Computational Methodology

My research relies heavily on advanced computational methods to simulate and analyse complex systems. Here are some of the key techniques and initiatives that underpin my work:

Agent-Based Modelling (ABM)

ABM simulates the behaviour and interactions of individual agents (such as people or households) based on defined rules. By observing these interactions, ABM helps us understand emergent larger-scale patterns and behaviours. I primarily use ABM to model human dynamics in migration and disaster contexts. The Flee ABM code is an example of a framework I have used and adapted.

Uncertainty Quantification (UQ) and Sensitivity Analysis (SA)

Uncertainty Quantification (UQ) identifies and quantifies sources of uncertainty in model inputs, parameters, and structure, and assesses how these propagate to outputs. Sensitivity Analysis (SA) investigates how variations in outputs can be attributed to variations in inputs. I apply UQ and SA to my migration and disaster models to enhance their transparency and robustness.

High-Performance Computing (HPC) and Exascale Computing

Many complex simulations require significant computational resources. HPC involves using powerful, often parallel, computer systems to tackle these problems. The field is moving towards exascale computing, enabling simulations of unprecedented scale and complexity.

ExCALIBUR Programme

ExCALIBUR is a UK national research initiative focused on preparing research software and algorithms for exascale supercomputers. My engagement with ExCALIBUR connects my work to these national efforts.

SEAVEA Project

SEAVEA develops open-source software tools to facilitate Verification, Validation, and Uncertainty Quantification (VVUQ) for complex simulations, making results more reliable and actionable.