Probability of... What, Exactly? Establishing a Simple Ground Rule for the Communication of Forecast Uncertainty

Improving the communication of forecast uncertainty has become a high priority within the National Weather Service, and the Weather Enterprise as a whole. Numerous efforts are currently underway to produce new products that convey the uncertainties of various forecast attributes and across hazard suites, including for hurricanes, flooding, severe weather, and winter weather. These endeavors are becoming centralized somewhat under large research initiatives, including the FACETs program. From the vantage of this program, new insights are emerging about the need for a common conceptual framework and language that can be used to understand what, exactly, we mean when we say we’re communicating forecast uncertainty. Active debates have arisen about bringing together different technologies, understanding what they mean as a whole, and how best to communicate them collectively. For example, is it enough to simply say "there is a 20% chance of tornadoes" to a broadcast audience? This presentation will give an example where this forecast, coming from two different communicators that use different spatiotemporal domains, could be problematic.

This presentation will provide a brief overview of contemporary challenges in this area, and propose a simple recommendation for communicating forecast uncertainty to make the media environment as unambiguous as possible while more products enter the market.