alternative DJF clim

[gleckler1]

@chengzhuzhang We briefly discussed an alternative to DJF clim some months ago to the command below. We could add this alternate so that both are available. We do not want to replace the existing DJF because it is well though out, defensible, and has been used for metrics plots across 3 generations of the IPCC model evaluation chapters among other things. Do you want to add the alternate so that both are routinely available?

PCMDIobs-cmor-tables/generate_clims/PCMDIobs_clims.py

Line 93 in 3e26c29

d_djf = cdutil.DJF.climatology(d)(squeeze=1).astype('Float32')

[chengzhuzhang]

Hi @gleckler1
I was trying to understand the default behavior of DJF clim calculation of cdutil.DJF.climatology (through https://cdat.llnl.gov/documentation/utilities/utilities-1.html). According the doc, the DJF avg is deemed to be different that that is calculated using ncclimo (a tool i often used for its support for native e3sm data and good performance). I think we can leave this issue open until we see there is noticeable difference between DJF avg climos generated with different options...