Task 2.1 Evidence-based development and evaluation of products and communication formats

This task will analyse which information users need, design evidence-based communication formats and empirically test how effective the developed prototypes serve those needs. In the first half of Coming Decade, we focus on the communication of expected extremes, a repeated yet unfulfilled request by users in their long-standing interactions with DWD. To thoroughly identify users’ needs, we will engage in user workshops from DWD and ClimXtreme II and rely on in-depth interview methods (e.g., cognitive task analysis), as decision processes are often not directly accessible. A follow-up survey conducted among the wider group of users will (a) ensure that the identified needs can be generalised across a sufficient number of users, and (b) identify users who need diverging climate services that will be developed during the second half of Coming Decade. To design alternative communication prototypes that are tailored to the identified user needs, we will implement existing evidence from risk communication research (for an overview, see e.g., Kause et al., 2020; Spiegelhalter, 2017; Stephens et al., 2012) in close collaboration with forecast developers. To highlight the relevance of climate information for users, the first important step is to tailor the information to their decisions, e.g. by providing cumulative risks for times and areas that are directly relevant to user decisions rather than providing all data available. Second, it is key to translate the climate information into potential impacts for users, e.g. by using examples of potential impacts for different “types” of users or use cases. Third, it is important to communicate probabilities in a transparent way to highlight that users’ current weather experience may not represent the risk of extreme, rare events that may unfold in the future. To ensure continuous user feedback, we will develop quick surveys including open feedback options. The surveys will be sent to random subsets of the users at different times during development to get repeated feedback without overfitting the needs of a particular user. Finally, we will use online experiments to empirically test and evaluate which prototypes work best—and to ensure that they are comprehensible, easy to use and meet users’s needs and preferences.

In the second half of Coming Decade, we rely on the same established procedures to develop further new climate services that are requested as highly relevant by multiple users during the initial survey. The results will allow thorough recommendations on how to communicate decadal climate predictions and their uncertainty to user-groups with similar information needs and provide a landscape of key climate services needed in the near future.

Contributors

Freie Universität Berlin
Prof. Dr. Henning Rust
MSc. Fabian Rüsenberg

Max Planck Institute for Human Development
Dr. Nadine Fleischhut

Deutscher Wetterdienst
Dr. Andreas Paxian

University of Hamburg