rd
level (Feranec et all., 1995). In the frame of the PHARE
programme an experimental CORINE Land Cover
interpretation in the scale 1/50 O00 started in selected
areas of Czech Republic, Hungary, Poland and Slovak
Republic at the end of 1995 and will be finished by the
end of 1996.
In the frame of the 1992 PHARE Regional Programme,
a Remote Sensing Programme was requested in order
to finalise the CORINE Land Cover Programme, but
also to focus on specific applications of remote sensing
for environmental management and assessment. This
project deals with GIS (Geographic Information System)
and the use of remote sensing as one of the basic data
sources, which permits to obtain a tool for decision
makers each within a specific area (Steenmans, C.
Willemsen,H. 1996). Such programmes in Central and
Eastern Europe are:
The Black Triangle GIS (Czech Republic, Poland with
the technical support of the Phare Land Cover Unit).
This project started in the Summer of 1995 and will
demonstrate by means of practical examples the
advantage of using GIS and integrated remote sensing
techniques. Cross-boundary comparable environmental
data are now available from the CORINE Programme.
Detailed land cover information at scale 1/50 000 have
been collected by means of integrated SPOT P and
Landsat TM data using the refined CORINE
nomenclature (level 4).
Danube Delta Project (Romania) is also carried out on
the base of refined CORINE nomenclature (level 4), in
the frame of the PHARE programme. These data will be
used mainly to identify and delineate mapping units
upon the nature and degree of intensity anthropic
induced changes in landscape and ecosystems i.e.
changes in vegetation and soil cover, in hydrological
regime, in silting intensity, soils salinization, wind erosion
on sand dunes etc. Finaly a map with restoration
problems will be compiled as a base for assessing the
priorities to ecological rehabilitation (Vajdea, Munteanu
1994).
As concering the definition, as a general frame, the
Corine Land Cover categories will be used for level 4
categories. But for improving thematic details new
definitions will be added whenever they appear
necessary. To have comparable results with other case
85
studies a coordination of the Central Land Cover Team
to necessary.
Danube Basin Remote Sensing Demo project
(Hungary). The aim of the project which started at the
end of 1995 is to demonstrate the possibility of the use
of CORINE Land Cover data for a practical application,
like hydrology/water management. The purpose is to
estimate, model and analyze the run-off of non-point
pollutants in the catchment of the Zagyva river. This can
be achieved using land cover, provided by the CORINE
project based on satellite imagery, other geographical
data (topography, soils), meteorological data and an
appropriate simulation model. Measured data (river flow,
pollution concentration) will be used for calibration
purposes (National Report of HSO, 1996).
Although the CORINE Land Cover programme
succesfulhas been and has created European standard
for land cover mapping and applications, the type of
methodology chosen (visual interpretation involving an
inevitably of subjectivity) brings serious difficulties for the
necessary updating of the land cover data base at
regular intervals. On the other hand, adopting automatic
or semi-automatic procedures updating, would probably
lead to redefining the original map/data base with a
different, reduced legend. The trade off is in fact difficult
to arbitrate at the present stage.
2. High resolution mapping by making maximum
use of automatic procedures
A number of mapping exercises using high resolution
satellite and automatic procedures were undertaken in
the last few years at national level in various countries,
including Finland, Germany, Great Britain, The
Netherlands (Thunnissen et al.,1993) and Sweden
(Rosengren et al.,1992). Landsat TM data were
generally preferred due to the increased discrimination
potential on soil and vegetation brought by the two
middle infrared bands of TM. SPOT data were also
associated in the case of Finland and Sweden, where
the emphasis was put on forest mapping. The
advantage of using automatic computer classification for
mapping is, of course, the reproducibility of the
processing which minimizes the amount of subjectivity
involved and allows easier regular updating of the data
base. The drawback of this solution lies in the more
limited number of classes included in the nomenclature
produced. They range within a minimum of 12 or 13 for
Finland or Sweden and a maximum of 25 for Great
Britain in the works mentioned above, although it must
be noted that six typical Mediterranean CORINE classes
are not present within those countries.
The scales range from 1/25.000 to 1/100.000 and the
objective of the projects was generally to produce and
regularly update a national land cover data base, with
particular emphasis on the updating requirement. For
Germany (Ellrott, \Wendt,1993), the work was
commissioned by a telecommunications company, for
optimizing the cellular mobile telephone network.
International Archives of Photogrammetry and Remote Sensing. Vol. XXXI, Part B6. Vienna 1996
