A continent-scale data-driven approach to map Critical Raw Materials potential in Europe

The energy transition presents a crucial challenge to Europe, with the necessity of securing a sustainable supply of Critical Raw Materials (CRM) as specified in the European Union's Critical Raw Materials Act. Reaching the goal set by the Act of securing 10% of the EU's annual consumption through domestic extraction by 2030 requires the assessment of Europe’s domestic potential for CRMs. The collection of available data regarding the known CRMs potential throughout Europe is needed to perform this assessment. This data collection allows to perform mineral potential mapping to highlight areas with potential for the discovery of new CRM deposits.

The EU-funded GSEU – Geological Service for Europe project, coordinated by EuroGeoSurveys, an international organization that brings together Europeans geological survey organizations, aims at providing harmonized pan-European geoscientific data and expertise to support policy and decision making. As part of this project, mineral prospectivity mapping methods are applied to outline areas with the highest likelihood to host potential mineralization. They allowed the production of pan-European prospectivity maps for a selection of CRM (Co, Cu, Li, Ni, Mg, Mn, Nb, Ni, Sb, Ta, V, W). Favorability maps highlight promising areas for mineral exploration, improving exploration benefit/costs ratio, reducing its environmental footprint and enabling informed decisions about land use, environmental protection, and resource management strategies. They provide crucial information to both industry stakeholders and policymakers.

These maps are produced using the “Disc-Based Association” (DBA) method in combination with a Random Forest supervised classification. This predominantly data-driven approach leverages spatial analysis and machine learning techniques to delineate prospective zones for mineral exploration, specifically targeting CRMs. The DBA method analyses neighboring associations of cartographic features over the studied area, producing a unique matrix presenting the multivariate features identified around each sample point. The Random Forest classification allows scoring of each sample points through a binary classification. The first class consist of sample points in the vicinity of known mineralization, accessed through the harmonized dataset of CRM deposits provided by the GSEU Raw Materials team, while the second class are all the other sample points. The classification process results in each point being given a score, displaying the favorability of an area for mineral exploration. The result of this classification allows the definition of favorable areas for mineral exploration throughout Europe.

In this presentation, we describe the methodology used to produce the favorability maps for CRMs in Europe using the data compiled by the GSEU Raw Materials team. We present some of the resulting favorability maps and discuss future developments and application of this methodology.