Notes on tests
All runs of CAM3.0_p1 are started on 1 September and spun up for 91 days to 1 December before any output data are used.
The following tests were done with forcing fields received from Pan Linlin:
x Test 01 - original T and VOR fields based on 5 levels; incorrect rotation due to bad metadata
Test 02 - same as 01, but with correct alignment of longitudes
x Test 03 - same as 02, but forcing fields are rotated +180 in lon
x Test 04 - same as 02, but forcing fields are rotated +187.5 in lon
Test 05 - new T and VOR fields based on 10 levels
Test 06 - ideal forcing (disk over N. Atlantic)
Test 07 - local forcing
Test 08 - new T and VOR R15L8 fields on sigma levels
Test 09 - new T and VOR R15L15 fields on sigma levels
VOR looks a lot like Test05 VOR
model blows up at day 903.89
x - based on incorrect assumptions; hence, need not be considered any further
Additional debugging was done with these cases:
PanLinLinVF - vorticity forcing only (Gritforc03)
PanLinLinTF - temperature forcing only (Gritforc03)
The above two tests were designed to reveal any significant differences between:
(a) runs with forcing on model levels constant in time and
(b) runs with forcing on sigma levels which are interpolated to model levels at each time step
The differences are insignificant.