A STUDY ON THE AUTOMATIC REVISION OF THE EUROPEAN COMMUNITY’S 
CORINE LAND COVER DATABASE USING SATELLITE DATA 
G. Wilkinson, S. Folving, K. Fullerton and J. Mégier 
Environmental Mapping and Modelling Group 
Institute for Remote Sensing Applications 
Joint Research Centre 
Commission of the European Communities 
21020, Ispra, Varese, Italy. 
ISPRS COMMISSION IV 
ABSTRACT: This paper describes work in progress at the European Commission's Joint Research Centre on the 
potential for automatic revision of land cover maps using remote sensing data for the whole of Europe at 1:100,000 
scale. The research on the automatic updating is concerned with a number of important issues in thematic mapping 
such as high accuracy image classification, image data generalisation, correspondences between image classes and 
pre-defined mapping classes. Studies are reported which have been undertaken in Portugal on inter-comparisons 
between first edition and revised land cover maps which reveal considerable subjectivity problems in the conventional 
mapping procedures. Some conclusions are drawn on the implications of these subjectivity issues for automatic 
updating via remote sensing and some possible approaches, involving the possibility of knowledge-based 
decision-making are described. 
KEY WORDS: Land Cover Mapping; Map Revision; GIS; Thematic Mapping. 
1. INTRODUCTION 
1.1. The CORINE Programme 
The CORINE programme of the European Commission 
is concerned with the production of a large Geographic 
Information System containing 22 data layers represent- 
ing the landscape of the twelve member states of the 
European Community (CORINE 1991). Examples of the 
data layers currently being produced are: 
e Biotopes 
e Water Resources 
e Soil Erosion Risk 
e Land Cover 
e Soil Types 
e Climate 
e Administrative Units 
These data layers are being assembled to provide an in- 
tegrated geographical and statistical view of the Euro- 
pean Community, although the cartographic scales of 
the datasets are not all the same. One of the most im- 
portant requirements of the complete information system 
is accuracy both in the spatial and temporal sense. 
Since it will be used widely througout Europe and will be 
used to provide "value-added" products, absolute accu- 
racy in the CORINE base maps is essential to avoid the 
problem of error accumulation in the derived products -a 
problem which can be very serious in using GIS data 
(Heuvelink, Burrough and Stein 1989). 
1.2. The CORINE Land Cover Database 
One of the most important layers in the overall CORINE 
system is the Land Cover database. This database will 
543 
eventually contain a land cover map of the entire Euro- 
pean Community at 1:100,000 scale as a digital polygon 
"coverage" plus attribute information. In order to produce 
such a database a uniform classification scheme - The 
Land Cover Nomenclature - has been devised to repre- 
sent the landscape in all the member states. The No- 
menclature is hierarchical and contains three levels with 
forty-four classes at the lowest (most detailed) level -see 
list below: 
Corine Land Cover Class Nomenclature:- 
1.1.1 Continuous Urban Fabric 
1.1.2 Discontinuous Urban Fabric 
1.2.1 Industrial or Commercial Units 
1.2.2 Road and Rail Networks and Associated Land 
1.2.3 Port Areas 
1.2.4 Airports 
1.3.1 Mineral Extraction Sites 
1.3.2 Dump Sites 
1.3.3 Construction Sites 
1.4.1 Green Urban Areas 
1.4.2 Sport and Leisure Facilities 
2.1.1 Non-Irrigated Arable Land 
2.1.2 Permanently-Irrigated Land 
2.1.3 Rice Fields 
2.2.1 Vineyards 
2.2.2 Fruit Trees and Berry Plantations 
2.2.3 Olive Groves 
2.3.1 Pastures 
2.4.1Annual Crops Associated with Permanent Crops 
2.4.2 Complex Cultivation Patterns 
2.4.3 Land Principally Occupied by Agriculture 
with Significant Areas of Natural Vegetation 
2.4.4 Agro-Forestry Areas 
3.1.1 Broad-Leaved Forest 
3.1.2 Coniferous Forest