Interactive visualization of uncertain spatial and spatio‐temporal data under different scenarios: an air quality example

Abstract

Abstract This paper introduces a method for visually exploring spatio‐temporal data or predictions that come as probability density functions, e.g. output of statistical models or Monte Carlo simulations, under different scenarios. For a given moment in time, we can explore the probability dimension by looking at maps with cumulative or exceedance probability while varying the attribute level that is exceeded, or by looking at maps with quantiles while varying the probability value. Scenario comparison is done by arranging the maps in a lattice with each panel reacting identically to legend modification, zooming, panning, or map querying. The method is illustrated by comparing different modelling scenarios for yearly NO2 levels in 2001 across the European Union. Keywords: Dynamic graphicsMapsProbability density functionCumulative density functionEnvironmental modelling Acknowledgements Much of the work for the development of the visualization tool presented here was supported by the EU‐project apmosphere (EVK2‐2002‐00577), one of the preliminary GMES projects under the 5th framework. Derek Karssenberg provided valuable suggestions. Tomislav Hengl and an anonymous reviewer are acknowledged for their helpful reviews.