Multispectral and multitemporal satellite data fusion for coastal zone dynamics assessment

摘要

This study focuses on the assessment of coastal zone land cover changes based on fusion technique of satellite remote sensing data. The evaluation of coastal zone landscapes is based upon different sub-functions which refer to landscape features such as water, soil, land-use, buildings, groundwater, biotope types. A newly proposed sub-pixel mapping algorithm was applied to a set of multispectral and multitemporal satellite data for Constantza and Black Sea coastal zone areas in Romania. A land cover classification and subsequent environmental quality analysis for change detection was done based on satellite images over a period of time from 1986 to 2006 , from Landsat TM , Landsat ETM, MODIS , IKONOS, Quickbird. Spectral signatures of different terrain features were used to separate and classify surface units of coastal zone and sub-coastal zone area.The change in the position of the coastline in Constantza area is examined in relation with the urban expansion. A distinction is made between landfill/sedimentation processes on the one hand and dredging/erosion processes on the other. A quasi-linear model was used to model the rate of shoreline change linked to anthropogenic and coastal erosion.