The distance of sound propagation in water is inversely proportional to the frequency; there for the low frequency noise generated by distance shipping contribute most to the low frequency components of the ocean ambient noise. In order to improve the performance of sonar application, understanding and predication shipping noise is important. In this study, noise data was collected by vertical line array (VLA), in South China Sea of Asian Seas International Acoustics Experiment (ASIEX) in 2001. Due to the limited access to the satellite imagery, and based on the assumption that fishing vessels in South China Sea has similar operational pattern and lighting in recent years, the night time satellite image data was acquired by Defense Meteorological Satellite Program DMSP - Operational Linescan System(DMSP-OLS) of 2005. Linear regression was analyzed between fishing light pixel distribution and low frequency noise level variation, fishing light was expressed in terms of total pixel number and total light intensity, and distance from fishing vessel to VLA was assumed by spherical and cylindrical spread. The results show high correlation between total light intensity and noise level, and cylindrical spreading is better assumption in shallow water on the continental shelf at low frequency. As a conclusion, this study successfully prove that by using the night satellite image, the low frequency fishing vessel noise can be predicted with reasonable accuracy.