Simulation study on the performance of modulation power spectrum estimation algorithm for the spectrometer

Based on the principle of wave directional spectrum measurement of space borne spectrometer, simulations of signals received by spectrometer are carried out under different sea state conditions and wind speed in this paper. And four different modulation spectral estimation methods, such as periodogram method, Welch method, AR model method and minimum variance method, are used to retrieve the ocean wave spectrum. The inversion performances of various modulation spectral estimation methods mentioned above are compared. The simulation results show that for the one-dimensional wave spectrum inversion, there is no absolute superiority for the wave spectrum performance of the inversion by different modulation spectral estimation methods. And for the two-dimensional wave spectrum inversion, the periodogram method is the worst under the condition of developed ocean wave, and the inversion performances of the other three methods have no absolute superiority. For the mature wind-sea, the minimum variance method has the best inversion performance on the integral energy error and the significant wave height error, while the periodogram method has the worst inversion performance on the dominant wave direction error and the dominant wavelength error. In the swell condition, the AR model method has better performance than the other three methods. In the different sea state conditions, the inversion performance will decrease with the incidence angle. Based on these simulation results, the minimum variance method is proposed to retrieve ocean wave directional spectrum for the case that the sea surface is fully developed, and the AR model method is proposed to retrieve the ocean wave directional spectrum under swell condition.