Laboratory-based hyperspectral imaging (HSI) is an optical non-destructive technology used to extract mineralogical information from bedrock drill cores. In the present study, drill core scanning in the long-wave infrared (LWIR; 8000–12,000 nm) wavelength region was used to map the dominant minerals in HSI pixels. Machine learning classification algorithms, including random forest (RF) and support vector machine, have previously been applied to the mineral characterization of drill core hyperspectral data. The objectives of this study are to expand semi-automated mineral mapping by investigating the mapping accuracy, generalization potential, and classification ability of cutting-edge methods, such as various ensemble machine learning algorithms and deep learning semantic segmentation. In the present study, the mapping of quartz, talc, chlorite, and mixtures thereof in HSI data was performed using the ENVINet5 algorithm, which is based on the U-net deep learning network and four decision tree ensemble algorithms, including RF, gradient-boosting decision tree (GBDT), light gradient-boosting machine (LightGBM), AdaBoost, and bagging. Prior to training the classification models, endmember selection was employed using the Sequential Maximum Angle Convex Cone endmember extraction method to prepare the samples used in the model training and evaluation of the classification results. The results show that the GBDT and LightGBM classifiers outperformed the other classification models with overall accuracies of 89.43% and 89.22%, respectively. The results of the other classifiers showed overall accuracies of 87.32%, 87.33%, 82.74%, and 78.32% for RF, bagging, ENVINet5, and AdaBoost, respectively. Therefore, the findings of this study confirm that the ensemble machine learning algorithms are efficient tools to analyze drill core HSI data and map dominant minerals. Moreover, the implementation of deep learning methods for mineral mapping from HSI drill core data should be further explored and adjusted.

It is widely accepted that non-energy minerals underpin modern economies since they are essential for manufacturing and renewable energy supply. Europe shows an inevitably growing and accelerated consumption of mineral commodities. The critical question is whether supply to meet these demands is adequate. However, no one can answer this with any certainty because secure supply of mineral RM is a matter of knowing the resources and the ability to exploit them sustainably.Europe’s strive to be become the world’s first climate-neutral continent by 2050 means implementing the “European Green Deal” by the EU Commission. Measures accompanied with an initial roadmap of key policies range from ambitiously cutting emissions, to investing in cutting-edge research and innovation, to preserving Europe’s natural environment.This green transition is a giant societal leap. However, of the clean and carbon-reducing technologies (e.g. wind turbines, photovoltaic panels, electric and hybrid vehicles), which allow energy production from renewable resources, use significant quantities of metals [e.g. REE, PGE, Nb, Li, Co, In, Ga, V, Te, Se] that are derived or refined from minerals, and of which Europe is strongly import dependent on. The high import dependence of strategic (SRM) and critical raw materials (CRM) has a serious impact on the sustainability of the EU manufacturing industry value chains and key enabling technologies (e.g. renewable energy industry, mobility sector and AI) and significant release of CO2 emissions due to foreign ore transport. Effectively knowing Europe’s subsurface and the potential mineral supplies that can be used in these manufacturing industries can achieve this. We need to calculate the volumes of CRM and SRM metals (e.g. Co, Nb, V, Sb, PGE and REE) and minerals currently not extracted in Europe. We further need to understand how high-tech elements are mobilised, where they occur and why some are associated with specific major industrial metals. This means a renewed and robust focus on advanced exploration for new mineral deposits on land and sea.FRAME(*)[1] addresses most of these concerns by focusing on at least four of the current objectives of the EU Commission: 1- CRM; 2- battery critical elements [graphite, Co, Li]; 3- The circular economy and; 4- the responsible sourcing of metals by combating conflict minerals.With focused work packages, FRAME aims to broadly deliver, 1- a new assessment of the SRM and CRM in Europe; 2- an innovative predictability of where the sourcing of some of these SRM and CRM may come from to reduce dependance on external supply sources, which in some specific metals such as Nb and Ta, fosters the sustainable and responsible supply and; 3- look at case specific sites for the reuse of mineral RM. Data will be made available through a structured data platform. Hence, FRAME is making a significant contribution to aid in the “European Green Deal”, activities such as the Battery Alliance and support legal actions like the new EU "conflict minerals" regulation effective from 1/01/2021. [1] (*) This project has received funding from the European Union's Horizon 2020 Research and Innovation Programme under Grant Agreement Nº 731166  

The H2020 GeoERAFRAME project (www.frame.lneg.pt) consists of a partnership of 11 European geological surveys. Geographical and geological information was collected including the genetic type of the different commodities. In the EU, data show that there are 1195 registered occurrences, prospects or deposits of Li, Co and graphite, of these only 17 are active. The data classify the occurrences according to their genetic type, occurrence type and production status. The data have been supplied from geological surveys national databases and in this compilation, we regard all Co deposits with a mean Co >100ppm as potential occurrences for Co. For the other commodities, Li bearing minerals or graphite must be identified or explored for to be included.Even if our compilation has shown that the different national resource databases contain data of variable quality, with a lot of shortcomings, inconsistences and errors, the overall quality is good enough to assess the EU potential.The Lithium deposits can be group into the following types: i) High grade Li deposits including Li-rich LCT pegmatites, rare metal granites and atypical stratiform deposits such as Jadar. The distribution of lithium in Europe shows a strong clustering highlighting the Li potential of the Variscan belt of south and central Europe. Representative examples are Sepeda pegmatites (103 000t Li2O – grade 1.0%) or Beauvoir rare-metal granite (325 260t Li2O – grade 0.78%). Medium-grade Li deposits are represented by hydrothermal deposits such as greisens and Li-bearing quartz veins associated to some peraluminous rare metal granites (Cinovec 2 715 010 Li2O – grade 0.42%).Cobalt is a common minor constituent in a number of different ore types. In Europe, most of the known Co-bearing deposits and showings are clustering in the Nordic countries (Finland, Sweden and Norway). In the Nordic countries, the deposits mostly represent magmatic Ni-Cu and Fe-Ti-V deposits and VMS deposits, whereas elsewhere in Europe genetic types are more varied from sediment-hosted, to lateritic and 5-element vein types, among others. The only active mines producing cobalt are located in Finland. Kevitsa mine in northern Finland is a large low-grade Ni-Cu-PGE deposit, which produced 591 t of Co in 2018. Kylylahti mine is a small-sized Outokumpu-type Cu-Zn-Ni-Co deposit, which produced 278 t of Co in 2018. Terrafame is a large, low-grade black-shale hosted Zn-Ni-Cu-Co mine that produces Co as by product to Ni and Zn.Graphite is a common mineral in rocks throughout Europe. However, find economically interesting deposits are rare. The bulk of the graphite occurrences occur in Archean or Proterozoic rocks of Fennoscandia and Ukraine. In addition, a number of amorphous graphite occurrences are found in Phanerozoic rocks in Austria. There are also a large number of showings where the genetic type is unknown. Active mines are situated in Ukraine, Austria and Norway. The graphite bearing rocks are typically organic rich para-gneiss often associated with carbonates and iron formations. The graphite content varies from 2-3% up to over 40%. The Trælen deposit in Norway is the world’s richest graphite deposit in current production with an average ore grade of 31%.