Underwater visible light communication (UVLC) technology based on a light-emitting diode is a type of underwater wireless communication approach that simultaneously realizes the functions of illumination and communication. UVLC can effectively solve many disadvantages of underwater acoustic communication, such as large delay, small bandwidth, and high energy consumption. However, it reduces the signal-to-noise ratio (SNR) of the receiving plane due to the attenuation of light propagation caused via various media in seawater, degrading the communication quality of the system. Additionally, the concentration of the underwater medium changes as the depth varies, causing the change in the light propagation absorption and scattering coefficient, which affects the parameter setting of the error-correcting coding. In this paper, we use the Monte Carlo approach to implement vertical UVLC channel modeling by operating on the Matlab simulation platform to obtain SNR that contributes to the design of an error-correcting coding scheme based on the polar codes, ensuring that the data transmission is reliable at different channel conditions by adjusting coding parameters. The simulation result shows that the proposed scheme can achieve error-free transmission at different SNRs by adjusting the code rate. Therefore, the proposed scheme can ensure the reliability of the UVLC system by adjusting the coding parameters obtained from the actual channel environment.