Geological and Mineralogical mapping in Moroccan central Jebilet using multispectral and hyperspectral satellite data and Machine Learning

Abstract

The Central Jebilet Massif is one of the main Palaeozoic outcrops in Morocco. This massif is characterized by its arid climate, its significant mining potential and the absence of plant cover, which favors the use of spatial remote sensing for geological mapping and mineral prospecting in this site. The objective of this study is the comparison of hyperspectral data from the Hyperion sensor of the Earth Observing-1 (EO-1) satellite and multispectral data from the Operational Land Imager (OLI) sensor of Landsat 8 in the discrimination of geological units and detection of iron caps in the study area. The classification by the Support Vector Machine (SVM) method allowed for a good mapping of the lithological units in the study area. The accuracy of the SVM classification of hyperspectral data is higher than that of multispectral data, which was demonstrated by the confusion matrix, notably an overall accuracy of 93.05% and 89.24%, respectively, and a kappa coefficient of 91.25% and 84.36%, respectively. Concerning the iron detection, the band rationing using both sensors have demonstrated a performance of detecting areas that contain more iron ores, especially, the iron caps of Kettara mine, with a small advantage of hyperspectral data. In overall, our results highlight the efficiency of machine learning classifier and hyperspectral data for the detection of iron ores and the discrimination of lithological units in arid regions. The use of hyperspectral and multispectral images has been shown to be a good technique for the characterization of iron deposits and lithological units, which may help in in mineral exploration engineering with reduced fieldwork and geochemistry.

Formation Features and Prediction Criteria for Gold-Sulphide deposits (East Kazakhstan)

“…The late Carboniferous age of the complex is reliably determined by geological data, intrusions and dikes of which break through the deposits of the Bukon' and Maitobe suites (С2-3), and themselves they are cut by Kalba granites (Р1), and this was registered in satellite images. The age of the Kunush complex is confirmed by radiogeochronological data obtained by different methods: K/Ar (305 Myr), U/Pb (306,7 Мл), (299,2 Мл) [11,21,22]. Geochemical characteristics of Kunush complex granitoids are inclined to low-K calc alkaline series, which is also proved by low alkaline sum Na2O+K2O (6.06-6.38 mass % and 5.20-6.81 mass % respectively) and by high Na2O/К2О ratio (3.83-3.10 and 3.93-6.43).…”

Multispectral data mining: A focus on remote sensing satellite images

“…This method involves both temporal and spectral criteria. In practice, both physical and image-based methodologies are combined in AC (Chakouri et al, 2020), for example, for regions with inland water bodies and haze.…”

A Systematic Review on the Application of Machine Learning in Exploiting Mineralogical Data in Mining and Mineral Industry

“…ML and AI are actively used for mining complex, high-level, and nonlinear geospatial data and for extracting previously unknown patterns related to geological processes [45]. These techniques were applied in the identification of mineralization related geochemical anomalies in China [45,47,[54][55][56], as well as generating a prospectivity map for targeting gold mineralization in Canada [49,50] and China [51], for the detection of iron caps in Morocco [57], for creating a continuous mineral systems model for chromite deposits in Iran [58], and geological mapping studies using the characteristics of rocks [59,60]. The authors in [61,62] integrated multi-sensor remote sensing techniques such as drone-borne photography and hyperspectral imaging for processing with ML algorithms in order to generate the geological mapping.…”

“…Therefore, it becomes of utmost importance to predict, relatively accurately, regions with higher potentials for new deposits based on the large datasets of various types of measurements. The dataset can contain lithogeochemical [49,82], spatial [49,50], geochemical [45,55,81,83], geophysical [81], geological [81], concentration of indicator elements [47,51,52,54,56,65,68], hyperspectral [57,60,61], spatial proxies [58], total magnetic intensity [52], isostatic residual gravity [52] data. It is worth emphasizing that in most of such studies mineralogical analyses results are either used to generate the input features for the ML models or ground truth for training such models.…”