风浪波高和周期的联合概率密度分布

The joint probability density distribution of crest height and crest period

本文采用有代表性的44008美国国家浮标站2003年1～3月实测的海浪谱密度资料,选取三次大风过程进行风浪谱分析。通过对实测平均海浪谱与PM谱、JONSWAP谱及Torsethaugen谱的比较,得到PM谱拟和最好。然后用国际上最先进的随机波分析方法,根据协方差矩阵的循环嵌套技术,以实测平均谱与PM谱为靶谱,对随机波面进行模拟。得到由模拟波面统计的特征值及估计的谱与实测谱结果极为相近,谱峰及谱峰频率都基本一致。说明利用模拟波面研究海浪具有代表性,它可以反映实测海浪的特征。利用Longuet-Higgins(1983)模型计算波高-周期联合概率密度分布,得到变换高斯过程计算的波高、周期联合分布与实测情况基本相同,更好地描述了波高-周期联合概率密度分布。

The wave spectrum data, collected at the American National Data Buoy 44008 from January through March, 2003. Three cold air processes were selected. Comparing observed mean spectrum to PM spectrum, JONSWAP spectrum and Torsethaugen spectrum, conclude the PM spectrum fit the observed spectrum best. Then, using the spectrum analysis of random wave method, it represents the latest developments. Taking the measured spectrum and PM spectrum as the target spectrum, fast and exact simulate stationary Gaussian process through circulate embedding of the Covariance matrix. We find the estimate spectrum is very closing the measured spectrum, and peak frequency almost same. It is a feasible method based on simulated wave surface, which can reflect the wave characteristics. Using Longuet-Higgins （1983） model, calculate joint pdf of crest height and crest period. We conclude that joint pdf of crest height and crest period calculated on the transform Gaussian sea is very closing the joint pdf of crest height and crest period derived by the measured data.