Risk-Based Visualization of Hurricane Forecasts

Abstract

The research presented here tests user comprehension of a new kind of risk-based visualization of hurricane forecasts versus other common representations: a cone of uncertainty (with and without a centerline) and an ensemble representation. Our landfall heat maps use intensity hues and a filled black circle with size representing hurricane category at the time of forecast, rather than representing the hurricane path. In two online studies we compare how different representations of the same scenarios affect viewers’ (a) judgments of the risk that a location is likely to be hit by a hurricane, (b) their likelihood of issuing an evacuation order for that location (if they were to be in a role of issuing evacuation orders), and (c) their confidence in their judgment. Findings contribute to developing a toolkit of visual materials that designers of hurricane visualizations can select from based on empirical evidence of user responses to their elements.