Dear all,
I am working using available data from the ‘seasonal forecast’ data-set. When using the variable ‘vsw’ (volumetric soil moisture) it has a vertical dimension called soilLayer with different extent for each available model.
vsw(number, forecast_reference_time, forecast_period, soilLayer, latitude, longitude)
However, in the downloaded files (netcdf version) it is not provided the actual value in cm or m of the ‘soilLayer’, instead only the index is provided.
double soilLayer(soilLayer) ;
soilLayer:_FillValue = NaN ;
soilLayer:long_name = "original GRIB coordinate for key: level(soilLayer)" ;
soilLayer:units = "1" ;
where it should be 2-dimensional to indicate, beginning and end of each layer, or assume that it is centered to the depth of the layer
I can not compare the differences between models, because I can not ensure the correspondence in depth of the actual values.
Is there a way to know at which vertical levels corresponds each value of the variable for each GCM?
I understand that for the ECMWF data would be:
soilLayer=[(0.,7.), (7.,28.), (28.,100.), (100.,255.)] cm
Many thanks in advance
Hi @Lluis_Fita_Borrell
unfortunately the information about the layers depths is currently not available within the GRIB files which are the base of the C3S seasonal forecast archive and therefore the netCDF files (created from the archived GRIB files) don’t contain that information.
Currently users can find the information about soil layer depths for each forecast system in its description page, for instance for the recently introduced Météo-France System 9 you would have them here. Similar tables are available for each one of the forecast systems.
Unfortunately, that information is currently not available in any machine-readable format and it can be only extracted from those documentation pages.
I hope that sounds useful.
Best regards,
Eduardo Penabad
Thank you Eduardo,
I see, following your web page I tried to figure it out, and I got the following values which I share here if any one is looking for something similar (all in meters), hope I do not misunderstood something:
- ‘BOM’: [[0.,0.1], [0.1,0.35], [0.35,1.], [1.,3.]]
- ‘CMCC’: [[0., 0.0175], [0.0175,0.0451], [0.0451,0.0906], [0.0906,0.1655], [0.1655,0.2891], [0.2891, 0.4929], [0.4929, 0,8289], [0.8289, 1.3828], [1.3828, 2.2961], [2.2961, 3.8019]]
- ‘DWD’: [[0., 0.065], [0.065, 0.319], [0.319, 1.232], [1.232, 4.134], [4.134, 9.834]]
- ‘ECCC’: [[0.,0.10], [0.1, 2.8]]
- ‘ECMWF’: [[0.0, 0.07], [0.07, 0.28], [0.28, 1.], [1., 2.89]]
- ‘MeteoFrance’: [[0., 0.01], [0.01, 0.04], [0.04, 0.1], [0.1, 0.2], [0.2, 0.4], [0.4, 0.6], [0.6, 0.8], [0.8, 1.0], [1.0, 1.5], [1.5, 2.0], [2.0, 3.0], [3.0, 5.0], [5.0, 8.0], [8.0, 12.0]]
- ‘JMA’: [[0.,0.02], [0.02, 0.07], [0.07, 0.19], [0.19, 0.49], [0.49, 0.99], [0.99, 1.99], [1.99, 3.49]]
- ‘NCEP’: [[0.,0.1], [0.1, 0.4], [0.4, 1.], [1.,2]]
- ‘UKMetOffice’: [[0., 0.1], [0.1, 0.35], [0.35, 1.0], [1.0, 3.0]]
When comparing between models, one should integrate the levels from each GCM that lay within a common set of levels used as reference.
NOTE: I realize that there are different forecast systems for some models. In some cases, there use different amount of soil layers.
When downloading data from Copernicus, we only can choose from a given set of system numbers (e.g. ECMWF: 4, 5, 51) that I choose accordingly to the dates I am working with, but I can not refer to which version of the forecast system to they correspond (this could be provided as additional metadata as global attributes in the netCDF format)
Thanks
The tables in this section of the CKB (Copernicus Knowledge Base) documentation can be useful for that mapping between the GRIB “system” keyword (used also in the CDS requests) and the names used in the documentation
Hi @Eduardo_Penabad,
Thank you, this was the information I was looking for, I would need to program that!
Do you think that there could be a way that Copernicus could ‘fix’ the issue in the short/mid term?