Ensemble Forecasting Scientist – AI & Probabilistic

ECMWF; Reading; Hybrid; GBP 60,000 - 90,000; Full time

Job summary

ECMWF is seeking a Scientist in the newly-formed Ensemble Modelling Team to advance operational ensemble prediction systems from medium-range to seasonal timescales. The role focuses on improving ensemble configurations, representing model and initial-condition uncertainty, and developing physics-based, AI-based and hybrid approaches.

The successful candidate will design experiments, test upgrades, and prepare training datasets, collaborating with data assimilation, forecast production and user

Qualifications

• Experience in Earth-system modelling and understanding of physical processes influencing predictions.
• Experience with stochastic methods used in probabilistic forecasting, including stochastic parametrisation schemes.
• Experience developing ensemble prediction systems for medium-range, sub-seasonal or seasonal timescales.
• Strong programming and data-analysis skills, experience in scientific software, HPC environments and robust configurations.

Responsibilities

• Enhance the representation of model uncertainty in IFS ensemble forecasts, including through the stochastically perturbed parametrisation scheme.
• Contribute to improving AIFS and nudged ensemble configurations, including through the preparation of training datasets and the design of calibrated ensemble systems.
• Optimise the initialization and representation of initial-condition uncertainty in ensemble forecasts across forecast lead times from days to seasons, including through singular vectors and use of initial conditions from the ensemble data assimilation system.
• Design numerical experiments, analyse results and implement robust code, while maintaining operational configurations and resolving emerging issues in the operational systems.
• Collaborate across ECMWF and with Member and Co-operating States to advance operational ensemble prediction systems, representing ECMWF where appropriate.

Skills

Analytical skills

Problem solving

Programming

Data analysis

Communication

Education

Advanced university degree

Job description

ECMWF is seeking a Scientist in the newly-formed Ensemble Modelling Team to advance operational ensemble prediction systems from medium-range to seasonal timescales. The role focuses on improving ensemble configurations, representing model and initial-condition uncertainty, and developing physics-based, AI-based and hybrid approaches.

The successful candidate will design experiments, test upgrades, and prepare training datasets, collaborating with data assimilation, forecast production and user