Using the wave-shaped features of remote sensing images, the wavelength of ocean waves can be determined based on the wavelet method. Shallow water depths can then be estimated from the wavelength because the wavelength becomes shorter as the water depth decreases. In this paper, remote sensing data were replaced by ideal elevation data, and numerical simulation data were used to study the performance of the Complex Morlet Wavelet method in estimating wavelength and water depth. In particular, the effects of data resolution and sub-image size on water depth estimation were explored. The results from the ideal elevation data shows that: when the wavelength has no spatial change and the size of the sub-image is greater than the wavelength, the data resolution has no substantial effect on the wavelength estimation if there are more than nine evenly distributed data grids in one image. This phenomenon can be explained by the wavelength-energy spectrum. When the wavelength changes spatially, accurate estimation of the wavelength requires that the sub-image size is larger than twice the wavelength and there are four data grids in one wavelength. The estimation of wavelength by numerical simulated data requires a similar size for sub-images and the data number. The error of water depth estimation increases slightly if the sub-image size is too large, and also increases slightly as the resolution of the data decreases.