Weather and Climate

Dr. Simon Driscoll has a background training in pure and applied mathematics, and a DPhil from the University of Oxford. Initially specialising in volcanic eruptions, stratospheric dynamics and geoengineering, he was part of the UK's Stratospheric Particle for Climate Engineering project. Accordingly he has extensive experience in atmospheric physics and climate modelling. His research has been covered in various newspapers and books around the world and he has featured in a documentary on geoengineering for VRT (Belgian national TV). Excited by the power and revolutionary potential of machine learning and AI techniques he refocused his research on machine learning methods as part of the Schmidt Sciences funded Scale Aware Sea Ice project. Based at the University of Cambridge he primarily focuses on building emulators of sub-grid scale parametrisations as well as conducting research into the AI weather forecasting models and other applications of ML and AI in weather and climate. Dr. Kieran Hunt is a NERC independent research fellow in tropical meteorology and AI at the University of Reading and National Centre for Atmospheric Science. His career has largely focused on developing and using methods to understand how extreme weather events develop over South Asia and the Himalaya, with implications for weather forecasting, water security, and energy systems. In recent years, he has shifted his focus to tackling these problems with explainable machine learning. This has led to a diverse range of applications, including: new dynamical understanding of monsoon systems, developing the first operational machine learning hydrology forecasts, vast improvements to paleoclimate modelling, and a state-of-the-art energy demand model for India. He has taught extensively and supervised numerous research projects on the applications of machine learning in weather.Dr. Laura Mansfield is a postdoctoral researcher with an interest in how machine learning and Bayesian statistics can improve climate prediction. Currently based at Stanford University, she uses machine learning to enhance the representation of subgrid-scale atmospheric gravity waves and the resulting stratospheric circulation in climate models. She develops machine learning approaches to both replace physics-based parametrisations and to aid calibration of physics-based parametrisations. She is passionate about uncertainty quantification and will soon be a Schmidt AI in Science research fellow at the University of Oxford, where she will be developing machine learning stochastic parametrisations, with the goal of estimating model uncertainty. She has a PhD from the University of Reading, where she focused on machine learning emulators of climate models to predict the climate response changing emissions. Dr Ranjini Swaminathan is a Senior Research Scientist at the Department of Meteorology and the National Centre for Earth Observation at the University of Reading. She has a Ph.D. in Computer Science from the University of Arizona and has previously been a postdoctoral research scientist at the University of Auckland and Texas Tech University. Her research expertise is in AI, focused on pattern recognition for computer vision and natural language processing applications. She is a core development team member of the UK’s flagship climate model, the UK Earth System Model (UKESM) and her research interests are in designing AI algorithms for climate model development and evaluation using observational data.Dr. Hong Wei is an Associate Professor in Computer Science. She received her first and master's degrees from Tianjin University, China, and a PhD degree from the University of Birmingham, UK. She is the author/co-author of over 100 papers and 3 textbooks. Since 2000, she has been at the University of Reading. Her research has been primarily focused on computer vision, image analysis, pattern classification, and applying machine learning to an extensive range of subjects, such as remotely sensed image interpretation, environmental monitoring, and precision agriculture.Dr. Eviatar Bach is a Lecturer (Assistant Professor) in Mathematics of Environmental Data Science at the University of Reading, after previously holding postdoctoral positions at the California Institute of Technology and École Normale Supérieure, and receiving a PhD at the University of Maryland. Eviatar's research interests include data assimilation and machine learning, predictability, and understanding climate through dynamical systems theory.Alison Peard is a nearly completed doctoral candidate at the University of Oxford’s Environmental Change Institute. Following training in Mathematical Sciences (University College Cork) and an MSc in Mathematical Modelling and Scientific Computing (University of Oxford), her research focuses on using generative AI and extreme value theory on climate reanalysis data to generate multivariate storm event sets to improve the accuracy and efficiency of large-scale coastal climate risk assessments. She has exhibited generative AI work at NeurIPS, peer-reviewed for ICLR, and her publications include work in Nature Sustainability. She has additionally worked on several national-scale climate risk assessments with UNOPS, the World Bank Group, and Oxford Infrastructure Analytics.

• 1. Introduction to Machine Learning - Fundamentals and Statistical Tools2. Introduction to Machine Learning Models3. Emulation and machine learning of sub-grid scale parametrisations4. AI/ML in weather forecasting and climate models5. XAI – explainable AI methods for understanding ML and AI models6. Generative AI in weather and climate research7. The interface of Data Assimilation and Machine Learning for Weather Forecasting8. Case studies of Machine Learning applied to Environmental Monitoring Data9. Future of ML and AI in climate and weather-related research10. References/Further Reading