Replace dids with dids_topo in gen_topo.f90

[edoddridge]

As identified by @anton-seaice in this forum post.

Somehow the script worked anyway for GEBCO data. It only failed when I pointed it at a different bathymetry product. Implementing this change meant that the script has worked for the other datasets I've tried.

I've run the script twice (with and without the bug fix) using GEBCO as the input and compared the outputs. I get different MD5 hashes for the outputs (presumably because of the attributes holding info about the processing history?), but checked using xarray.DataArray.equals ("True if two DataArrays have the same dimensions, coordinates and values; otherwise False.") and the topography.nc files are the same.

[anton-seaice]

Thanks Ed!

I think it was conincidence it works for GEBCO.

It might be because the first variables in GEBCO are lon & lat and the first variables in the output are lon & lat, but the order is different in the dataset you were trying (DTU_DKL_6_Ker50_pre_processed)