Different types of first and second fixed surface: Why error?

[lkugler]

Hi, I tried to load a grib2 containing 'Low Cloud Cover (800 hPa - Soil)' from opendata.dwd.de.

This raises a TranslationError: Product definition section 4 has different types of first and second fixed surface.

I don't see why this should be a problem, can anyone clear things up? Thanks!

grib_dump shows

#-Unknown code table entry ()
typeOfFirstFixedSurface = 100;
#-READ ONLY- unitsOfFirstFixedSurface = 100;
#-READ ONLY- nameOfFirstFixedSurface = 100;
scaleFactorOfFirstFixedSurface = 0;
scaledValueOfFirstFixedSurface = 80000;
#-Unknown code table entry ()
typeOfSecondFixedSurface = 1;
#-READ ONLY- unitsOfSecondFixedSurface = 1;
#-READ ONLY- nameOfSecondFixedSurface = 1;
scaleFactorOfSecondFixedSurface = 0;
scaledValueOfSecondFixedSurface = 0;
topLevel = 800;
bottomLevel = 0;
shortNameECMF = ccl;
shortName = CLCL;
nameECMF = Cloud cover;
name = Cloud Cover (800 hPa - Soil);

[pp-mo]

If I understand right, what this represents is a cloud fractional coverage, for all the atmosphere between ground level and the 800 hPa level.

So I think this is a problem from Iris' point of view, because Iris data-model is all aligned with netcdf-CF rules, and CF just does not provide a means of describing what is intended here.
In CF terms this information "should" translate into a one-point vertical coordinate with a point value and bounds. However, CF expects upper and lower bound values to have common units and meaning, whereas here the correct upper bound is a pressure value, but the lower bound (ground level) can not be given as a (fixed) pressure.

So the key problem is how should this data appear in Iris ?
I can think of a few possibilities :

• use a nominal fixed pressure value for the lower surface, e.g. 1000 hPa (standard) or 100e6 (improbably large); .. or ..
• .. define a new (Iris-specific) nonstandard phenomenon identity meaning 'cloud below given level', and load with a pressure point-value only (no bounds or equal-value bounds)
• and (optionally) add case-specific metadata to track metadata adjustments

Such solutions can be potentially be implemented by extending iris-grib. However, a user-code-specific solution is obviously more flexible as you don't need to resolve the core problem in a generally acceptable way.
You can modify low-level data before translating, with a callback function or via iris_grib.load_pairs_from_fields (see docs).

It really depends on what you then need to do with the data.
If you just needs a means of loading (working around the load error), I think I would just set the type and value of the second fixed surface to that of the first in this case, using a load callback.
That will yield a cube that apparently has a specific cloud fraction value between levels of 800 to 800 hPa, which is nonsensical but at least no original information is lost.

What kind of solution do you need ?

[marfel]

Just wanted to note that I just ran into the exact same problem. For me, it would be enough to change the second fixed surface to look like the first, as you suggested. How would such a callback look like?