Measured Wave Crest Distributions in Central and Southern North Sea Storms

摘要

Design criteria based on wave crest height are important for offshore engineering applications, including specification of platform deck elevation. This paper aims to assess how measured individual crest elevations fit each of three theoretical probability distributions. The Rayleigh distribution has long been used to determine wave height in a narrow-banded sea. However, nonlinearities in real sea-states cause individual waves to become more peaked with increased crest amplitudes and decreased trough amplitudes which are inherently non-Rayleigh. Corrections to the Rayleigh distribution have been proposed by Jahns & Wheeler (1970) and Tung & Huang (1985). Amoco have recorded wave profile data at platforms in the Southern and Central North Sea, amounting to a total data period of approximately eight years. The locations are Leman, 38 metres depth, and Lomond and North Everest, 90 metres depth. The wave sensors employed are a Thorn EMI infrared laser at Leman and Marex wave radars at Lomond and Everest. The data were processed to reveal wave height, period and crest elevation of more than eleven million individual waves. Measured crest height distributions of this extensive data-set were compared with the two modified Rayleigh models. We conclude that, because the Tung and Huang distribution shows good consistency with the data and is based on sound theory, we recommend it for design applications.