EAMxx Resolutions, Common Terms `F2010-SCREAMv1`
Current defaults
NE | se_tstep | ATM_NCPL | phys dt (computed, don’t try to change – use ATM_NCPL) | N elements | dtt | dtr | nu_top | hvr | pgc |
---|---|---|---|---|---|---|---|---|---|
30 | 300 | 48 | 1800 | 5400 | 6 | 2 | 250000 | 6 | 1 |
120 | 75 | 96 | 900 | 86400 | 6 | 2 | 100000 | 0 | 0 |
256 | 33.33333333333 | 144 | 600 | 393216 | 6 | 2 | 40000 | 6 | 1 |
512 | 16.6666666666666 | 432 | 200 | 1572864 | 6 | 2 | 20000 | 0 | 0 |
1024 | 8.3333333333333 | 864 | 100 | 6291456 | 6 | 2 | 10000 | 6 | 1 |
The physics dt (also called dtime) is computed 86400/ATM_NCPL
, where ATM_NCPL
can be changed with xmlchange
. But note while it’s also possible to change dtime, one should not, see user should not be allowed to set dtime in user_nl_cam · Issue #509 · E3SM-Project/E3SM I added the last 2 columns here only because they were oddly different between cases and we’ve found why that’s the case. We don’t depend on ne120/ne512 for production case currently, but we would like to correct this (Mark T is aware).
dtt = dt_tracer_factor
dtr = dt_remap_factor
hvr = hv_ref_profiles
pgc = pgrad_correction
dtt=se_tstep*dtr
phys dt = 86400/ATM_NCPL
dt_tracer_factor = (time step of tracer transport) / (dynamics time step)
dt_remap_factor = (time step of vertical remap) / (dynamics time step)
se_tstep ATM_NCPL phys dt nelem dtt dtr nu_top hvr pgc
ne30 300 48 1800 5400 6 2 250000 6 1
ne120 75 96 900 86400 6 2 100000 0 0
ne256 33.33333333333 144 600 393216 6 2 40000 6 1
ne512 16.6666666666666 432 200 1572864 6 2 20000 0 0
ne1024 8.3333333333333 864 100 6291456 6 2 10000 6 1
actual resolution/compsets used:
ne30pg2_ne30pg2.F2010-SCREAMv1
ne120pg2_ne120pg2.F2010-SCREAMv1
ne256pg2_ne256pg2.F2010-SCREAMv1
ne512pg2_ne512pg2.F2010-SCREAMv1
ne1024pg2_ne1024pg2.F2010-SCREAMv1