ATLAS

Abstract

Due to technological advances in data collection and distribution methods, data is increasing in size and complexity. Challenges and opportunities have emerged in the design of data visualization tools, particularly in the visualization of geospatial and multidimensional data. Traditional visualization approaches are falling behind as they lack effective design solutions for usability issues posed by the complex relationship between spatial and numeric data. Interdisciplinary approaches are essential to address issues in geovisualization, thus the field of human-computer interaction can act as a useful lens upon the problems of data discovery within seemingly disparate fields such as climate and agricultural sciences. ATLAS is a tool for the discovery and visualization of multidimensional geospatial data (MDGSD) and aims to aid data discovery. It proposes a new approach to the visualization of MDGSD: creating a single, 'multi-modal' interface for both spatial and time-series information. Benefits and failures inherent in this approach were identified though user testing. Tests were administered to nine college students pursuing various fields of study. Preliminary analysis of data suggests that participants recognize a relationship between the spatial and time-series information; however, subjects disagreed about the significance of colors across them. Our future work aims to utilize eye tracking data to determine whether data discovery is successfully enabled in ATLAS by examining how participants visually assess and connect the data. ATLAS offers insight on the emerging opportunities of interdisciplinary work in human-computer interaction and data visualization.