NCAR HPC Specific Setup CUPID
We will be using the NSF-NCAR High Performance Computing (HPC) Machines, Derecho and Casper. You could also use NSF-NCAR JupyterHub to work here. You need a Linux Terminal from your laptop in order to do the work. On MacOS or Linux machines you can simply using the Terminal Application. On Window's machines you'll need to download PuTTY (or something similar) or use MS VSCode to do the work (perhaps with the Remote SSH extension).
For the LDF and ILAMB work you'll login to Casper and run jobs interactively on Casper. For the CTSM case work -- you'll login to Derecho and work there.
References in YAML files that mention "cam_" just refer to your CLM data files. For example, cam_hist_loc refers to the archive directory where your case is after the short term archiver step, which will be $SCRATCH/$USER/archive/$CASE/lnd/hist.
We tried to get a setup that will just work when instructions are followed. But, this process is new and under development, so it admittedly has flaws, and you'll likely see that. We also learn more when things go wrong, so I hope you'll mistype something and experience a job or task failing because of that. The real work of modeling is figuring out how to fix it and move on. And learning is usually about playing with something new to learn how it works. So feel free to try things out, so you learn how to get it to work. Here you have someone to help you, but later when you are on your own it'll be good to have experienced how to identify problems and handle error messaging.
As part of this look over the various type of log files and output you'll see as well as your input settings and YAML file to prescribe how CUPID works. Get comfortable checking if directory locations and files exist where expected.
Next to the Installing CUPID page
Back to the Main LMWG-CUPiD Tutorial Page
