This would be useful for modeling non-rectangular domains (e.g., single catchment, volcanic island, etc.) using uniform, raster grids.
This would require an additional node_status label value for skipping nodes that are outside of the domain.
Regarding diffusion_adi_eroder, it only supports solving for the whole grid but I guess one workaround would be to interpolate or extrapolate elevations for each row (column) from the domain boundaries to the grid borders before running the solver in each direction.
This would be useful for modeling non-rectangular domains (e.g., single catchment, volcanic island, etc.) using uniform, raster grids.
This would require an additional
node_statuslabel value for skipping nodes that are outside of the domain.Regarding
diffusion_adi_eroder, it only supports solving for the whole grid but I guess one workaround would be to interpolate or extrapolate elevations for each row (column) from the domain boundaries to the grid borders before running the solver in each direction.