南海南部海域深海环境噪声统计特性分析

To solve the problem of the background interference caused by oceanic noise on sonar signal processing and sonar performance evaluations, in this study, we analyzed statistical characteristics such as the depth dependence, empirical fitting, and probability density function (PDF) of ambient noise in the deep sea of southern South China Sea (SCS) based on experimental observation data obtained in summer and winter. The correlation coefficients between noise levels (NLs) and wind speed or significant wave height (SWH) are provided. Using a parameterized model, we obtained empirical formulas between the NLs and wind speed, SWH or spectrum levels at 1 kHz. We propose a method of noise spectrum level PDF fitting based on Weibull and Burr distributions. The results show that the median NLs in summer is 6 dB greater than that in winter for frequencies less than 400 Hz and depths less than 800m. The PDFs of low-frequency shipping-radiated and high-frequency wind-driven NLs show agreement with Burr and Gaussian distributions, respectively. The statistical differences of the noise spectrum levels have close relationship with eddy distribution, sound channel transmission loss and distribution of ship noise sources in summer and winter.