Sentinel-2 imagery, particularly from the Sentinel-2A platform, is widely used in Earth observation due to its high spatial and temporal resolution. However, for specialised applications such as satellite-derived bathymetry (SDB), preprocessing remains a critical step. While some users assume that Sentinel-2 imagery—especially Level-2A products—is “ready to use,” the reality is more nuanced.

Level-1C data, provided as top-of-atmosphere (TOA) reflectance, clearly requires atmospheric correction prior to analysis. Level-2A data, which offers bottom-of-atmosphere (BOA) reflectance processed through Sen2Cor, reduces the atmospheric burden but may still underperform in coastal and aquatic environments due to limitations in aerosol modelling and adjacency effects. Consequently, further atmospheric correction—using tools such as ACOLITE or iCOR—is often necessary to retrieve accurate water-leaving reflectance.

Geometric and radiometric corrections are already applied to Sentinel-2 products; however, for SDB, researchers must often go beyond standard radiometric preprocessing. Factors such as sun-glint, turbidity, and variable water column absorption require additional corrections to ensure spectral consistency and depth inversion accuracy.

Therefore, while Sentinel-2 offers a robust platform, preprocessing cannot be overlooked when working with water-dominated targets. For reliable SDB results, users must select cloud-free, low-turbidity imagery and apply a tailored workflow that includes both optical and atmospheric corrections specific to shallow aquatic environments.

Yes, it requires atmospheric, cloud, and optionally topographic correction.

Of course.

Try to open a Landsat image on a Sentinel-2 image.... you will find a disaster.

Dissertations can be conducted in this research point.