Improving how we communicate the uncertainty of flood predictions made with probabilistic weather forecasts

Abstract:

Floods are consistently the second deadliest natural hazard in the United States. One of the reasons for this is due to flood warnings being regularly ignored by the general public. Warnings can be ignored for a multitude of reasons, but some are public distrust in predictions and a lack of comprehending the severity of the danger. It is proposed that the scientific community has failed the general public when warnings are not taken seriously. As such, we have developed the Hydrologic Risk Forecaster (Hydro RF) to improve hydrologic warning systems in remote areas of the United States. Hydro RF ingests ensemble precipitation data from operational weather forecasts from both NOAA/NWS and the USAF. The conglomerated ensemble of precipitation forces hydrologic and hydraulic HEC models to produce ensembles of stream velocity and stage at a specific river crossing. These ensembles are then used to communicate hydrologic hazards through visual aids and threshold values set per vehicle class. This communication is valuable in that it is a two-way development process. Our group has proposed what we believe is our best first step in producing a graphic to communicate hazard. The next step to this is that we are continually meeting with the eventual users of our product to hear their wants and needs directly from them. As such, Hydro RF is continually evolving to be tailored to the needs of its users, and most importantly keeping the public safe.