A climatic model of mixed waves in the ocean is considered in the form of an ensemble of states generated by the Monte Carlo method from a given multivariate distribution of elements of wind-driven waves and ripples. Simulated results are presented in terms of time- and position-dependent spectral densities. Using the wave field in the tropical Pacific and North Atlantic as an example, the method of statistical linearization is shown to be appropriate for calculating the average climatic wave spectrum and the spectrum ensemble variance. A probabilistic field model allowing for reproduction of the state ensemble of the wave surface from a given spectrum on a quasi-stationary interval and quasi-homogeneous segment is suggested.