The marine controlled source electromagnetic (MCSEM) signals are susceptible to various kinds of noises, which affects the accuracy of the inversion interpretation of subsequent data. Waveletbased denoising theory and methods have been widely used in the field of MCSEM signal denoising. However, the applied wavelet bases are generalpurpose ones, and the denoising effect needs to be improved. This study proposes to construct a new wavelet base specially used for MCSEM signals. Firstly, through the particle swarm optimization (PSO) algorithm, the optimal coefficients of the filter banks were iteratively calculated taking the average similarity of the new wavelet function and the MCSEM signal as the constraints, and then the new wavelet base was constructed using the obtained coefficients. Secondly, aiming at the seawater turbulence noise in deep sea exploration, a denoising method based on the new waveletbase was designed. An experiment was conducted to compare the new waveletbased denoising method and the traditional waveletbased denoising method using the simulated noisy data. The signal to noise ratio (SNR) and mean square error (RSE) were used to evaluate the denoising effects, and the results demonstrate that the new waveletbased denoising method is superior to the traditional waveletbased denoising method. Finally, the new waveletbased denoising method was applied to the actually measured MCSEM data. The time domain signals and magnitude versus offset (MVO) curves before and after denoising were compared and analyzed, the results show that the proposed method can not only remove the noise of sea water turbulence, but also extend the interpretation range of the MVO curves, which proves the effectiveness and practicability of the new waveletbased denoising method.