Temporal visualization of boundary-based geo-information using radial projection

  • Authors:
  • Y. Drocourt;R. Borgo;K. Scharrer;T. Murray;S. I. Bevan;M. Chen

  • Affiliations:
  • Department of Computer Science and Department of Geography, Swansea University, UK;Department of Computer Science and Swansea University, UK;Department of Geography, Swansea University, UK;Department of Geography, Swansea University, UK;Department of Geography, Swansea University, UK;Department of Computer Science and Swansea University, UK

  • Venue:
  • EuroVis'11 Proceedings of the 13th Eurographics / IEEE - VGTC conference on Visualization
  • Year:
  • 2011

Quantified Score

Hi-index 0.00

Visualization

Abstract

This work is concerned with a design study by an interdisciplinary team on visualizing a 10-year record of seasonal and inter-annual changes in frontal position (advance/retreat) of nearly 200 marine terminating glaciers in Greenland. Whilst the spatiotemporal nature of the raw data presents a challenge to develop a compact and intuitive visual design, the focus on coastal boundaries provides an opportunity for dimensional reduction. In this paper, we report the user-centered design process carried out by the team, and present several visual encoding schemes that have met the requirements including compactness, intuitiveness, and ability to depict temporal changes and spatial relations. In particular, we designed a family of radial visualization, where radial lines correspond to different coastal locations, and nested rings represent the evolution of the temporal dimension from inner to outer circles. We developed an algorithm for mapping glacier terminus positions from Cartesian coordinates to angular coordinates. Instead of a naive uniform mapping, the algorithm maintains consistent spatial perception of the visually-sensitive geographical references between their Cartesian and angular coordinates, and distributes other termini positions between primary locations based on coastal distance. This work has provided a useful solution to address the problem of inaccuracy in change evaluation based on pixel-based visualization [BPC*10].