Postdoctoral Researcher – HiSCORE: High resolution data assimilation with Spatially and temporally Correlated ObseRvation Errors

This is a postdoctoral research role to undertake research into new, numerically efficient methods for the treatment of large, dense observation datasets and their uncertainties, with the aim of improving weather forecasts via data assimilation. The project will be done in close collaboration with the European Centre for Medium-Range Forecasts (ECMWF) and the Met Office, applying the concepts in a realistic global weather forecasting system. We plan to exploit and develop state-of-the-art approaches in numerical linear algebra, in combination with physical understanding for practical benefits in hazardous weather prediction.

The role contributes to the overall goals of the Advancing the Frontiers of Earth System Prediction (AFESP) programme, both in data assimilation and across the three AFESP themes (see https://research.reading.ac.uk/earth-system-prediction/our-research/science-plan/ ).

This role is funded by the research programme on Advancing the Frontiers of Earth System Prediction (AFESP) - a £30million 15-year investment by the University of Reading, in partnership with ECMWF, the UK Met Office and the National Centre for Atmospheric Science. It will deliver sustained investments to tackle some of the far-term (10–15 year) and difficult (high-risk, high-reward) research challenges in global Earth System prediction. By enhancing our capabilities in global data assimilation, simulation and analysis, the research programme will deliver a new class of accurate, reliable and usable forecasts, aiming to increase the medium-range predictability limit and enable a wide range of new scientific and societal applications. 

Full time, fixed term post for 5 years.

Interview date 18/03/2024

Main duties and responsibilities 

Undertake collaborative research and make significant contributions to the following activities:  

• Research new data assimilation techniques that are capable of handling dense data. The research will  
• Estimate and compare satellite observation uncertainties using metrological studies and DA diagnostics from the ECMWF and Met Office systems
• Develop new numerical linear algebra approaches for large matrix-vector products and test them in idealized systems to develop a deep understanding of the methods, and their effects on the assimilation in terms of analysis accuracy at a range of scales, convergence of the minimization and computational cost
• Implement the chosen methodology in ECMWF’s NWP system and test the outcomes in terms of analysis accuracy, forecast skill and wall-clock time
• Design a strategy for maximizing the benefits of high resolution observations for global km-scale forecasting, from a range of observation types,
• Attend, contribute to, and organise relevant project meetings.
• Report on progress and results of the research through appropriate methods, including papers for submission to scientific journals, presentation of results at conferences and workshops, etc.
• Maintain awareness of current progress in relevant research areas to ensure that the research remains at the cutting edge.
• Liaise with collaborators from other parts of the programme on a regular basis, including travel to visit project partners as required.

Information on further research posts on the working on the Advancing the Frontiers of Earth System Prediction (AFESP) research programme available here

Informal contact details

Contact role: Professor of Data Assimilation

Contact name: Professor Sarah Dance

Contact email: [email protected]

Alternative informal contact details

Contact role:

Professor of Climate System Science & Climate Hazards and Director of University of Reading - ECMWF Research Collaboration

Contact name: Professor Pier Luigi Vidale

Contact email: [email protected]