Some fixes are already implemented for CESM2 models, but with the new checks it turns out that the level coordinate has wrong negative values as well as wrong units ('hPa' instead of '1') and the already implemented fixes are not enough:
netcdf cl_Amon_CESM2_historical_r1i1p1f1_gn_185001-201412 {
...
lev = -3.64346569404006, -7.59481964632869, -14.3566322512925,
-24.6122200042009, -35.9232500195503, -43.1937500834465,
-51.6774989664555, -61.5204982459545, -73.7509578466415,
-87.8212302923203, -103.317126631737, -121.547240763903,
-142.994038760662, -168.225079774857, -197.908086702228,
-232.828618958592, -273.910816758871, -322.241902351379,
-379.100903868675, -445.992574095726, -524.687174707651,
-609.778694808483, -691.389430314302, -763.404481112957,
-820.858368650079, -859.53476652503, -887.020248919725,
-912.644546944648, -936.198398470879, -957.485479535535,
-976.325407391414, -992.556095123291 ;
}
I guess the points should be recalculated using one of those formulas? I do not have much experience with these types of coordinates though.
Describe the dataset issue
Due to #598 , the CMOR checker will be able to detect issues in the hybrid sigma pressure coordinate for models that so far were passing the checks.
Some fixes are already implemented for CESM2 models, but with the new checks it turns out that the level coordinate has wrong negative values as well as wrong units ('hPa' instead of '1') and the already implemented fixes are not enough:
According to cfconventions:
The hybrid sigma-pressure coordinate for level k is defined as a(k)+b(k) or ap(k)/p0+b(k), as appropriate.I guess the points should be recalculated using one of those formulas? I do not have much experience with these types of coordinates though.
fullpath to the file(s) that are causing the problem, e.g. on CEDA Jasmin or DKRZ): /badc/cmip6/data/CMIP6/CMIP/NCAR/CESM2/historical/r1i1p1f1/Amon/cl/gn/latest/cl_Amon_CESM2_historical_r1i1p1f1_gn_185001-201412.nc