Validating river diagnostics on non-global domains

[Alice Miller (alice-miller)]

I am currently adding rivers to the Regional Climate Model (see u-dx147) and need to initialise some dump fields on grid type 23 (STASH codes 151, 152, 153). I am using a regional domain, this causes mule/validator.py to fail:

mule.validators.ValidateError: Failed to validate
File: /home/users/alice.miller/cylc-run/u-dx147/rivers/share/cycle/19801201T0000Z/first.transplant.astart
Field validation failures:
  Fields (742,743,744)
Field is river routing diag, which is invalid for non-Global domains

Regional rivers were added at UM13.8, so river routing diagnostics should now be valid for non-global domains (see UM ticket 7880 and JULES ticket 1555). This leaves the following elif block from line 751 of mule/validators.py defunct.

 elif field.stash.grid == 23:
        # River routing grid
        if umf.fixed_length_header.horiz_grid_type != 0:
            return ["Field is river routing diag, which is invalid "
                    "for non-Global domains"]
        else:
            # Note the river routing diagnostics are restriced to a very
            # specific fixed grid; any deviation from this is incorrect
            lon_start_exp = 0.5
            lat_start_exp = -89.5
            n_row_exp = 180
            n_col_exp = 360
            lon_spacing_exp = 1.0
            lat_spacing_exp = 1.0

I think the solution may be to simply delete this section of code, but I'm not sure if anything needs to change elsewhere. Would anyone be able to advise?

[Andrew Clark (arjclark)]

Not an expert on the rivers model and definitely not familiar with what's been done to make it work for regional models but I think you'd need to keep the validation for the global model part (the else here) as that defines what's valid for a global domain and that shouldn't have changed. Whoever did the regional rivers work would need to inform what would constitute a valid regional river routing grid (it may be that the if block for horiz_grid_type != 0 just reports that its regional and continues on its merry way).

[sberthou]

Hello,
I am part of the team who implemented regional rivers.
the regional grid for rivers is the same as the atmospheric grid (rivers run on same grid as land and atmosphere), therefore I think the solution you propose ("reports it's regional and continues its merry way") would be correct.

[James Bruten (james-bruten-mo)]

Thanks both, sounds like this is reasonably simple then.
Alice, I'll get in touch about providing a build for you to use, so you're unblocked while we don't have a release

[Alice Miller (alice-miller)]

That would be great, thanks James!