Massimo Bollasina

PhD opportunity at the University of Edinburgh in Enhanced predictability of European weather extremes via machine learning . This project asks whether European heatwaves, floods, and other extreme weather events can be predicted further ahead than current forecasting systems allow. The research combines deep learning , atmospheric dynamics , and climate science to study ocean-atmosphere signals in the Pacific and Atlantic, including persistent patterns such as Rossby waves and blocking. The PhD will use reanalysis data , sub-seasonal-to-seasonal (S2S) forecasts , and climate-model simulations to identify early warning signals of European extremes, test whether the signals are physically meaningful using explainable AI and atmospheric diagnostics, and investigate how these relationships may change under global warming, including possible effects linked to a weakening Atlantic Meridional Overturning Circulation (AMOC) . The opportunity is fully funded by NERC and is part of the E5 Doctoral Training Partnership . The post is aimed at candidates with a strong quantitative background in atmospheric or climate science, physics, applied mathematics, computer science, statistics, or engineering. You do not need prior expertise in both climate dynamics and machine learning; training will support skill development across both areas. Deadline: 2 December 2026, 12 noon GMT. How to apply: Review the project information on the E5 DTP website and the linked project page, then submit your application before the deadline. Share the opportunity with anyone who may be interested.