AuroraMap

Brief

AuroraMap is a research visualization method for mapping multivariate 2D spatial distributions. It projects clusters from a spatial distribution to a boundary band and color-encodes their densities so dense internal structures can be read without filling the main map area.

Problem

Multidimensional spatial data can be difficult to compare when multiple variables, clusters, and complex geographic boundaries compete for screen space. Traditional encodings often obscure cluster density or make it hard to compare several spatial distributions in one view.

Approach

The work proposed an abstract contextual visualization that moves cluster density information to the boundary of spatial distributions. This keeps the central spatial shape readable while preserving cues about cluster location, scope, and density.

Highlights

• Proposed a boundary-homographic projection for encoding cluster structure around spatial distributions.
• Used continuous color encoding to communicate cluster densities.
• Evaluated whether users could detect relative cluster locations, cluster scopes, and peak density values.
• Demonstrated the method with real-world religious infrastructure data and multivariate geospatial comparisons.

Outcome

The research contributed a visual strategy for saving map space, supporting quantitative cluster-density estimation, and comparing multiple spatial distributions in one graph.

Links

Watch IEEE Vis AuroraMap
Read the Purdue thesis record

Stack

Data visualization research, spatial distributions, density encoding, visual analytics, geospatial visualization.