Rebecca Simpson

Testing Machine Learning Model Representation of the Climate System in the Southern Hemisphere using Rossby Waves

Rebecca Simpson is a DPhil student in the Intelligent Earth CDT at the School of Geography and the Environment, supervised by Dr Neil Hart, Professor Seth Flaxman, and Dr Mark Rodwell (ECMWF).

Her research focuses on the representation of large-scale atmospheric dynamics in machine learning weather and climate models. In particular, she investigates whether these models capture the physical behaviour of Rossby waves, which are large-scale atmospheric waves that influence global weather patterns and extremes.

Rebecca applies diagnostics to next-generation machine learning and hybrid climate models, comparing their performance with observations and reanalysis datasets. Through this work, she aims to identify where the models successfully reproduce underlying physical processes, where they deviate, and how their physical fidelity might be improved to support more reliable subseasonal predictability. She has a particular interest in the Southern Hemisphere, where sparser observations present a challenging but valuable testing ground for model evaluation.

Before joining the Intelligent Earth CDT, Rebecca completed a BA in Geography at the University of Oxford followed by three years working as a management and strategy consultant in the insurance sector.