PhD Studentship: NERC GW4+ DTP Studentship for 2024 Entry - Improving Local Weather Forecasts for Urban Areas Using Machine Learning

The GW4+ DTP consists of the Great Western Four alliance of the University of Bath, University of Bristol, Cardiff University and the University of Exeter plus five Research Organisation partners: British Antarctic Survey, British Geological Survey, Centre for Ecology and Hydrology,  the Natural History Museum and Plymouth Marine Laboratory. The partnership aims to provide a broad training in earth and environmental sciences, designed to train tomorrow’s leaders in earth and environmental science.

For eligible successful applicants, the studentships comprises:

• An stipend for 3.5 years (currently £19,237 p.a. for 2024-25) in line with UK Research and Innovation rates
• Payment of university tuition fee;
• A research budget of £11,000 for an international conference, lab, field and research expenses;
• A training budget of £3,250 for specialist training courses and expenses

Project Background

Numerical Weather Prediction (NWP) models produce forecasts that generally perform well at capturing large-scale atmospheric features. However, high-resolution weather models can still exhibit considerable forecast errors, especially over complex terrains such as urban areas. These inaccuracies negatively impact the usefulness of the forecast, and perception of forecast performance by the public. Statistical post-processing techniques can help to reduce forecast errors by training machine learning models on data sets of past forecasts and observations. This project aims to develop and apply novel post-processing techniques to improve weather forecasts, with particular focus on high-impact events over urban areas.

Project Aims and Methods

This project will explore machine learning approaches for post-processing high-resolution weather forecasts, including traditional regression-based approaches, as well as more modern techniques such as artificial neural networks. One challenge will be to develop efficient methods that can combine large amounts of data, from ensembles of model runs and observation data from a variety of different sources, including crowdsourced observations that are particularly dense in urban areas. The project will identify a number of past high-impact events, including convective thunderstorms, to demonstrate the performance of the improved forecasts. The final output of the project will be a well-calibrated and thoroughly tested post-processing model that can be used to specifically improve forecasts over urban areas under uncertain weather conditions. 

This project builds on the existing relationship between the UK Met Office Post-Processing team and the University of Exeter Statistical Science group. The successful student will have the opportunity to collaborate with Met Office scientists and be part of the pull through of research into operational weather forecasting. The student will benefit from Exeter’s expertise in Statistics, Data Science, and Machine Learning, and be part of a vibrant community of bright PhD students working on environmental problems. The student will be aligned with the Exeter Environmental Intelligence CDT, and benefit from community activities, training opportunities and collaborations. 

The supervisors encourage co-creation of research projects with prospective candidates, and are happy to adapt the project scope to better match the research interests of the student. Please email s.siegert@exeter.ac.uk to discuss your own project ideas.

Advert information