Consumption of natural land for settlements, industry and transport infrastructure proceeds with an alarming pace in Europe and disagrees with the stagnating population. Every three years we lose a surface area equivalent to the size of Luxemburg. Structural Funds of the EU have been held partially responsible for the ill-development. In order to obtain an objective and independent monitoring tool for this and many other purposes, the EU has now launched the “Urban Atlas”, a harmonised map of current land-use/land-cover and their changes over time.
{sidebar id=293 align=right} Outset and purpose
The Urban Atlas is the first large-scale geodataset ever produced operationally from satellite data. Production costs are in the order of a few Euro per km². The new dataset must have been awaited because the announcement on 16 January 2009 had caused several 100000 Google references only a month later.
There have been precursors in the research domain, most notably MURBANDY (Monitoring URBAN DYnamics) and MOLAND (MOnitoring LANd use/cover Dynamics) as well as many other methodical studies inside and outside Europe [2], but are inferior in scale (mostly 1:25000 instead of 1:10000) and patchy. Clearly the new endeavour has only become possible with the advent of satellite imagery in the metric resolution range. Already the precursors show that all 25 cities under study, albeit East and West Europe, had grown, 1/3 of them even doubled in size from the 1950’s – 1990’s, mainly at the expense of natural space [Fig.1].
By the time when the operational “Fast-Track” Services of GMES (Global Monitoring for Environment and Security and since re-named “KOPERNIKUS”) were launched [3], problems had accrued in Europe. Not only that the Gothenburg Summit had adopted the European Sustainability Strategy requesting a GMES by 2008 [4], but a number of policies had been set into force fighting environment degradation or hazards, or simply governing a uniform geo-information, most notably:
{sidebar id=294 align=right}
{sidebar id=295}
Moreover, an “Urban Audit” had been launched by the Commissioner in charge of Regional Development in 2001 (= first reference year), edited 2004 and covering 258 cities.
The first update was published 2007 with 367 (reference year 2004) and further updates are planned with a three-annual periodicity targeting at over 500 cities. It is a database giving a useful pan-European statistical overview by a set of over 300 indicators across the domains Demography, Social Aspects, Economic Aspects, Civic Involvement, Training and Education, Environment, Travel and Transport, Information Society, Culture and Recreation [5]. However it lacks an adequate geographic component to cope with the policies listed above.
The Urban Atlas as part of the GMES Land Monitoring Core Service (LMCS)
The only source of land use or land cover information that is EU wide is CORINE-Landcover (CLC) which was primarily designed to measure natural areas and agricultural land. Due to the coarse resolution and high density thresholds for urban classes, CLC fails to capture urban areas in sufficient thematic and geographic detail. Nevertheless CLC is widely established and used and any data model of the Urban Atlas must, in a higher level of aggregation, be compatible.
Old CLC will as well be improved and has therefore become part of the LMCS. Artificial coarsening to 25ha minimum mapping units was no longer timely, led to incorrect statistics due to overestimation of dominant / underestimation of inferior classes (Fig. 2) and hence got criticised by the users. Other criticism addressed the update period which should be shortened from 10 towards 3 years reflecting the increased speed of land-use changes.
Both, CLC and Urban Atlas are components of the LMCS which have been defined in a geographically and thematically well represented user workshop end 2005. Approximately 100 participants from 22 countries and relevant European organisations wished to get “regularly updated core land cover / land use data for information requirements at European, national, regional and local level” [6] in the form of a “Continental Component” (= Europe) of medium scale (= improved CLC) and a “Local Component” (selected “hot spots”) of large scale (Fig. 3).
{sidebar id=296 align=left} A “Global Component” was to follow later. For all components they wished to access both, the satellite images and interpreted data in the form of land-use/land-cover classes, preferably automatically derived and compatible with existing systems.
For the Local Component the Directorate General for Regional Policy of the European Commission provided operational funds to cover 305 LUZ with an Urban Atlas as a GMES “Fast-Track” Service to be rolled out together with the next edition of the Urban Audit [7]. In order to include monitoring of the city-hinterland relationships, where growth normally occurs along the communication highways, areas to be observed have been extended to “Larger Urban Zones” (Fig. 4).
For explanation: the average LUZ surface of the 305 agglomerations covered is 1978 km² whereas the administrative boundaries delineate an acreage of only 229 km² average size. The overall surface of these LUZ corresponds to 15% of the EU-27, but encompasses almost 50% of the population.
Technical requirements as to the level of detail have of course been risen with regard to the precursor projects, reflecting the progress in satellite image resolution [8]:
{sidebar id=297 align=right}
With these specifications production prices of under 10 Euro per sq. km. have been calculated including the cost of data. This would imply automated or at least semi-automated information extraction. Main input data should be very high-resolution satellite images, i.e. 2.5m pan-sharpened colour images from Spot-5 in the first exercise; for later repetitions sub-meter images have been envisaged.
They will be blended with commercial street navigation vector maps to enhance geometry and attributing, as well as topographic maps of 1:25000 scale whose provision is coordinated by Eurogeographics. In any case of conflict between these data the (presumably more up-to-date) satellite information should prevail. 2006 was fixed as the reference year for all imagery for the whole of Europe.
A list of the first 185 cities to be mapped in the first year [7]. But beyond them there might be maps of more cities available from preparatory projects. All LMCS services and products are available from [9] and readers are invited to test and feed back.
{sidebar id=298 align=right}
Applications
Before a call for tender was issued for the whole exercise, a proof-of-concept test was performed taking Bremen as a medium-sized north-western European agglomeration and Badajoz as a smaller southern European city for test beds.
Results were validated against ground control points, about 900 for Bremen (5920 km²). During these tests, in which local professionals with good knowledge of the area as well as specialised value-adding companies from all over Europe took part, the mapping guide [8] was iteratively improved to finally reach a good confidence level.
Meetings were also held with experts from different fields of local management of different cities to exchange ideas about use cases for the new dataset, not only on the European but also on the local and regional level.
The value for the European level is clear: The Urban Atlas will provide an neutral and independent tool to monitor effects of structural measures, be they heading in the good or the bad direction.
They can involve both Ex-ante assessments (e.g. Environmental Impact Analyses) in areas where most of structural funds are destined and where impacts include a visible physical footprint; Impact monitoring including support indicator systems for measuring and reporting on progress of structural funding. If sufficient topicality is assured, political decisions can be examined and eventually amended earlier in time.
{sidebar id=289 align-right}For the cities themselves, in addition to easing their mapping and reporting obligations under the European policies, some innovative “downstream services” appear feasible when local information is added on top of the Urban Atlas, such as:
{sidebar id=299 align=right}
A possible measure to counteract urban sprawl could consist in identifying and quantifying abandoned inner-urban land and support legislation that renders re-conversion of these surfaces more attractive.
Regular overpass of satellites makes it possible to single out parcels in repetitive images, on which absolutely no changes occur over time. In some literature these areas are designated as “brownfields”, but this implies usually contamination which is normally not visible from above. In the nomenclature of the European Urban Atlas, all presumably abandoned areas are comprised in class 1,3 and 4 as “Land without current use” (Fig. 8).
{sidebar id=300 align=right} Another interesting example are areas with exposure to risks where, especially in cross-border areas, traditional maps fail to represent the area in sufficiently harmonised detail and up-to-dateness.
That is where the unique potential of satellite data can fully unfold and show citizens of such regions a common European identity (Fig. 9). Calls from within the European Regional Development Fund (ERDF) will further accompany GMES by including applications into relevant calls, for example [11].
Outlook
In a recent communication [12] the EU Commission outlines the future organisation structure of GMES once it becomes fully operational with the launch of the next budgetary term 2014 (transitory budget requested from 2011).
{sidebar id=301 align=left} For the observation infrastructure in space the recent ministerial conference has set the course: under German (37%) and French (18%) leadership a “GMES Segment-2” budget of 822 million Euro was released. Land will be monitored by a fleet of “Sentinel-2” high-resolution optical satellites complemented by radar observations and very high resolution images from satellites from member states.
The “Local Component” could perhaps evolve from regular updates of the Urban Atlas towards inclusion of other “hot spots” outside urban areas like transport hubs, construction or mining sites, waste dumps, coastal areas, specific agricultural or forest activities, protected sites under NATURA2000 and other places of high sensitivity or rapid change to be mapped with the same method.
New satellites on the horizon such as hyperspectral imagers, offering some 250 channels rather than below 10 of presently used sensors [13], might allow the distinction of far more land-use/land-cover classes at reasonable cost. Improving census data by capturing night-time light emissions [14] might equally become operational.
{sidebar id=302 align=right}
Another promising field could be to put air quality measurement on a new fundament. (Fig.10) shows particulate matter measurements from satellites that were hybridised with administrative boundaries to provide a neutral evidence basis for communes to decide on emission restrictions when threshold levels are exceeded.
Furthermore, the third dimension could be included in the mapping. This could range from tracing gentle up and down motions of urban soils due to groundwater extraction, mining, subway construction or tectonic processes, in centimetre accuracy with radar satellite interferometry [15], up to documenting newly built storeys with modern elevation modelling [16].
In conclusion, urban monitoring is a promising field when it comes to mastering our future living space and the Urban Atlas has opened a wide gate into this important domain.
————————————————————————
Wolfgang Steinborn is with the German Aerospace Centre in charge of Earth Observation Networking. He has in the last four years worked for the European Commission in Brussels where he was responsible for the build-up of the GMES Land Monitoring Service.
Address: Koenigswinterer Str. 522-524, 53227 Bonn, Germany
50°43’5”N, 7°9’29”E, 65mH
E-Mail: [email protected]
Phone: +49-228-447-356
Acknowledgement
The work of Lewis Dijkstra from the European Commission, the great driving motor behind the Urban Atlas, is gratefully acknowledged.
References
1] European Environment Agency: “Urban sprawl in Europe – The ignored challenge”, EEA Report No 10/2006, http://www.eea.europa.eu/publications/eea_report_2006_10
2] European Environment Agency (Ed.): “Towards an urban atlas – assessment of spatial data on 25 European cities and urban areas”. Environmental Issue Report No. 30 (2002), ISBN 92-9167-470-2, available online from http://www.eea.eu.int/, and references therein
3] COM(2005)565: „Global Monitoring for Environment and Security (GMES): From Concept to Reality”,
4] COM(2001)264: ”A Sustainable Europe for a Better World: A European Union Strategy for Sustainable Development”
5] http://www.urbanaudit.org/
6] W.Steinborn: “The GMES Land Service – a Building Block in the Common European House”, GEO:connexion, M.Ward publishers, Cambridge, April 2009
7] Europa Press Releases: “Urban Atlas: Europe’s eye in the sky provides cities with tools for smart development”, Brussels, 16 January 2009 http://europa.eu/rapid/pressReleasesAction.do?reference=IP/09/65&format=HTML
8] S.Meirich: Mapping Guide for a European Urban Atlas”, 2008 http://ec.europa.eu/enterprise/calls/files/08_029/guide.pdf
9] LMCS internet portal: http://www.land.eu/ (in operation since the GMES-Conference in Lille on 16/09/2008)
10] C.Sprecher, M.Fritsch, M.Huhmann: “Effizientere Nutzung von Industrie- und Gewerbezonen – Die mögliche Rolle von interkommunalen Landumlegungen”, Géomatique Suisse, 9/2007, p.454-458
11] INTERREG-IVc programme “Regions for economic change” has GMES as one of its priorities in “Theme 1.9: Improving monitoring of Environment and security by and for the regions”, http://ec.europa.eu/regional_policy/cooperation/interregional/ecochange/doc/theme19.pdf
12] COM(2008)748: „Global Monitoring for Environment and Security (GMES): we care for a safer planet”
13] http://www.enmap.org/
14] David J. Briggs, John Gulliver, Daniela Fecht, Danielle M. Vienneau: “Dasymetric modelling of small-area population distribution using land cover and light emissions data”, Remote Sensing of Environment 108 (2007), p.451–466
15] see the GMES “TERRAFIRMA”-Project for reference: http://www.terrafirma.eu.com/
16] http://www.infoterra.de/terrasar-x/tandem-x-mission.html
