NWS Forecaster Perceptions of New AI Guidance for Coastal Fog Prediction

Forecasting and providing decision-support for coastal fog are challenging, in part because there are limited guidance options for predicting fog events that can have costly impacts for industries like shipping and aviation. At the same time, artificial intelligence (AI) is gaining momentum from researchers and developers as a tool for developing new guidance. In this study, we leverage these advances to explore how National Weather Service (NWS) Forecasters perceive the use of new AI guidance for forecasting coastal fog. We also specifically examine which guidance features are important for how forecasters assess the trustworthiness of new AI guidance. To this aim, we conducted virtual, structured interviews with NWS forecasters from across the Eastern, Southern, Western, and Alaskan Regions. The interviews covered the forecasters’ approaches and challenges for forecasting coastal fog, perceptions of AI and its use in forecasting, and reactions to a prototype of new AI coastal fog guidance. During the interview, the forecasters went through a self-guided review of different sets of information about the development (spin-up information, AI technique, developers of the guidance, training, and inputs) and performance (explainable AI analyses, verification, a case study, and a comparison of the new product to existing guidance) of a prototype for new non-operational coastal fog AI guidance (Kamangir et al., 2021). The forecasters then assessed how the information influenced their perception of how trustworthy the guidance was and whether or not they would consider using it for forecasting. In this presentation, we synthesize the range of the interviewed forecasters’ perceptions of AI and its use in forecasting, as well as which features of the guidance were important for their overall perceptions of how trustworthy the new AI guidance was to them. We conclude with broader implications about integrating new AI tools for forecasting and decision-support in ways that are meaningful for forecasters.