Operational test of wave - forecasting models during the Canadian Atlantic Storms Program (CASP)

  • Date: 1987-08
  • Size (MB): 121


This report describes the work funded by the Environmental Studies Research Funds (ESRF) to set–up and test a regional spectral ocean wave model for providing real–time wave forecasting during the Canadian Atlantic Storms Program (CASP), 15 January – 15 March, 1986. The ODGP spectral Ocean Wave Model was selected for this test. The model is a fully directional spectral model which resolves the spectra into 24 directional bands and 15 frequencies. A modified version of the ODGP was developed to include shallow–water equations describing wave propagation in the CASP–OC area.

The model was driven by wind fields obtained from the CMC spectral operational weather prediction model. The CMC winds were extracted at 0.998 sigma–level, which is approximately 16 m above the ocean surface in standard atmosphere. This version of the wave model referred to as ODGP–CMC. In addition, the results from the operational version of the ODGP (i.e. ODGP–OPR which was running in real–time as part of MacLaren Plansearch/Oceanweather forecasting service) were obtained and compared to those provided by the ODGP–CMC. The ODGP–OPR version is identical to the ODGP–CMC deep–water model, but driven by winds produced from reanalysis of the NOAA LFM/NGM, pressure fields in a man–machine forecast system. Also the results from the two models were compared against the METOC wave height predictions.

A large amount of field measurements (both wind and wave) was collected from a number of locations on the U.S. east coast, Georges Bank, Scotian Shelf and the Grand Banks of Newfoundland. These data sets were used for the evaluation of both wind and wave predictions.

This study provided a comprehensive evaluation and statistical measures of the accuracy of the operational wave model when driven by two different wind fields (CMC and OPR). The CMC provided winds with a large positive bias, which grew with forecast projection time. This in turn, contributed to the large errors in wave height predictions from the ODGP–CMC model. The winds obtained from a man–machine mix procedure has shown to be superior.

In addition, a 1–D shallow–water wave model was tested within the CASP–OC area. The accuracy of this model is a function of input spectra (obtained from the deep–water model) at the end of the 1–D array, treatment of local winds(particularly offshore winds), and shallow–water propagation algorithm. When excluding the periods when the input deep–water spectra were incorrect (i.e. the last two weeks of January), the model predicted the wave height with reasonable skill.


If the following document is not accessible to you, please contact sciencebulletin@dfo-mpo.gc.ca to obtain an appropriate format, such as regular print, large print, Braille or audio version.