Testing Visual Synthesis of Probabilistic Forecast Information: Results from a Field Survey of Emergency Management Officials

The integration of social and behavioral science within the weather enterprise has emerged as a critical field that has provided new guidance on how to communicate probabilistic forecast information to key public safety decision makers and the public. To date, much of the research and progress has predominantly been situated around text or verbal communication, with less focus on visual or the graphical design of forecast information. Given the inherent greater complexity of probabilistic forecast information over traditional or deterministic weather forecast data, there is a need for the weather enterprise to understand how to best compose and synthesize probabilistic forecasts to facilitate the end user’s understanding. Recently, National Weather Service meteorologists collaborated with the University of Oklahoma's Institute for Public Policy Research and Analysis to deploy a survey to hundreds of emergency management officials from around the country to find their preferences, test understanding, and evaluate the effectiveness of various graphical information styles of probabilistic information. This survey evaluates three main research questions: i) When phrasing 90th Percentile values as a proxy for reasonable worst case rain amounts, does the intentional inclusion of areal context impact the risk perception of end users? ii) Does the risk perception of a user presented with commonly used categorical scales change based on the categories presented on the outlook? iii) Does the manner of visual presentation of probabilistic temperature range forecasts and probabilistic snow accumulation forecasts impact interpretation? The results, which will be discussed in this presentation, point toward potential recommendations the weather enterprise could employ to improve visual communication of probabilistic weather forecast information, and will also highlight areas for additional research.