Abstract: The complex noise in multibeam water plume image is one of the main obstacles affecting plume target recognition. Aiming at the complex noise interference and denoising problems in multibeam water column image, we analyzed the distribution characteristics of the main noise in water column image, and then compared the effects of the current main threshold method, background subtraction method and symmetrical subtraction method in the suppression of plume noise, for which the simulation data of water plume and the measured data of deep sea were combined. Results show that the threshold method had good applicability and compatibility, but it could not remove the seabed echo. The background subtraction method could effectively eliminate the environmental noise and seabed interference, but it was not enough to suppress the sidelobe noise. The symmetrical subtraction method could suppress the noise of acoustic scattering layer well, but it might destroy the target shape across the central beam, which can thus be used for preprocessing. Therefore, a variety of methods shall be integrated to better suppress the noise of multibeam water column image upon the work demand.