Multisensor image fusion for Romanian Black Sea coastal zone analysis

Abstract

This paper applied satellite data fusion technique for changes assessment and environmental impact over Romanian North-Western Black Sea and Danube deltaic coastal area. Since single sensor solutions for automatic object recognition provide only partial answers was proposed a multi-sensor, multi-resolution approach to be effective tool of exploiting the complimentary nature of different data types for change-detection studies. In order to extract environmental features for determining surface patches and surface boundaries, grouping surface patches based on spatial proximity, geometric and radiometric properties was performed object recognition for North-Western Black Sea coastal zone based on different satellite data (Landsat MSS, TM, ETM, ERS, SPOT XS, ASTER and MODIS). Preliminary results show significant coastline position changes of North-Western Black Sea during the period 1975-2004. The direct impacts are clearly shown, but it is less straightforward to link the changes in coastline to indirect impacts of the changing land use/cover. As an indication of land use/cover change, the extension of the road network and the urban areas were compared. The growth of coastal urban areas generates a range of threats to the nearby shoreline habitats. Direct physical damage is caused by construction works on harbors, airports and tourist resorts. Also the coastline change is examined and linked to the urban expansion in order to determine if the changes are mainly human induced or natural. A distinction is made between landfill/sedimentation processes on the one hand and dredging/erosion processes on the other.