The World Weather Open Science Conference 2014, held on 16-21 August 2014 in Montreal, Canada, heard results from the German WEXICOM project, presented by Thomas Kox, project lead researcher at Institute of Meteorology of Freie Universität Berlin, Germany. See presentation: Verification of the linguistic uncertainty of warning uncertainty

The WEXICOM project (2011-2014) examines weather warnings from extreme event information to communication and action, looking inter alia at risk perception, risk analysis, media meteorology and forecast uncertainty and damage modelling. Next to FU Berlin, the national weather service DWD and the Research Forum on Public Safety and Security, also DKKV, the German Committee for Disaster Reduction which functions as IRDR National Committee, is among the research consortium partners.

The paper presented focused on a DWD survey on how the communication of weather warnings to the emergency community and how professional end-users convert the warnings into mitigation measures. “Results show that the emergency service personnel who participated in this survey generally have a good appreciation of the uncertainty of weather forecasts. Although no single probability threshold could be identified for organisations to start with preparatory mitigation measures, it became clear that emergency services tend to avoid forecast based on low probabilities as basis for their decisions.

This paper suggests that when trying to enhance weather communication by reducing the uncertainty in forecasts, the focus should not only be on improving computer models and observation tools, but also on the communication aspect, as uncertainty also arises from linguistic origins. Here, improvements are also possible and thus uncertainty might be reducible.”

The paper has been accepted by “Atmospheric Research” and is available in: http://www.sciencedirect.com/science/article/pii/S0169809514001240

In phase two of the project, WEXICOM will assess the communication and usage of uncertainties in weather warnings, with the main goal of testing the usefulness of providing users in emergency service institutions, private businesses, and the general public with probabilistic warning information and risk-based warnings.