I am doing a niche modelling in a global scale for a shallow water coral specie, therefore I need a good resolution for bathymetry data close to the shore in a big scale.

I am using ETOPO1, but it returns wrong results when estimating seafloor grid in shallow waters, probably because it uses nearby soundings both as constraints and to save gravity-bathymetry relationships.

So I was thinking about using a interpolation method to artificially get a higher resolution data set that represents better the coral niche. But I am not sure if this would work as I expect and what interpolation method I should use. Can anyone help me with that?