VERIFY: Out of sample testing for early warning systems using past climate

Start date: 1 April 2025
End date: 31 March 2030
Funder: Advanced Research + Invention Agency (ARIA)
Value: £5 million
Primary investigator: Dr Ruza Ivanovic
External primary investigator: David Thornalley, University College London
Co-investigators: Dr Lauren Gregoire
External co-investigators: Ruth Amey, University of Leeds (Project Manager), Ryan Patmore, Joel Hirschi, National Oceanography Centre | Chris Brierley, Erica Thompson, Serge Guillas, Richard Chandler, UCL | James Rae, Andrea Burke, Graeme MacGilchrist, University of St Andrews | Paola Moffa-Sanchez, Durham University | Liz Thomas, British Antarctic Survey | Xavier Fettweis, University of Liege | Mohamed Iskandarani, University of Miami

This team brings together experts in past climates (modelling, analytical geochemistry and sedimentology), oceanography, atmospheric science, glaciology, artificial intelligence, machine learning and modelling for decision making. Together, they will build Digital Twins – computer simulations trained or merged with real-world data – of past Greenland ice sheet and North Atlantic subpolar gyre tipping to evaluate the performance of new technologies developed by the programme.

Through this approach, the team will hone observational systems, physical process-based models and early warning detection algorithms against real-world realisations of tipping behaviour to build a robust and trusted early warning system.