The shape, period and wavelength of high storm waves

It has been known for nearly 30 years that high waves in a linear sea have shapes which tend to that of the autocorrelation function. This fact forms the basis of the Shell New Wave Theory, but its acceptance by the offshore industry has been slow. For the present paper, 1000 random wave records have been simulated from a JONSWAP spectrum, and the highest wave in each record examined. The results are used to demonstrate the validity of the concept and its limitations. The statistics of the period and length of these individual high waves are examined together with the relationship between them. Their mean zero-upcross period does not tend to T-1 as has sometimes been stated, and in fact cannot be stated simply in terms of spectral moments.