In this essay for our “Chancing the Storm” series, Susan Joslyn discusses how research in cognitive psychology can uncover the thought processes behind the use and misuse of weather information. Myriad experts and practitioners of various disciplines are involved in the production of every forecast, but Joslyn outlines how the mental models each individual uses to interpret information can cause confusion, even among experts. Drawing on her research at her Decision Making with Uncertainty Lab at the University of Washington, Joslyn shows how simple phrasing choices can cause decision errors. By accounting for the thought processes behind the interpretation of information, can we improve the practice of information sharing and decision-making in uncertain weather conditions?
Understanding how decisions are made under conditions of uncertainty is one of the perennial questions of social science. While the current moment offers plenty of uncertainty with respect to a range of social and environmental dynamics, the stakes are especially high when we consider the risks brought about by the anticipation, occurrence, and aftermath of extreme weather conditions. Uncertainty pervades the trajectory of potentially disastrous events, such as hurricanes, tornadoes, and flooding, where it is typically conceived through predictions based on probabilities of an event’s physical occurrence, strength, and direction. But uncertainty also figures into how individuals and groups interpret forecasts, perceive risks, and respond to and recover from hazardous weather.
Much research exists on meteorological uncertainty in weather forecasts, how it is communicated, and how it is understood, but less attention has been paid to the ways in which other types of uncertainty emerge through the lifecycle of a hazardous weather event or across multiple events. People engaged at every stage of the process, from prediction to recovery—including atmospheric scientists, operational forecasters, emergency managers, broadcasters, and members of the public—interpret, infer, and consider uncertainty in their perceptions and decision-making. Social scientists and others need to pay more systematic attention to nonmeteorological types of ambiguity that arise from social, cultural, political conditions. How do such contexts shape understandings of a forecast or warning, perceptions of risk, and decision-making?
For instance, when there is a threat of a tornado, uncertainty abounds: Will the weather prediction models accurately anticipate rapid changes? What is the chance the tornado will form? Will a forecaster identify it as a threat? This gets even more complicated as information about the elements and characteristics of the threat are disseminated: What path will it take? How long will it last? Whether people will receive threat information, understand it and perceive it as relevant to them, and have the ability to respond are also uncertain. Even if all of these conditions are met, variables in each individual’s situation can mean that taking protective actions will result in other costs, which must be weighed in each decision.
By engaging scholars across the social sciences, “Chancing the Storm” aims to address this “other” uncertainty. These essays emerged from a panel series at the 2019 meeting of the American Meteorological Society that showcased ways in which social scientists are building a more holistic understanding of all aspects of uncertainty when extreme weather threatens and occurs, and how this understanding can be coupled with cutting-edge atmospheric science research to better characterize uncertainty. Through this series, participating authors aim to connect divergent streams of research and practitioner knowledge, help build more comprehensive theory, and provoke methodological innovation in research about uncertainty in extreme weather and the people who are touched by it. Through deeper understanding of the tacit and unseen ways that uncertainty shapes expert and public decisions, we hope to reveal places where social scientists and practitioners might intervene to improve the way we communicate extreme weather conditions and risks.
Melissa Bica’s contribution to the “Chancing the Storm” series engages crisis informatics—research on “how people use personal information and communication technology, including social media, to respond to disaster…and cope with uncertainty.” Bica draws on her research on the use of Twitter during the 2017 hurricane season as a tool for experts to communicate and for people to evaluate the uncertainty of information about potential risks of major storms. Using a human-centered data science approach, she analyzes, in both quantitative and context-specific ways, conversations on Twitter as they took place in real time—using the example of how people interpret “spaghetti plots” used by meteorologists to represent the degree and location of risk.
Writing for our “Chancing the Storm” series, Robert Soden discusses the multiple meanings and uses of “uncertainty” in understanding floods and people’s responses to information about them. Building on extensive research on the design of floodplain maps in Colorado that declare some places, but not others, at risk, Soden argues the techno-science understanding of uncertainty that these maps represent is important but limiting. To supplement this perspective, he calls for imagining uncertainty as productive (“generative”) and as socially and politically structured (“systematically produced”), drawing on examples from the floodplain mapping project.
Scott Gabriel Knowles opens our “Chancing the Storm” series with a reflection on how uncertainty—and an engagement with contingency and multicausality—has come to be embraced by historians, not least by those who study the history of disasters. Building on his own research on the history of engineering, Knowles emphasizes temporality and how disasters are both events and technological, environmental, and social processes that unfold slowly over time. Knowles also calls attention to space and scale, especially in the era of the Anthropocene, and how disaster history can make possible “a fusion of the analytical and the irrational—the graph and the story, the cost-benefit analysis and the social analysis” in ways that bring ostensibly opposing approaches together.