GEM-Cu

GEM-Cu is developing a revolutionary new approach to mineral exploration by creating a digital “Virtual Earth” capable of reconstructing the geological conditions that generated copper deposits throughout Earth’s history – including millions to billions of years in the past. 

The project combines advanced geodynamic modelling, plate tectonic reconstructions, machine learning, and decision science to identify, predict and inform where copper-favourable conditions have existed across space and time. 

Using world-leading modelling platforms including G-ADOPT and GPlates, researchers will reconstruct more than one billion years of Earth’s tectonic evolution. By combining these 4-dimensional digital Earth models with geological and geophysical datasets, the team can identify the deep Earth processes associated with known copper deposits and use that knowledge to predict where undiscovered resources may exist.

Rather than relying solely on historical discoveries, this process-based approach enables exploration teams to anticipate where copper is likely to occur before evidence appears at the surface. 

The result is a more efficient, targeted and sustainable exploration strategy that reduces uncertainty, lowers environmental impact and opens entirely new search spaces for future resource discovery. 

What exactly are we doing?

The project is organised into three interconnected research work packages (WP) that together create a new framework for predictive mineral exploration. 

• Reconstructing Earth’s geodynamic history 

Researchers are combining global plate tectonic reconstructions with advanced mantle convection simulations to identify and understand how copper-forming environments evolved over the last billion years. This work creates a detailed four-dimensional picture of Earth’s changing geology through space and time and forms WP1. 

• Building predictive exploration models 

Machine learning and statistical modelling techniques are used in WP2 to identify the geodynamic signatures associated with copper mineralisation. These insights are then transformed into predictive prospectivity models capable of highlighting regions where copper-favourable conditions likely existed. 

• Developing decision-support tools 

The project goes beyond scientific discovery by exploring how people interpret, trust, and use predictive information. As part of WP3, researchers are co-designing visualisation tools, training resources, and decision frameworks that help organisations confidently integrate predictive geoscience into exploration and planning processes. 

• Creating open-access resources 

The project will deliver an openly available global database linking copper deposits with their tectonic and geodynamic origins, providing a valuable resource for researchers, policymakers and industry. 

Together, these activities create a complete innovation pipeline that connects deep Earth science with real-world exploration decisions. 

• Reconstruct the tectonic and mantle processes responsible for copper mineralisation over the past billion years.  

• Develop a global database linking known copper deposits to their geodynamic origins.  

• Create predictive machine learning models capable of identifying copper-favourable regions before discovery.  

• Deliver decision-support tools that help organisations confidently apply predictive geoscience.  

• Reduce exploration risk, cost and environmental impact through more targeted resource discovery.  

• Improve understanding of how scientific predictions are interpreted, trusted and adopted by users.  

• Create a scalable framework that can be extended to other critical minerals, including nickel, cobalt and rare earth elements.  

Ultimately, GEM-Cu aims to transform mineral exploration from a reactive search for surface clues into a predictive science rooted in Earth’s deep history. By combining advanced modelling, artificial intelligence and human-centred design, the project seeks to accelerate the discovery of the resources needed for a cleaner, more sustainable future.