Visualising Epidemiological Simulation Results for Decision Support

Sammanfattning

Epidemiological simulations are widely used to study how infectious diseases spread and to support decision-making during epidemics, but their outputs can be difficult to interpret, especially for users without specialised modelling knowledge. As a result, simulation systems may be perceived as a “black box”, where the final outcomes are visible but the development leading to them is unclear. This thesis investigates how interactive visualisations can improve the interpretability of compartmental epidemiological simulation results by making temporal trends, regional spread, demographic differences, and scenario differences easier to explore. To address this, a web-based prototype was developed in Python using Streamlit, combining an SEIR-based simulation model with interactive maps, line graphs, animations, demographic views, scenario comparison, and a difference map. The prototype was evaluated through a use case study and a small user study with domain experts. The results indicate that combining spatial, temporal, demographic, and comparative visualisations supports exploratory analysis of simulation results and helps users understand how outbreaks develop over time and across regions. However, the evaluation also identified limitations, including unclear colour scales, small graph text, limited numerical detail, and the need for improved interaction features such as zooming. Overall, the thesis shows that interactive visualisation can make epidemiological simulation outputs more accessible and interpretable, reducing the “black box” effect by showing not only the outcome of a simulation, but also how that outcome develops.