This session examines how observations, NWP and forecasters’ skill may be combined in systems designed to aid in the identification of possible high-impact weather and in the production and issue of forecasts and warnings. Such systems have become increasingly automated, and there is a need to ensure the data used are an optimal combination of various model and observational sources, with proven accuracy in the forecasting of high-impact weather events. In addition, with many users adopting a risk-based approach to mitigation of the impacts of weather, warnings are increasingly becoming probabilistic in nature, and our warning systems must be able to reliably capture and communicate the uncertainty of a situation.

Presentations are invited on the themes above. Some suggested topics include:

• Automated forecasting and warning systems.
• Use of ensemble prediction systems in warning processes.
• The role of the forecaster in warning systems.
• ‘Best data’ creation from deterministic and probabilistic NWP and observations.