When applying Tanioka, the bathymetry should be read from a highly resolved bathymetry file (e.g. NetCDF file), not from SeisSol geometry, which incorporates most of the time coarser bathymetry data (else the SeisSol mesh would be huge).
This is not the case currently, as the bathymetry file is not given as an argument.
https://github.com/SeisSol/SAMPLER#rasterize-seafloor-using-taniokas-method
see https://gitlab.lrz.de/samoa/displacement-converter/-/tree/thomas/tanioka_reconstruction for an example of implementation.