(CLM-only) Add a pftdyn test with constant pft's to give same answers as constant pft case

[ekluzek]

Erik Kluzek < erik@ucar.edu > - 2008-10-23 11:10:11 -0600
Bugzilla Id: 845
Bugzilla CC: dlawren@ucar.edu, lawrence@ucar.edu, muszala@ucar.edu, oleson@ucar.edu, slevis@ucar.edu,

Suggestion from Sam...

Keith and I confirmed in our own recent use of the clm test-suite that dyn. landuse tests have been passing (as Erik said, too). This includes multi-year point simulations and few-day global simulations, all with arbitrary initial conditions. Keith suggested trying a longer global run, but also a run in which the dyn. landuse data remains the same every year to show that this gives the same answers as a run without dyn. landuse.

Erik, couldn't the latter could be added to the test suite because it's an objective test?

[ekluzek]

Erik Kluzek < erik@ucar.edu > - 2008-12-10 15:44:33 -0700

I'd like to do this, but I'd like to hold off until I have a better grasp of pftdyn file creation and we have more files to actually test with. So I'm going to lower the priority until later.

[ekluzek]

Erik Kluzek < erik@ucar.edu > - 2013-08-29 15:24:15 -0600

Good idea. Let's think about later.

[billsacks]

Maybe do a single-point smallville test where we construct the landuse_timeseries file to be constant in time, and make a new test type that compares that with a run that doesn't have a landuse_timeseries file.