Burning by pysheds

[mvhulten]

Ideally burning and related routines are done by pysheds.

Blindly following these steps

grid = Grid.from_raster('dem.tif', data_name='dem')
grid.fill_depressions('dem', out_name='filled')
grid.resolve_flats('filled', out_name='flat')
grid.burn_streams('flat', streams='rivers.shp', burn_depth=10, out_name='burned')
grid.compute_gradient('burned', out_name='grad', spacing=(grid.dx, grid.dy))

results in

Traceback (most recent call last):
  File "/p/home/jusers/vanhulten1/jureca/.local/lib/python3.12/site-packages/pysheds/sgrid.py", line 2404, in _get_nodata_cells
    assert (isinstance(data, Raster))
            ^^^^^^^^^^^^^^^^^^^^^^^^
AssertionError

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "/p/scratch/cdetect/vanhulten1/testing/ParFlow_static-file-gen.25C18/mkslopes/create_pfl_flow_direction.py", line 16, in <module>
    grid.fill_depressions('dem', out_name='filled')
  File "/p/home/jusers/vanhulten1/jureca/.local/lib/python3.12/site-packages/pysheds/sgrid.py", line 2151, in fill_depressions
    dem_mask = self._get_nodata_cells(dem)
               ^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/p/home/jusers/vanhulten1/jureca/.local/lib/python3.12/site-packages/pysheds/sgrid.py", line 2406, in _get_nodata_cells
    raise TypeError('Data must be a Raster.')
TypeError: Data must be a Raster.

This should still be tested more carefully.

[mvhulten]

Use Niklas' burning routines.