perturbation: IBM(P4) - growth of a perturbation on bluesky difference: SGI(O3800)-IBM(P4) - growth of the difference between runs on tempest and bluesky difference: IBM(AMD64)-IBM(P4) - growth of the difference between runs on lightning and bluesky
In the plot the blue curve lies on top of the red curve for times less than about .9 days, and lies on top of the black curve for times greater than about 1.1 days.
At roughly day 1.1 the perturbation curve on bluesky exhibits a single step jump in RMST of about 2 orders of magnitude. This type of jump is not observed when CAM is run with the FV dycore and the standard physics package, hence is likely due to a parameterization in the WACCM physics or chemistry packages. A package that produces an RMST change of this magnitude in response to a small perturbation cannot be reliably tested with this procedure.
The large single step jumps in the difference curves for both tempest and lightning may well be due to the same process that is causing the large one step jump in the bluesky perturbation curve. But we can't verify that without further analysis, and another possibility is that there is a problem in the ports. So the test is inconclusive.
Source tag: cam3_1_9_brnchT_2Dfix01 This tag is a branch of cam3_1_9 with 1. a workaround to allow WACCM to run with a 2D decomposition 2. a modification to the time_manager to allow setting the perpetual used by aqua-planet mode. This is necessary because WACCM_MOZART On IBM and SGI % configure -chem waccm_mozart -res 1.9x2.5 -pergro -spmd -smp On AMD64 cluster % configure -chem waccm_mozart -res 1.9x2.5 -pergro -spmd -nosmp % build-namelist \ -namelist "&camexp nsrest=0 nelapse=-2 mss_irt=0 nrefrq=0 \ nhtfrq=1 ndens=1 mfilt=145 hfilename_spec='h%t.nc' \ empty_htapes=.true. fincl1='T:I','PS:I' aqua_planet=.true. \ ncdata='/fis/cgd/cseg/csm/inputdata/atm/waccm/ic/cami_2000-07-01_1.9x2.5_L66_aqua_c051022.nc' \ perpetual_ymd=20000701 /" For the run that adds a perturbation to the initial temperature field add the namelist variable "pertlim=1.e-14" to the above namelist.