metpy/plots/mapping.py >> Unhandled projection: rotated_latitude_longitude

[blaylockbk]

What should we add?

Canada's HRDPS model is on a rotated latitude longitude grid. The following returns an error:

ds = # load grib file with xarray/cfgrib
ds = ds.metpy.parse_cf()
ds.metpy_crs.item().to_cartopy()

# OUT
# ValueError: Unhandled projection: rotated_latitude_longitude

Is the solution to this simply adding the ccrs.RotatedPole projection in metpy/plots/mapping.py?

MetPy/src/metpy/plots/mapping.py

Line 87 in 0b93500

raise ValueError(f'Unhandled projection: {self.name}') from None

HRDPS GRIB2 Data:
https://dd.weather.gc.ca/model_hrdps/continental/2.5km/00/000/

Reference

https://eccc-msc.github.io/open-data/msc-data/nwp_hrdps/readme_hrdps-datamart_en/#technical-grids-specification

[dopplershift]

Yes, it's a matter of:

1. Add a new function in mapping.py to turn CF metadata into a Cartopy projection. There a lot's of examples to borrow from there
2. Handle mapping the CF metadata attributes to appropriate Cartopy kwargs
3. Register it using the decorator: @CFProjection.register('rotated_latitude_longitude')

[erin6541]

I'm going to try to work on this through a PR over the next couple weeks!

[blaylockbk]

Thanks @erin6541. I started looking into this in my MetPy fork but didn't get around to testing it.