Determine which diagnostics to include for Sea Ice component

[TeaganKing]

Provide diagnostics from the sea ice component to be included in #3

[dabail10]

• Compute climatological averages for annual and seasonal (JFM, AMJ, JAS, and OND). These should allow for arbitrary start and end years. This average file will include all variables.

This actually works using the ADF.

• Compute hemispheric (northern and southern) sums of sea ice volume, snow on sea ice volume, and ice area. These six quantities will be monthly mean timeseries. (Different from issue Single variable time series generation #18)

• Compute sums of sea ice volume, snow on sea ice volume, and ice area for regions. See mask files in /glade/campaign/cesm/development/pcwg/ice/data.

• Plot NH and SH timeseries data for a single run or two runs. There are seven main plots here.

• The monthly mean timeseries for NH and SH

• The monthly mean timeseries versus PIOMAS (only for historical runs)

• Climatological seasonal cycle versus NSIDC extent.

• Annual and seasonal mean timeseries.

• Annual and seasonal mean timeseries versus CESM1-LENS.

• Annual and seasonal ice volume versus PIOMAS.

• Annual and seasonal anomaly plots.

• Same timeseries plots, but for each masked region.

• Annual and seasonal mean polar projection contour plots. Single run or difference between two runs. These are either two panel plots for a single run (NH and SH) or three panel plots (two runs and difference). NH and SH separated for differences. Includes the following variables: sea ice concentration (aice), sea ice volume (hi), snow cover (hs), snow/ice surface temperature (Tsfc), snow / ice broadband albedo (albsni), latent heat flux (Flat), sensible heat flux (Fsens), salt flux (fsalt), basal ice growth (congel / growb), frazil ice growth (frazil), snow-ice conversion (snoice), basal ice melt (meltb), top ice melt (meltt), lateral ice melt (meltl), ice volume tendency due to thermodynamics (dvidtt), ice volume tendency due to dynamics (dvidtd), ice area tendency due to thermodynamics (daidtt), and ice area tendency due to dynamics (daidtd).

• Annual and seasonal mean polar projection vector plots. Single run or difference between two runs. These just used the uvel, vel velocity components. Speed is contoured as well.

• Notes:

1. When the two runs are on the same grid, the difference plots are very straightforward.
2. When the two runs are on different grids, they both get mapped onto a common 1x1 degree grid for the differences.
3. Contour bar labels are set based on standard deviations and customized for each variable. We do print the absolute min and max for the difference plots, but only show differences between -2sigma to 2sigma. Generally we try to maximize the ranges to highlight the horizontal gradients.
4. For high resolution (0.1-degree) we make slight tweaks to the plots. Mostly, we use the TriangularMesh in NCL and subsample the velocities.

For an example of all of the plots go here:

Single case plots:
https://webext.cgd.ucar.edu/GIAF/CESM2.3.alpha10b.TL319_t061m/ice/CESM2.3.alpha10b.TL319_t061m-obs/yrs1981-1985/

Difference plots:
https://webext.cgd.ucar.edu/B1850/b.cesm3_cam058_mom_e.B1850WscMOM.ne30_L58_t061.camdev_cice6.026i/ice/b.cesm3_cam058_mom_e.B1850WscMOM.ne30_L58_t061.camdev_cice6.026i-b.cesm3_cam058_mom_e.B1850WscMOM.ne30_L58_t061.camdev_cice5.026g/yrs1-33