PhD Studentship: Incorporating Machine Learning Methods and Post-processing to Produce Optimal Weather Forecasts - NERC GW4+ DTP PhD Studentship for September 2025 Entry Ref 5407

About the Partnership

This project is one of a number that are in competition for funding from the NERC Great Western Four+ Doctoral Training Partnership (GW4+ DTP).  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 further details about the programme please see nercgw4plus.ac.uk.

Project details

For information relating to the research project please contact the lead Supervisor via f.kwasniok@exeter.ac.uk

Project Aims and Methods

Weather forecasts are usually generated by ensembles of numerical weather prediction models, each with different initial conditions to quantify the uncertainty present in atmospheric phenomena. While nowadays errors are small in large-scale synoptic variables such as geopotential height, there are still significant biases and errors in dispersion in smaller-scale local weather elements, thus necessitating the application of statistical post-processing techniques to alleviate these issues and produce accurate and well-calibrated probabilistic forecasts.

This project will develop and explore novel statistical and machine learning approaches for turning raw ensembles into probabilistic forecasts. The main interest will be on non-Gaussian variables such as precipitation, wind speed and wind gusts. Relationships between large-scale weather regimes and local scale forecast errors may be investigated and harnessed for forecast improvement. Emphasis will be on multivariate methods which consider and preserve cross-site, cross-temporal and cross-variable correlations. The project may also look at the efficient blending of forecasts from different sources and in particular combinations of physics-based models and machine learning models. General forecast performance will be assessed with a particular focus on high-impact extreme events. 
This studentship will include the opportunity of a work placement for the student at the Met Office as CASE partner. 

 Project partners

The Met Office will contribute through (i) Providing co-supervision by the Met Office supervisor Dr Gavin Evans for the duration of the project. (ii) Providing the opportunity for the student to spend time physically located at the Met Office (at least three months) during their PhD including gaining an insight into the day-to-day concerns of the post-processing teams at the Met Office. (iii) The work undertaken by the student will also have the potential to influence the operational post-processing of weather forecasts at the Met Office. 

 Training

The DTP offers funding to undertake specialist training relating to the student’s specialist area of research.

Advert information