Détection semi-automatique des cheminées doléritiques du bassin de Tindouf par imagerie satellite Sentinel 2. - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2019

Semi-automatic detection of doleritic pipes in the Tindouf basin by Sentinel 2 satellite imagery.

Détection semi-automatique des cheminées doléritiques du bassin de Tindouf par imagerie satellite Sentinel 2.

Résumé

Doleritic pipes are known in the Tindouf Basin in Algeria since the works of Jaqcuemont (1952) and Gevin (1960), and are assumed to belong to the Central Atlantic Magmatic Province (CAMP). The Tindouf Basin is characterized by an abundance of magmatic formations belonging to the CAMP (Chabou, 2008; Bersi, 2016) which outcrop mainly as sills and dykes. These magmatic formations represented mainly by dolerites, were emplaced at the Triassic-Jurassic boundary (201 Ma) (Chabou, 2008). The sills inject the devono-carboniferous formations and their total thickness can reach a hundred meters (Chabou,2008; Bersi, 2016). Doleritic chimneys outcrop in both flanks of the basin as sub-circular massifs that intersect the Paleozoic formations. On the northern flank, these manifestations are observed in the heart of the Zemoul, Oum El Ksi, Igma, Smeïra and Tinfouchy anticlines, whereas in the southern flank, the chimneys are located in the clay-limestone formations of the Viséen. On the other hand, they are very rich in sulphides and may represent the equivalent of PGE (platinum group elements) rich LIP'S differentiated rocks such as that of the Siberian magmatic province (Norils'k). The present work focuses on the mapping of these magmatic structures using remote sensing. In geological mapping, the treatments usually performed on satellite images give good results using PCA and band ratios (Bersi, 2016; Hachemaoui et al., 2018). However, these treatments seem to be ineffective in discriminating chimneys from the rest of the dolerite formations. In order to characterize doleritic chimneys by remote sensing, we used a semi-automatic detection technique with the Target Detection Wizard tool under ENVI software. This technique allows us to detect targets by correlating all image spectra with spectra obtained in the laboratory. To address the problem of the effects of patina and other disturbances (related to weathering) on the spectral responses of rocks, we opted for the creation of a spectral library specific to chimneys, by retrieving spectral signatures from pixels from known chimneys. The results obtained allowed us to map structures and areas with spectral signatures identical to those of the pipes in the entire Tindouf basin. However, the fact that some mapped areas coincide with the presence of known dolerite sills indicates that the spectral signature used likely corresponds to contact metamorphism that develops intensely around dolerite pipes, but also in the contact zone of the sills with the sedimentary host (Bersi, 2016; Chabou, 2008; Jacquemont, 1952; Gevin, 1960). Field investigations confirmed the latter result. Sampling of dolerites from chimneys allowed the use of this technique with new spectral signatures specific to these rocks, which allowed discrimination of facies corresponding to dolerite chimneys.
Fichier non déposé

Dates et versions

hal-03891382 , version 1 (10-12-2022)

Identifiants

  • HAL Id : hal-03891382 , version 1

Citer

Oubaida Hachemaoui, Charaf Chabou, Chrystèle Verati, Mohand Bersi. Détection semi-automatique des cheminées doléritiques du bassin de Tindouf par imagerie satellite Sentinel 2.. Mediterranean School on Modeling & Multi-scale Imagery in Earth Sciences (MOMIES), Faculty of Earth Sciences, Geography and Country Planning (FSTGAT/USTHB), Sep 2019, Alger, Algérie. ⟨hal-03891382⟩
12 Consultations
1 Téléchargements

Partager

Gmail Facebook X LinkedIn More