Extending bicomponent temporal trend maps: an abstracted view on geographic time series
Open Access
- Author:
- Dennis, Aaron Patrick
- Area of Honors:
- Geography
- Degree:
- Bachelor of Science
- Document Type:
- Thesis
- Thesis Supervisors:
- Cynthia Ann Brewer, Thesis Supervisor
Roger Michael Downs, Thesis Honors Advisor - Keywords:
- cartography
geography
visualization
temporal
time series
bicomponent
maps
animation
JavaScript
web - Abstract:
- In an increasingly data driven world, repeated or constant measurements over time provide time series datasets that record potentially insightful trends about the dynamics of our systems. Those datasets that focus on geographic regions or features provide the opportunity to contextualize trends in attribute measurements with their spatial relationships, history, environment, demographics, and culture. The difficulty is communicating visually the major time series trends while still providing a rich geographic context. Common approaches to cartographically presenting geographic time series include small multiples and animation. However, it is uncertain whether map-readers can effectively derive accurate and comprehensive understanding of a time series dataset from these methods. This thesis focuses on a novel technique called bicomponent temporal trend mapping and presents a JavaScript library for implementing this computationally complex method in a web based medium. Here, I also discuss how this technique can be applied to a variety of datasets and what qualities of these datasets cartographers should consider when designing bicomponent temporal trend maps. Bicomponent temporal trend maps communicate patterns in an abstract—but still objectively consistent—approach that may be a better choice than map animation or small multiples, and the technique is now accessible through this library to other cartographers.