Analysis of the significantly statistical duration to clarify the uncertainty of ambient noises due to environmental changes

摘要

The ambient noise is very important in prediction systems of sonar performance, because it determines the detection ranges always in passive sonar and usually in active sonar. In the uncertainty issue for sonar performance, it's necessary to know this factor's statistical characteristics that are only obtained by data processing from the measurement data. So this study first presents the statistically analyzed results from the ambient noise data in the South China Sea experiment of ASIAEX (Asian Seas International Acoustics Experiment) in 2001. The vertical line array was deployed on May 2-18 in a depth of 124 m on the continental-shelf. The statistical results include the mean and standard deviations of noise levels in both of high and low frequency bands. These statistics are then correlated with the environmental factors, such as surface wind speeds, wave heights and shipping densities. The correlations of noise levels and environmental factors are obtained in various periods of one hour, 4 hours, and 24 hours. But the time spans of periods are selected subjectively; it may neglect factors of shorter durations in such processing. This paper presents the analysis of the statistical duration for various factors affecting ambient noise, such as high/low sea states, different frequency bands, and noise bandwidth. The probability density function of noise levels is assumed a normal distribution in the test model, so the effective confidence level is used for testing and determining the significant statistical duration (i.e. confidence time). Finally, the results provide the uncertain characteristics of noise levels in the South China Sea. For example, the confidence time frequently occurs few hours are noise level in 1k Hz at low sea states and in 100 Hz. The confidence time of noise level in 1k Hz may arise over few days at high sea states.