CF issues related to euro cordex datasets (CF errata)

[larsbuntemeyer]

EUR-11.MIROC-MIROC5.UHOH.UHOH-WRF361H.historical.mon

this datasets seems to have wrong coordinate definitions, e.g., tas has lon, lat coordinates in rcp85 while it has rlon, rlat in historical.

dset_dict['EUR-11.MIROC-MIROC5.UHOH.UHOH-WRF361H.rcp85.mon']

[larsbuntemeyer]

EUR-11.MIROC-MIROC5.CLMcom.CLMcom-CCLM4-8-17.rcp26.mon

scenario has missing lon_vertices, lat_vertices

[larsbuntemeyer]

EUR-11.MOHC-HadGEM2-ES.MOHC.MOHC-HadREM3-GA7-05.historical.mon

historical part is missing december 2005:

[larsbuntemeyer]

EUR-11.ECMWF-ERAINT.CNRM.ALADIN53.evaluation.mon

tas units attribute is K although it's obviously degC

Already reported in errata.es-doc

[larsbuntemeyer]

it's actually all ALADIN53 datasets have units attribute K but the values are actually degC

This is also an inssue in the CDS.
Already reported in errata.es-doc

[larsbuntemeyer]

see also https://errata.es-doc.org

[jwohland]

MOHC-HadREM3-GA7-05 uses a different grid than the others (413 instead of 412 rlatsteps, shifted by about half a grid box).

[larsbuntemeyer]

thanks @jwohland for this detailed report! that's really a mess 😠

[larsbuntemeyer]

I tried to report issues to es-doc once but i failed miserably, here is the reply i got from [email protected] in case we want to report:

To get the proper format and syntax, ideally head over to your nearest ESGF data node and search for the dataset in question. You'll find the dataset identifier under the id metadata key.
For example: cordex.output.EUR-11.SMHI.MPI-M-MPI-ESM-LR.historical.r3.i1p1.RCA4.v1.day.evspsblpot.v20191001
I know it is sort of a hassle to get the correct identifier with this degree of strictness, but this is the only way automated systems can understand the different facet values.
That being said, some cordex datasets might still raise validation flags on the errata backend. This is because we rely on a vocabulary archive that needs constant updates.

[larsbuntemeyer]

ALADIN and ALARO datasets have incomplete grid_mapping definitions and wrong x and y units, e.g.

alaro = cx.tutorial.open_dataset("tas_EUR-11_ECMWF-ERAINT_evaluation_r1i1p1_RMIB-UGent-ALARO-0_v1_mon_198001-198012")
alaro

The units of x and y should be m, even if i convert those, i can not correctly plot it in it's native coordinates, since there is some unknown shift...

alaro['x'] = alaro.x * 1000.
alaro['y'] = alaro.y * 1000.

import cartopy.crs as ccrs
transform = ccrs.LambertConformal(central_longitude=9.9, central_latitude=49.0, 
                                  false_easting=0.0, false_northing=0.0, 
                                  secant_latitudes=None, standard_parallels=None, 
                                  globe=None, cutoff=-30)

import cartopy.feature as cf
import matplotlib.pyplot as plt
plt.figure(figsize=(10,8))
ax = plt.axes(projection=transform)
ax.gridlines(draw_labels=True, linewidth=0.5, color='gray',
             xlocs=range(-180,180,10), ylocs=range(-90,90,5))
alaro.tas.isel(time=0).plot(ax=ax, transform=transform)
ax.coastlines(resolution='50m', color='black', linewidth=1)
ax.add_feature(cf.BORDERS)