Plain-language explanation.
Mathematical modelling is the process of using mathematics to represent, analyse, and predict the behaviour of real-world systems. It bridges pure mathematics with applications in science, engineering, economics, biology, and policy.
Core concepts and standard treatment.
Core mathematical modelling covers the modelling cycle (problem formulation → model construction → analysis → interpretation → validation → refinement), deterministic models (differential equation models — population dynamics, SIR epidemics, Newton's cooling), optimisation models (linear programming — simplex method; integer programming; network optimisation), statistical models (regression, time series, Bayesian models), and discrete models (Markov chains, cellular automata, agent-based models). Model assumptions, limitations, and sensitivity analysis are central to good modelling practice.
Deeper theory, debates and edge cases.
Advanced mathematical modelling covers multi-scale and multi-physics modelling (coupling micro/meso/macro scales — materials science, biology), stochastic modelling (SDEs, Monte Carlo methods, Gillespie algorithm for biochemical reaction networks), control theory (optimal control — Pontryagin's maximum principle, LQR; model predictive control — MPC), simulation methods (finite element, finite volume, spectral element — COMSOL, ANSYS; agent-based — NetLogo, Mesa), and uncertainty quantification (UQ — sensitivity analysis, surrogate modelling, polynomial chaos expansion).
How it is applied in practice.
At the applied mathematician, operations researcher, and simulation engineer level, mathematical modelling is the core deliverable: climate model development (Earth System Models — CESM, UKESM; parameterisation of sub-grid processes); pandemic modelling for policy (SAGE advisory models — SEIR + age structure + vaccination; scenario uncertainty — SPI-M-O consensus approach); supply chain optimisation (MILP formulations — Gurobi, CPLEX; discrete event simulation — AnyLogic, SimPy); financial risk modelling (Monte Carlo VaR, CVA, stress testing — Basel III); and structural engineering simulation (FEA — Abaqus, ANSYS Mechanical).