A real-time forecasting system to predict coastal storm impacts

Brandy N. Armstrong, John C. Warner, Richard Signell

U.S. Geological Survey, Coastal and Marine Geology Program, Woods Hole Science Center, Woods Hole, MA 02543

ABSTRACT

In the coastal zone, storms are one of the primary environmental forces causing coastal change. These discrete events often produce large waves, storm surges, and flooding, resulting in coastal erosion. In addition, strong storm-generated currents may pose threats to life, property, and navigation. The ability to predict these events, their location, duration, and magnitude allows resource managers to better prepare for the storm impacts as well as guide post-storm survey assessments and recovery efforts.

As a step towards increasing our capability for prediction of these events we have developed an automated system to run the Coupled Ocean – Atmosphere – Wave – Sediment Transport (COAWST) Modeling System as a daily forecast. Management of the system is controlled by the Windows Scheduler to start Matlab® and run scripts and functions. Data required by the modeling system include daily modeled wave, wind, atmospheric surface inputs, and climatology fields. The Unidata Internet Data Distribution/Local Data Manager is used to download NCEP GFS global 5 degree data and NCEP NAM Conus 12km data to a local server. The Matlab “structs” tool and NJ-Toolbox are used to access these large data sets on the local server as well as data sets available on the Nomads http://nomads.ncep.noaa.gov site. The data are used to create the required inputs for the WRF, ROMS and SWAN models. Currently the output from the coupled ROMS-SWAN system is displayed at http://woodshole.er.usgs.gov/project-pages/cccp/public/COAWST.htm. We will describe the modeling system, data acquisition, and visualization methods required for the forecasting system.