Spinup: Create NEON initial conditions files

[wwieder]

As part of the NCAR-NEON project we need to provide initial conditions (restart files for Jan-2018) for each site.

Let's start by running a SASU spinup from cold start for each site, looping over data 2018-2020.
@negin513 can you talk to @olyson for how to best do this?

We'll need to check the steady state thresholds for this, which can build on this script https://github.com/ESCOMP/CTSM/blob/master/tools/contrib/SpinupStability.ncl

This is connected to #1354.

[ekluzek]

@wwieder in order to do a SASU spinup we'll need to get one of the CN matrix branches updated to ctsm5.1.dev039. Or wait until CN matrix is on the trunk.

You probably want to start with a normal spinup and get the protocol or scripts setup to do that first. The AD spinup is the first step anyway, so the final version will just add some extra steps.

[wwieder]

Perfect. @negin513 let's just start with the 'normal' AD spinup. I typically run for 200 years in AD and 100 years in postAD. We should check pools are spun up, especially as we get data for cold sites.

[wwieder]

@jedwards4b will provide these with input_data on the next tag that gets made for the container.

The wrapper script (#1444) that @negin513 created and @jedwards4b modified will also have options to run the:

• full transient case (starting for the postAD restart file), as well as with just the
• latest data (starting from the end of the transient simulation for when new data are available from NEON).

[wwieder]

This is still TODO, using new input data and surface dataset updates, but Gordon emphasized that this may not be critical until we have 'final' datm data from NEON (which Will thinks is close or done, except for outstanding missing sites).