DLR Logo
 
 
 
 
 
 

Brief Description of the Projects CORINE Land Cover 2000 and CORINE Land Cover 2006

Project objectives

In the European Union's CORINE Land Cover 2000 Program (CLC2000), land use and land cover information was to be brought up-to-date for all of Europe with the help of satellite data. This activity was based on a uniform methodology and a classification key which has been harmonized since the first survey was undertaken in 1990. The goal was to create a common database which permits comparisons of land cover and land use as well as identification of any changes occurring in the time interval between 1990 and 2000, and made possible ecological appraisals with their consequences for regional planning both at European and national level.

A further update of the data base CORINE Land Cover was performed between 2007 and 2010. The reference year was 2006. One main aspect of CLC2006 in the European context was the derivation of changes between 2000 and 2006.

Methodology

Within the framework of the European CLC2000 project the data base of the first survey 1990 was updated for all of Europe using the year 2000 as a base year, and changes with respect to CLC1990 were mapped.

Identification of land use changes was accomplished by visual interpretation supplemented by automated processes in a GIS supported system. The classification for Germany includes 37 land cover classes in three hierarchical levels. The main categories are developed areas, agricultural land, forests and natural landscapes, wetlands, and water bodies. The recording scale is 1:100,000, whereby newly identified areas have to be included if over 25 ha in size, and changes in land use if the new borders alter the area by at least 5 ha. Areas of linear form such as rivers are to be included if wider than 100 m.

The datasets for the year 2000 were uniformly orthorectified Landsat-7 data from 1999-2001 for all of Europe (responsible institution: JRC, in the Image2000 project). At DFD, the CLC1990 vector database and the satellite data from the 1990 activities were adjusted to these orthorectified data. The significant result of CLC2000 were two data sets: an up-to-date mapping of land cover, CLC2000, and mapping of changes with respect to CLC1990. In addition, any corrections found to be needed to the 1990 data set as a result of the interpretation process were published within a CLC1990 revised data set.

The update of the CLC dataset regarding the reference year 2006 was done using a database of SPOT-4, SPOT-5 and IRS-P6 LISS-3 data, which have been collected on behalf of ESA under the name IMAGE2006. Besides a summer coverage (data between 2005 and 2007), an additional coverage, a spring coverage, was built-up, which was to enable an improved separation of arable land and pasture land in agriculture. The ortho-rectification of the IMAGE2006 data was done at the Remote Sensing Technology Institute (IMF), located at DLR in Oberpfaffenhofen.

The update CORINE Land Cover 2006 was done mainly due to the same specifications as for CLC2000. However, the change layer CLC2006_Change was in the foreground. In CLC2006_Change, also changes with the minimum of 5 ha were recorded which have no effects on an extension of present areas more than 25 ha in size, or on new areas fulfilling the minimum mapping unit of 25 ha.

Project Structure

On the European level, the European Environment Agency (EEA) was responsible for the management and the coordination with the member states, assisted by the European Topic Centre for Terrestrial Environment (ETC-TE). A technical team of ETC-TE was built for the technical assistance, the training, the data integration and validation. The Joint Research Centre (JRC) of the European Commission was responsible for the subproject Image2000. The Image2000 team performed the selection, acquisition and ortho-rectification of the satellite data (Landsat-7 ETM+) assisted by the national project team.

At national level, project teams were built which performed the interpretation of satellite data, the mapping of land cover and land cover changes, the quality assurance and the validation. Responsible for the management and the performance of the German subproject was the German Remote Sensing Data Center of DLR in Oberpfaffenhofen, on behalf of the Federal Environmental Agency as national responsible institution. By a tender process, the companies EFTAS and GAF AG as well as a consortium of Infoterra GmbH, Hugin GmbH and Delphi IMM were integrated in the German CLC2000 project.

The German CLC2000 project was funded in the frame of the UBA F+E project under the number FKZ 201 12 209 with a share of funding by the European Union.

The update CORINE Land Cover 2006 for Germany was funded in the frame of the UBA F+E projects under the number FKZ 3707 12 200 and FKZ 3708 12 200.



Brief Description of the Project CORINE Backdating - "Backdating" of DLM-DE of the Reference Year 2009 to the Year 2006

Project objectives

For the update 2012 of land cover according to CORINE Land Cover classes, a specific national approach has been chosen in Germany. This approach uses the more accurate geometry of the Official Topographical Cartographic Information System ATKIS of the land survey authorities. The "Digital Land Cover Model" DLM-DE, derived from ATKIS, is to be applied as base for the deduction of CLC classes. In the responsibility of the German Federal Agency for Cartography and Geodesy (BKG), the DLM-DE 2009 database was built-up for the reference year 2009 with a minimum mapping unit (MMU) of 1 ha. Currently, an updated database DLM-DE 2012 is established by BKG. Using methods of generalization, this database will be transferred into the CLC2012 database (CORINE Land Cover 2012, having 25 ha MMU due to the EU specifications). The adjustment of the update is conducted using satellite data of 2012.

Besides CLC2012, a change layer of land cover between 2006 and 2012 is needed which describes the development of land cover categories (with 5 ha MMU). The update of DLM-DE 2009 to the year 2012 grants only directly the deduction of changes between 2009 and 2012. A comparison of changes of the conventionally derived CLC2006 (with 25 ha MMU) with DLM-DE 2009 or DLM-DE 2012 would mainly yield differences because of different geometries. Therefore it was necessary to derive spatial information on the changes regarding the more detailed geometry of DLM-DE 2009.

The modelling of changes between 2006 and 2009 in the detailed geometry of DLM-DE 2009 was task of DLR`s German Remote Sensing Data Center (DFD). For that purpose, a database "CLC2006_backdating" had to be created by a "backward look" from DLM-DE 2009 to the situation of land cover / land use in 2006.

Methodology

The backdating was done using an approach including the 46 most common change classes (based on the changes between 2000 and 2006). As far as possible, automatic or semi-automatic methods were used.

Concerning the expected large number of changes within agricultural areas, it was very helpful to use multi-seasonal satellite data of the AWiFS time series from 2005 / 2006. The multi-seasonal data delivers information about the different phenological behavior of arable fields in comparison to grasslands. Temporal statistic values acquired from the NDVI vegetation index (from AWiFS) were used for a separation of arable land against grassland. In addition, high resolution geo-information products created within the GMES / Copernicus program were used, e. g. two variants of the soil sealing layer 2006. Furthermore, interim layers derived from high resolution thematic raster products were generated for the status 2006, e. g. a water mask and several forest layers.

For the derivation of change areas concerning the main change classes, a combination of several approaches was applied: Particularly the calculation of zonal statistics; segmentations of high resolution LISS-III data (summer and spring) of the IMAGE2006 database and Landsat summer data, using the geometry of the polygons of DLM-DE 2009, and a combination of geometric segmentation information and thematic raster information of the status 2006. Final steps of interactive checks and manual corrections significantly improved the automatically created change database. An area-wide product of the status of 2006, called CLC2006_backdating, was delivered to BKG. The product comprises the polygons of modelled changes larger than 1 ha and a combined data base.

Project Structure

The project "Updating the Land Use and Land Cover Database CLC for the Year 2012 - "Backdating" of DLM-DE of the Reference Year 2009 to the Year 2006" was funded under the registration number UBA FKZ 3712 12 282 by the Federal Environment Agency. It was performed in a cooperation with BKG, UBA and the European Environment Agency (EEA) in Copenhagen.



Figure: Exemplary results of CLC2006_Backdating for the South of Leipzig region and the new Leipzig lakelands: Development of numerous water bodies and recreation areas as a result of re-cultivation and re-naturation of former lignite open-pit mines. Upper left: CLC2006_Backdating dataset, color codes see CLC legend; upper right: LISS-III summer scene with overlay of change areas and the codes for the status 2006; lower left: changes between 2006 and 2009 in the status of 2006; lower right: changes between 2006 and 2009 in the status of 2009 (color codes see CLC legend as well). Source satellite data: Copyright ANTRIX, Euromap / GAF AG (LISS-III)

 

 

 

Copyright © 2007 Deutsches Zentrum für Luft- und Raumfahrt (DLR). Alle Rechte vorbehalten.
Copyright © 2007 Deutsches Zentrum für Luft- und Raumfahrt (DLR). Alle Rechte vorbehalten.