Colors in a Multivariate Attribute Space Over TimeLuc Guillemot, UC Berkeley David O'Sullivan, UC BerkeleyHow can colors be used to unravel spatiotemporal patterns in a multivariate geographical space? Perceptually consistent color spaces such as L*a*b* or L*c*h* are well defined, but their use in qualitative cartography is still relatively rare. Furthermore, qualitative color palettes are often randomly selected and do not relate the distance between colors to degrees of difference between categories depicted on the map. This study presents a tool allowing to select colors and automatically connect them to a multivariate space. It is applied to a geodemographic map of the San Francisco Bay Area where colors for 15 clusters can be algorithmically selected to reflect similarities between clusters in the attribute space or to maximize contrast between spatially contiguous clusters. This study shows that careful consideration of a color palette and its relation to the mapped data space can assist in the visualization of complex spatiotemporal patterns.
View Slides »Mapping Demographic Change with Cartes et DonnéesKazimierz Zaniewski, University of Wisconsin-OshkoshDemographic change over time can be shown on several types of quantitative thematic maps. Some types (e.g. choropleth and bar graph maps) have been frequently used to portray spatial and temporal trends in population change and its two components, natural increase and migration. Other less popular types (e.g. nightingale and polar graph maps) can also be used for displaying demographic trends from a more complex (i.e. multidimensional) perspective. This paper discusses several cartographic techniques, available in the latest version of a French software packages Cartes et Données, for mapping population change in the United States at the state and county level.
View Slides »
How to Build a Space/Time Directory?Bert Spaan New York Public Library - NYPL Labs
The NYC Space/Time Directory will make urban history accessible through the kinds of interactive, location-aware tools used to navigate modern cityscapes. It will provide a way for scholars, students, and enthusiasts to explore New York City across time periods, and to add their own knowledge and expertise. Over the past five years, NYPL Labs - the R&D and digitization lab of the New York Public Library - has worked on many digitization, crowdsourcing, and digital cartography projects, all aiming to make the collection of the New York Public Library more accessible. The NYC Space/Time Directory, a two-year project funded by the Knight Foundation, will build on top of those efforts, and it will create new connections between previously unconnected library collections and data sources, allowing people to tell new stories about the history of New York City.
Determining current uses of cartographic animation in geography through a journal content analysis -
10 Minute TalkJoanna Merson, Arizona State UniversityAnimation offers a captivating and informative avenue for representing dynamic data in cartography. Likewise, leading cartographic research aims to improve animation use through data and user evaluation to establish best-practice guidelines. But how many of these guidelines actually reach the research community? This question is investigated using a content analysis determine how cartographic animation is used in major geography journals in the past 5 years. I specifically examine what types of animation are used, the purpose behind their use, and the congruence between the animations and the data represented. This analysis is used to examine both how cartographic animation is used outside of the cartographic research community, and if in the era of digital maps, there is a need for better facilitating methods for including animations in academic publications.
View Slides »
Beyond Choropleth Animation and Small Multiples: A Cubist Approach to Temporal Data - 10 Minute TalkAaron Dennis 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 talk focuses on ideas like using principle component and cluster analysis to derive significant trends from time series datasets and displaying these patterns cartographically. We can apply the thinking of 20th century cubist artists, who analyzed and reassembled objects from more than one viewpoint for greater context, to how we present geographic time series. Instead of dynamic animations of choropleth maps, we might use dynamic processing techniques to create more abstract and insightful maps.
View Slides »
