LUT inconsistency, radice is the diameter of ice and not the radius

[Chiil]

While comparing the LUT of ice to that in ecRAD, we discovered that one needs to increase the radius with a factor of 2 to get the same optical properties. After a big of digging into the construction of the table, we think that where radice is used as a dimension name, this is in fact the diameter.

[RobertPincus]

Thanks @Chiil. See also the issue in the code repo.

[RobertPincus]

Let's see, we can rename things with

ncrename -v radice_lwr,diamice_lwr -v  radice_upr,diamice_upr -v radice_fac,diamice_fac rrtmgp-clouds-lw.nc
ncrename -v radice_lwr,diamice_lwr -v  radice_upr,diamice_upr -v radice_fac,diamice_fac rrtmgp-clouds-sw.nc

That leaves the attributes unchanged e.g.

ncdump -h rrtmgp-clouds-lw.nc | grep ice | grep rad
		diamice_lwr:standard_name = "radice_lwr" ;
		diamice_upr:standard_name = "radice_upr" ;
		diamice_fac:standard_name = "radice_fac" ;

so we need to

ncatted -a standard_name,diamice_lwr,m,c,diamice_lwr \
               -a standard_name,diamice_upr,m,c,diamice_upr \
               -a standard_name,diamice_fac,m,c,diamice_fac  rrtmgp-clouds-lw.nc
ncatted -a standard_name,diamice_lwr,m,c,diamice_lwr \
               -a standard_name,diamice_upr,m,c,diamice_upr \
               -a standard_name,diamice_fac,m,c,diamice_fac  rrtmgp-clouds-sw.nc