Communicating the future: Dynamic implications of probabilistic climate forecasts

Abstract

The state of the future world is uncertain. Will it rain tomorrow? Will Donald Trump still be president in 2019? Will the world’s average temperature rise by more than 2 ºC? These questions can only be answered probabilistically, at best. The likelihood of rain may be estimated to be 50%, or somewhere between 40% and 60%. The aim of the present thesis is to investigate how laypeople perceive such uncertain predictions, especially in the climate domain. Two types of probabilistic forecasts are studied: First, how do people perceive revised forecasts, such as when an event goes from being 50% likely to 60% likely to occur? Second, how do people perceive single-bound probability estimates, such as if the chance of an outcome is said to be “more than 40%”? The results of this thesis have important implications for communicating forecasts and risks that are expressed probabilistically, in areas like climate science, medicine, weather forecasting and intelligence analysis. Communicators should be aware that receivers may read more into their forecast than they may intend.