PA21A-2153
Exploratory Data Analysis Using a Dedicated Visualization App: Looking for Patterns in Volcanic Activity

Tuesday, 15 December 2015
Poster Hall (Moscone South)
Saskia M van Manen, Open University, Milton Keynes, United Kingdom and Sarada Chen, Parsons the New School, New York, NY, United States
Abstract:
Here we present an App designed to visualize and identify patterns in volcanic activity during the last ten years. It visualizes VEI (volcanic explosivity index) levels, population size, frequency of activity, and geographic region, and is designed to address the issue of oversampling of data. Often times, it is difficult to access a large set of data that can be scattered at first glance and hard to digest without visual aid. This App serves as a model that solves this issue and can be applied to other data. 
To enable users to quickly assess the large data set it breaks down the apparently chaotic abundance of information into categories and graphic indicators: color is used to indicate the VEI level, size for population size within 5 km of a volcano, line thickness for frequency of activity, and a grid to pinpoint a volcano's latitude. The categories and layers within them can be turned on and off by the user, enabling them to scroll through and compare different layers of data.
By visualising the data this way, patterns began to emerge. For example, certain geographic regions had more explosive eruptions than others. Another good example was that low frequency larger impact volcanic eruptions occurred more irregularly than smaller impact volcanic eruptions, which had a more stable frequencies. Although these findings are not unexpected, the easy to navigate App does showcase the potential of data visualization for the rapid appraisal of complex and abundant multi-dimensional geoscience data.