Numerical geodynamo modelling focused on running, managing, and analysing large-scale spherical dynamo simulations
on high-performance computing systems, with emphasis on reproducibility, physical robustness, and end-to-end data
processing from raw model output to decision-ready scientific results.
Modelling and Simulation Experience
Designed, configured, and ran three-dimensional spherical geodynamo simulations on national and university high-performance computing facilities.
Executed long-duration, multi-stage simulations using batch-scheduled HPC environments.
Worked with complex, parameter-rich numerical models spanning multiple dynamical regimes.
Maintained and extended modelling codebases obtained via version-controlled repositories.
Managed large simulation ensembles rather than isolated single-model studies.
High-Performance Computing and Software
Ran production simulations on UK HPC systems including ARCHER2 and ARC4.
Prepared and maintained machine-specific build and runtime environments.
Worked with batch schedulers, job arrays, restarts, and storage constraints.
Used Linux-based workflows and shell scripting for automation and orchestration.
Integrated modelling and analysis across multiple systems and architectures.
Model Output Processing and Data Pipelines
Extracted spherical harmonic Gauss coefficients from raw simulation output.
Developed reproducible pipelines to transform raw model files into analysis-ready datasets.
Applied consistent truncation and conditioning strategies to remove transients and unstable phases.
Converted large time-dependent outputs into standardised, structured data products.
Combined scripting (shell, Python) and MATLAB for post-processing and analysis.
Quantitative Analysis and Diagnostics
Computed magnetic and kinetic energy budgets from model output.
Calculated ohmic dissipation and ohmic fraction as key physical diagnostics.
Produced time-averaged and time-resolved summaries of dynamo behaviour.
Ensured derived quantities reflected statistically stable regimes rather than numerical artefacts.
Supported ensemble-level comparison across large numbers of simulations.
Reproducibility, Documentation, and Quality Assurance
Developed and maintained detailed procedural documentation for the full modelling workflow.
Ensured simulations and analyses were fully reproducible and auditable.
Version-controlled analysis scripts and maintained consistent data structures.
Enabled new simulations to be added to existing analysis frameworks with minimal friction.
Prioritised transparency and governance in computational research workflows.
Integration with Research Outputs
Integrated modelling results into peer-reviewed research publications.
Supported comparative and statistical studies across simulation ensembles.
Produced figures, summaries, and diagnostics suitable for publication and presentation.
Communicated modelling results clearly to both specialist and non-specialist audiences.