Weather and large waves along the Oregon coast: atmospheric controls on a numerical wave model

Graduation date: 2017In this study, the effects of implementing different wind input or physics packages in a numerical wave model to recreate large wave conditions are explored. Three large wave events are simulated with WaveWatch III. The wind inputs which are compared are NCEP's Global Forecasting System (GFS) with 0.5 degree resolution and Climate Forecast System Reanalysis (CFSR) with 0.312 degree resolution, and the physics packages which are compared are ST2 (Tolman and Chalikov, 1996) and ST4 (Ardhuin et al, 2010). The modelled output, including spectral shape and bulk parameter time series, are compared with National Data Buoy Center buoy observations offshore of Newport, OR. The atmospheric conditions which generate these large waves include a wind feature called a coastal jet along with a distant cyclone. The energetic contribution of these simultaneously occurring atmospheric features results in a wave field characterized by bi-modal energy spectra for two events and uni-modal energy spectra for the third event. The analysis of model output includes evaluates bulk parameter time series significant wave height, mean period and mean wave direction derived from partitioned energy spectra. A consistent underestimation in wave energy emanating from the southwestern direction is found for the output associated with all model configurations. This wave energy is generated by the coastal jet. An overestimation in swell energy emanating from the northwest is also found for all model configurations. The model configuration which features the combination of CFSR winds with Ardhuin et al (2010) physics results in the best performance for the largest wave heights, with a reduction in error for overall bulk parameters as well as partitioned bulk parameters