International Journal of Applied Earth Observation and Geoinformation
The MURBANDY Project: development of land use and network databases for the Brussels area (Belgium) using remote sensing and aerial photography
Introduction
The urbanisation process in Europe is still going on, but above all, the diffusion of the urban phenomenon is becoming more and more extensive. Very often, the city’s growth is studied indirectly, in terms of population change: net increase of population, daily or permanent migrations … (Hall, 1981, Cheshire, 1995, Le Jeannic, 1997, Vandermotten, 1983, Vandermotten et al., 1999). This is due to the availability of very accurate socio-economic data from the 10-year censuses or annual national registers. The real geographical dimension of the growth, which means the 'physical’ expansion of the city, is forgotten. In the developing countries, these socio-economic data are non-existent or unusable and the spatial study using remote sensing is the only way to follow the very rapid urban growth (Edmonds and Kyle, 1998, IAURIF, 1989, IAURIF, 1990, Turkstra, 1996, Virgo and Subba, 1994, Yeh and Li, 1997, Yeh and Li, 2001). However, although the problem is not really comparable and worrying as in the developing countries, there is in Europe an important need to measure, analyse, understand and forecast the urban expansion of the cities. Very few studies fulfil this need and often use raster data (Dagorne and Canavese, 1988, Himiyama, 1999, IFHPCRU, 1972, Lenco, 1995, Milanova et al., 1999). Moreover, it is also necessary to monitor the pressure of the urbanised areas on the surrounding agricultural and semi-natural zones. This topic is more than ever in the headlines at a time when politics are concerned with sustainable development and quality of life.
Section snippets
MURBANDY Project
The MURBANDY Project, initiated from the Space Applications Institute (SAI) of the European Joint Research Centre (JRC) in Ispra, aims to fulfil these objectives, that is measure, analyse, understand and forecast the urban expansion of some European cities. It can be compared to the Lacoast Project (Wolff et al., 1998), funded also by the SAI, which assessed the land cover changes in the European coastal zones since 1930 till 1995.
The project is divided into three modules. The CHANGE module
Why Brussels?
Brussels is a very remarkable city to study due to the important urban sprawl since World War II. Before the war, new houses filled up the existing neighbourhoods inside the actual administrative region, but after the war, new housing estates, large space consumers (single family free standing houses surrounded by gardens), were built up outside the city, in the suburbs, leading to a generally concentric structure. Fig. 1 reveals the cumulated percentages of space consumption in function of the
Study area
The area under study has been delineated using a morphological city’s definition (Vandermotten et al., 1999), slightly modified with the extent of the contiguous artificial surfaces of the CORINE land cover database (Jasselette et al., 1995), designed at 1:100,000. Then, according to a formula established for the MURBANDY Project, a buffer around the morphological city has been calculated with a width of (equal to 5.8 km for Brussels). The resulting study area covers 1365 km2, slightly
Scale and legend
The adopted working scale is 1:25,000, which is a good intermediate scale between a fine scale needed for the intensive urban land use of the city (1:5,000 or 1:10,000) and the necessity to cover a large study area to monitor the urban sprawl. The smallest mapping unit equals 1 ha inside the city and 3 ha outside the city. All the linear features are included in the network databases, but the elements larger than 25 m are also represented as polygons in the land use databases. The legend under use
Data
The reference image was a panchromatic IRS/1C image dating from 21 August 1997 with a spatial resolution of 5.8 m. In addition to these data, a complete set of orthophotos from the National Geographic Institute (NGI) dating since 1995 was used; they had a spatial resolution of 5 m, but their effective resolution was much better than the IRS. Furthermore, the NGI topographic database at 1:10,000 from the years 1994–1995 was helpful to interpret the land use (Wolff et al., 2000). Some additional
Method
The reference land use database was interpreted on screen according to the MURBANDY legend using an approximate scale of 1:10,000 in a GIS software (ArcView or ArcInfo according to the team). The satellite imagery and the orthophotos were displayed alternately under the edited coverage: the first one for an up-to-date image and the second for a precise delineation. A numerical copy of the survey maps was also available for the interpreter. For the Brussels area and prior to the interpretation,
Results
The land use maps for the years 1955 and 1997 are presented hereafter (Fig. 3, Fig. 4). The comparison between these two land use maps reveals a strong increase in built up areas, inside the administrative region of Brussels, but above all outside the city. The built up continuum now overlaps the region. In the past, these built up areas did not extent beyond the Soignes forest, which is not the case today. The most important part of the Flemish and Walloon regions in 1955 was still devoted to
Analyses
The set of produced databases allowed us to perform various analyses.
Annual changes of surface land use by period are revealed in Fig. 7. The rhythm of change for all classes has slowed down considerably since the 1950s: practically 2,500 ha per year between 1955 and 1970, less than 1,000 ha per year between 1985 and 1997. Fig. 8 reveals land use structures and changes between 1955 and 1997. Agricultural zones are increasingly urbanised—industrial, commercial, infrastructure and administrative
Conclusions and perspectives
The first results of the MURBANDY Project allow to assess the urban expansion of Brussels, not only in terms of population growth but by measuring the spatial increase in residential areas. We can postulate for a strong relation between these two variables, unless we highlight spatial variations in space consumption practices: a very open settlement will generate a lower population density per residential hectare than a grouped settlement. Such space consumption practices have to be put in
Acknowledgements
This research was undertaken by the IGEAT-ULB2 (Institut de Gestion de l’Environnement et d’Aménagement du Territoire—Université Libre de Bruxelles—Belgium) in collaboration with the NGI-B3 (National Geographic Institute, Belgium), and more specifically Eric Bayers and Jacques Peeters. Both teams would like to thank the MURBANDY team of the Space Applications Institute, Joint Research Centre in Ispra, for its financial support. The authors also thank
References (24)
- et al.
Land use in Macau: changes between 1972 and 1994
Land Use Policy
(1998) Historical information bases for land use planning in Japan
Land Use Policy
(1999)- et al.
Land use/cover change in Russia: mapping and GIS
Land Use Policy
(1999) Etude géographique des modes de faire-valoir en Belgique
Bulletin de la Société Royale Belge de Géographie
(1956)A new phase of urban development in Western Europe? The evidence for the 1980s
Urban Studies
(1995)- et al.
Loose-coupling a cellular automaton model and GIS: long-term urban growth prediction for San Francisco and Washington/Baltimore
Int. J. Geogr. Inf. Sci.
(1998) - et al.
Télédétection aéroportée et cartographie diachronique en zone urbaine. Le cas de Nice-Ouest (France)
Photo-interprétation
(1988) - European Community, 1993. CORINE Land cover-Guide technique. Office des Publications Officielles des Communautés...
Urban change in Europe. Lund Studies in Geography
Hum. Geogr. Ser. B
(1981)- Institut Bruxellois de Gestion de l’Environnement (IBGE), Institut de Gestion de l’Environnement et d’Aménagement du...
Cited by (7)
The urban environmental indicator "biotope Area Ratio" - An enhanced approach to assess and manage the urban ecosystem services using high resolution remote-sensing
2012, Ecological IndicatorsCitation Excerpt :A well-established approach in environmental management is to use indicators as standardized tools which provide comparable and comprehensible information for a specific planning or management goal. Indicators hence depend on their specific application aims and scales, such as for measuring sprawl or monitoring small-scale urban development (Fricke and Wolff, 2002; Herold et al., 2002; Jaeger et al., 2009; Kasanko et al., 2006; Schwarz, 2010; Yang and Liu, 2005), for identifying large-scale landscape functions in semi-urban areas (Walz, 2008) or for detecting fine-scale land cover heterogeneity (Cadenasso et al., 2007). One of the successfully and widely applied indicators in urban planning is the sealing degree describing the quantitative proportion of built-up or non-built-up sealed urban surface to the total surface.
Urban sprawl at the expense of cultivated land: decadal land use and land cover changes and future projections in the upper Awash basin of central Ethiopia
2023, Frontiers in Ecology and EvolutionUrban Climate Science
2018, Climate Change and Cities: Second Assessment Report of the Urban Climate Change Research NetworkEffects of historical urbanization in the brussels capital region on surface air temperature time series: A model study
2009, Journal of Applied Meteorology and ClimatologyOccurrence of Ips typographus (Col., Scolytidae) along an urbanization gradient in Brussels, Belgium
2005, Agricultural and Forest Entomology