SDB is often employed in planning hydrographic surveys for marine regions that have not been surveyed or where bathymetric data is outdated. This method has gained significant popularity in recent years, relying on analytical modeling of light penetration through the water column in visible and infrared bands. In this study, the SDB method was applied using free-of-charge Landsat-8 and Sentinel-2 satellite images to obtain bathymetric data along the North Coast of Cherchell, Algeria. The objective of this research is to compare the accuracy of SDB derived from Landsat-8 and Sentinel-2 imagery after selecting and validating the best correlation model, with the implementation of the Stumpf Empirical Algorithm. The data analysis and SDB processing procedures are described in detail. Furthermore, the processed satellite data and the resulting bathymetric maps were compared through a comprehensive statistical analysis for SDB calibration and model validation. Echosounder measurements were collected from September 15–26, 2019, with a total of 1131 depth points recorded and geolocated. Two satellite images were utilized: a Landsat-8 image from September 24, 2019, and a Sentinel-2 image from February 24, 2020. Statistical indices such as R2, RMSE, and MAE were calculated, with results indicating that Sentinel-2 images provided a better correlation between SDB-predicted values and in-situ depth data (Sentinel-2: R2 = 0.7892; Landsat-8: R2 = 0.7233, RMSE = 2.176, MAE = 4.735). Overall, the generated models demonstrate the potential to predict bathymetry on a large scale using only satellite images, highlighting the importance of spatial resolution for accurate SDB estimation.