PASTA. Pilotprojekt zur Anwendung der SAR-Technik fuer die Agrarstatistik und die Inventur der Landnutzung Schlussbericht

Satellite remote sensing allows the survey of agricultural practice for large areas and the inspection of land use for the purpose of the agricultural policy of the European Community. The existing techniques are based on optical remote sensing data, which in Europe are often disturbed by clouds. SAR-data of the ERS-satellites however are operationally available without these limitations caused by clouds or sun illumination. They provide additional information related with the physical properties of the illuminated objects. It was the main goal to analyse how ERS-SAR data in combination with optical data and official GIS basis data could be used for a common landuse classification and which improvements could be achieved. Several different classification methods were used for a multisensoral and multitemporal approach. The main cultivated crops could be identified with a higher accuracy of roughly 10% compared to pure optical classification. A further substantial improvement (up to 20%) could be achieved by the application of a majority filter on the classification results. The methods are suited for a fast acquisition of land use and for the control of the EC-policy of financial compensation related with acgricultural land use. It is however necessary to improve the classification accuracy if, as it is the case in Baden-Wuerttemberg, the agricultural fields are of very small size. Higer resolution satellite data and additional GIS information could lead to these improvements. (orig.)SIGLEAvailable from TIB Hannover: F97B1403 / FIZ - Fachinformationszzentrum Karlsruhe / TIB - Technische InformationsbibliothekBundesministerium fuer Bildung, Wissenschaft, Forschung und Technologie, Bonn (Germany); Deutsche Agentur fuer Raumfahrtangelegenheiten (DARA) GmbH, Bonn (Germany)DEGerman