Environment of Ghannouch in the south-east of Tunisia is characterized by the wide-spread hypersaline soils, typically colonized by halophytes. The study of their distribution is required in order to reveal the extent of salinization and its dynamic. Mapping and monitoring with a remote sensing approach are foreseen as the ways to trace the spatial and temporal dimensions of the phenomenon. The identification of halophyte vegetation can take advantage by analyzing optical remote sensing data. Here, we propose using a decision tree approach applied to European Space Agency Sentinel-2 imagery, for an accurate land cover mapping of Ghannouch district in Gab`es governorate. Data pre-processing was carried out using the European Space Agency's Sentinel Application Platform and the SEN2COR toolboxes. The mapping approach combines the spectral information in several channels of the visible-near-infrared spectrum. The land cover identification was performed following a spectral classification approach, exploiting several optical indices, normalized difference water index, normalized difference vegetation index, and several soil salinity index, in order to elaborate a decision tree algorithm. As a result, for an area of interest of 50?inspace??inspace50?km2, at least 68% was classified as halophyte land cover. This mapping exercise represents an important step toward improved halophytes mapping in Tunisia and could be used to monitor the status of other salinity prone regions in the world.
Decision tree for mapping of halophyte cover around Ghannouch, Tunisia
Sarti;Maurizio;Ciolfi;Marco;Lauteri;Marco;
2018
Abstract
Environment of Ghannouch in the south-east of Tunisia is characterized by the wide-spread hypersaline soils, typically colonized by halophytes. The study of their distribution is required in order to reveal the extent of salinization and its dynamic. Mapping and monitoring with a remote sensing approach are foreseen as the ways to trace the spatial and temporal dimensions of the phenomenon. The identification of halophyte vegetation can take advantage by analyzing optical remote sensing data. Here, we propose using a decision tree approach applied to European Space Agency Sentinel-2 imagery, for an accurate land cover mapping of Ghannouch district in Gab`es governorate. Data pre-processing was carried out using the European Space Agency's Sentinel Application Platform and the SEN2COR toolboxes. The mapping approach combines the spectral information in several channels of the visible-near-infrared spectrum. The land cover identification was performed following a spectral classification approach, exploiting several optical indices, normalized difference water index, normalized difference vegetation index, and several soil salinity index, in order to elaborate a decision tree algorithm. As a result, for an area of interest of 50?inspace??inspace50?km2, at least 68% was classified as halophyte land cover. This mapping exercise represents an important step toward improved halophytes mapping in Tunisia and could be used to monitor the status of other salinity prone regions in the world.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.