Visualizing multilayer spatiotemporal epidemiological data with animated geocircles

Area of Science:

• Geographic Information Systems (GIS)
• Epidemiology
• Data Visualization

Background:

• The COVID-19 pandemic highlighted the need for effective geospatial visual analytics.
• Existing systems struggled to represent multiple, interacting, time-varying geospatial variables.
• Understanding disease spread requires analyzing complex spatiotemporal data.

Purpose of the Study:

• To investigate how epidemiologists interact with visual analytics tools.
• To develop a unified view for multiple time-varying geospatial variables.
• To assess the utility of complex spatiotemporal encodings for experts and non-experts.

Main Methods:

• Proposed an encoding method using animated, concentric, hollow circles with color encoding.
• Developed CoronaViz, a browser-based tool implementing this method.
• Conducted task-based evaluations with non-experts and interviews with epidemiologists.

Main Results:

• Epidemiologists confirmed the need for multivariate, spatiotemporal queries and found CoronaViz useful.
• Non-experts preferred animation over multi-view dashboards for spatiotemporal queries.
• The animated encoding effectively conveyed multiple variables without occlusion.

Conclusions:

• Conveying complex multivariate data involves trade-offs, necessitating complementary visualization strategies.
• Animated multivariate spatiotemporal encoding addresses crucial needs for data exploration and presentation.
• CoronaViz empowers experts and the public for future disease outbreaks and is available open-source.