Item talk:Q149520

Forecasts of coastal change hazards

Model predictions of severe storm impacts provide coastal residents, emergency managers, and partner organizations valuable predictive information for planning and response to extreme storm events. The foundation of this work is a USGS-developed numerical model to forecast storm-induced coastal water levels and expected coastal change, including dune erosion, overwash, and inundation. The model is operated in three modes: generalized scenarios, real-time storms, and an operational forecast, with each mode requiring slightly different water level inputs. To evaluate and improve the accuracy of the models, we collect data on water levels and coastal change. In particular, observations before, after, and during storm conditions are used to test the different model applications. Forecast validation for Hurricanes Matthew (2016) and Irma (2017) illustrate three cases with demonstrated forecast skill and three cases with poor skill, and reveal elements of the modeling and/or testing approach which require improvement.