Author/Authors :
Abdallah, C National Council for Scientific Research - Remote Sensing Center, Lebanon , Bou Kheir, R National Council for Scientific Research - Remote Sensing Center, Lebanon , Khawlie, M National Council for Scientific Research - Remote Sensing Center, Lebanon , Faour, G National Council for Scientific Research - Remote Sensing Center, Lebqnon
Title Of Article :
Comparative analysis of drainage networks extracted from DEMS and conventional approches in Lebanon
Abstract :
Drainage networks are major elements of watershed assessment as they reflect its topographic, geologic, hydrogeologic and tectonic characteristics. The conventional identification ofdrainage networks from aerial photographs and topographic maps is effort and time consuming. Therefore, many algorithms have been developed in the last three decades for the extraction of drainage networks from satellite imageries (Spot, Landsat, Radar). But these digital delineations still prove to have some gaps and disagreements with those obtained by conventional methods. Accordingly, this paper aims to analyze the gap by a comparative analysis between drainage networks extracted from a digital elevation model (DEM) obtained from radar imageries and those delineated from topographic maps at scale 1:50,000. Accordingly, a coincidence value ofoverlap ofsame features (39%) was found between the two approaches. Adaptation of various thresholds in the algorithms of the digital pixels, taking into account the geological and geomorphologic properties of the basins, has improved the agreement between the automated and the manually delineated drainage network. An improvement of 7% was achieved in the elevated crests, 4% in the gentle slopes, 10% in the plateau and 26% in the plain, giving a total overlapping of51%.
NaturalLanguageKeyword :
drainage network , digital elevation model , morphology , automated extractions
JournalTitle :
Lebanese Science Journal